National Academies Press: OpenBook

Hydroacoustic Impacts on Fish from Pile Installation (2011)

Chapter: Hydroacoustic Impacts on Fish from Pile Installation

Page 1
Suggested Citation:"Hydroacoustic Impacts on Fish from Pile Installation." National Academies of Sciences, Engineering, and Medicine. 2011. Hydroacoustic Impacts on Fish from Pile Installation. Washington, DC: The National Academies Press. doi: 10.17226/14596.
×
Page 1
Page 2
Suggested Citation:"Hydroacoustic Impacts on Fish from Pile Installation." National Academies of Sciences, Engineering, and Medicine. 2011. Hydroacoustic Impacts on Fish from Pile Installation. Washington, DC: The National Academies Press. doi: 10.17226/14596.
×
Page 2
Page 3
Suggested Citation:"Hydroacoustic Impacts on Fish from Pile Installation." National Academies of Sciences, Engineering, and Medicine. 2011. Hydroacoustic Impacts on Fish from Pile Installation. Washington, DC: The National Academies Press. doi: 10.17226/14596.
×
Page 3
Page 4
Suggested Citation:"Hydroacoustic Impacts on Fish from Pile Installation." National Academies of Sciences, Engineering, and Medicine. 2011. Hydroacoustic Impacts on Fish from Pile Installation. Washington, DC: The National Academies Press. doi: 10.17226/14596.
×
Page 4
Page 5
Suggested Citation:"Hydroacoustic Impacts on Fish from Pile Installation." National Academies of Sciences, Engineering, and Medicine. 2011. Hydroacoustic Impacts on Fish from Pile Installation. Washington, DC: The National Academies Press. doi: 10.17226/14596.
×
Page 5
Page 6
Suggested Citation:"Hydroacoustic Impacts on Fish from Pile Installation." National Academies of Sciences, Engineering, and Medicine. 2011. Hydroacoustic Impacts on Fish from Pile Installation. Washington, DC: The National Academies Press. doi: 10.17226/14596.
×
Page 6
Page 7
Suggested Citation:"Hydroacoustic Impacts on Fish from Pile Installation." National Academies of Sciences, Engineering, and Medicine. 2011. Hydroacoustic Impacts on Fish from Pile Installation. Washington, DC: The National Academies Press. doi: 10.17226/14596.
×
Page 7
Page 8
Suggested Citation:"Hydroacoustic Impacts on Fish from Pile Installation." National Academies of Sciences, Engineering, and Medicine. 2011. Hydroacoustic Impacts on Fish from Pile Installation. Washington, DC: The National Academies Press. doi: 10.17226/14596.
×
Page 8
Page 9
Suggested Citation:"Hydroacoustic Impacts on Fish from Pile Installation." National Academies of Sciences, Engineering, and Medicine. 2011. Hydroacoustic Impacts on Fish from Pile Installation. Washington, DC: The National Academies Press. doi: 10.17226/14596.
×
Page 9
Page 10
Suggested Citation:"Hydroacoustic Impacts on Fish from Pile Installation." National Academies of Sciences, Engineering, and Medicine. 2011. Hydroacoustic Impacts on Fish from Pile Installation. Washington, DC: The National Academies Press. doi: 10.17226/14596.
×
Page 10
Page 11
Suggested Citation:"Hydroacoustic Impacts on Fish from Pile Installation." National Academies of Sciences, Engineering, and Medicine. 2011. Hydroacoustic Impacts on Fish from Pile Installation. Washington, DC: The National Academies Press. doi: 10.17226/14596.
×
Page 11
Page 12
Suggested Citation:"Hydroacoustic Impacts on Fish from Pile Installation." National Academies of Sciences, Engineering, and Medicine. 2011. Hydroacoustic Impacts on Fish from Pile Installation. Washington, DC: The National Academies Press. doi: 10.17226/14596.
×
Page 12
Page 13
Suggested Citation:"Hydroacoustic Impacts on Fish from Pile Installation." National Academies of Sciences, Engineering, and Medicine. 2011. Hydroacoustic Impacts on Fish from Pile Installation. Washington, DC: The National Academies Press. doi: 10.17226/14596.
×
Page 13
Page 14
Suggested Citation:"Hydroacoustic Impacts on Fish from Pile Installation." National Academies of Sciences, Engineering, and Medicine. 2011. Hydroacoustic Impacts on Fish from Pile Installation. Washington, DC: The National Academies Press. doi: 10.17226/14596.
×
Page 14
Page 15
Suggested Citation:"Hydroacoustic Impacts on Fish from Pile Installation." National Academies of Sciences, Engineering, and Medicine. 2011. Hydroacoustic Impacts on Fish from Pile Installation. Washington, DC: The National Academies Press. doi: 10.17226/14596.
×
Page 15
Page 16
Suggested Citation:"Hydroacoustic Impacts on Fish from Pile Installation." National Academies of Sciences, Engineering, and Medicine. 2011. Hydroacoustic Impacts on Fish from Pile Installation. Washington, DC: The National Academies Press. doi: 10.17226/14596.
×
Page 16
Page 17
Suggested Citation:"Hydroacoustic Impacts on Fish from Pile Installation." National Academies of Sciences, Engineering, and Medicine. 2011. Hydroacoustic Impacts on Fish from Pile Installation. Washington, DC: The National Academies Press. doi: 10.17226/14596.
×
Page 17
Page 18
Suggested Citation:"Hydroacoustic Impacts on Fish from Pile Installation." National Academies of Sciences, Engineering, and Medicine. 2011. Hydroacoustic Impacts on Fish from Pile Installation. Washington, DC: The National Academies Press. doi: 10.17226/14596.
×
Page 18
Page 19
Suggested Citation:"Hydroacoustic Impacts on Fish from Pile Installation." National Academies of Sciences, Engineering, and Medicine. 2011. Hydroacoustic Impacts on Fish from Pile Installation. Washington, DC: The National Academies Press. doi: 10.17226/14596.
×
Page 19
Page 20
Suggested Citation:"Hydroacoustic Impacts on Fish from Pile Installation." National Academies of Sciences, Engineering, and Medicine. 2011. Hydroacoustic Impacts on Fish from Pile Installation. Washington, DC: The National Academies Press. doi: 10.17226/14596.
×
Page 20
Page 21
Suggested Citation:"Hydroacoustic Impacts on Fish from Pile Installation." National Academies of Sciences, Engineering, and Medicine. 2011. Hydroacoustic Impacts on Fish from Pile Installation. Washington, DC: The National Academies Press. doi: 10.17226/14596.
×
Page 21
Page 22
Suggested Citation:"Hydroacoustic Impacts on Fish from Pile Installation." National Academies of Sciences, Engineering, and Medicine. 2011. Hydroacoustic Impacts on Fish from Pile Installation. Washington, DC: The National Academies Press. doi: 10.17226/14596.
×
Page 22
Page 23
Suggested Citation:"Hydroacoustic Impacts on Fish from Pile Installation." National Academies of Sciences, Engineering, and Medicine. 2011. Hydroacoustic Impacts on Fish from Pile Installation. Washington, DC: The National Academies Press. doi: 10.17226/14596.
×
Page 23
Page 24
Suggested Citation:"Hydroacoustic Impacts on Fish from Pile Installation." National Academies of Sciences, Engineering, and Medicine. 2011. Hydroacoustic Impacts on Fish from Pile Installation. Washington, DC: The National Academies Press. doi: 10.17226/14596.
×
Page 24
Page 25
Suggested Citation:"Hydroacoustic Impacts on Fish from Pile Installation." National Academies of Sciences, Engineering, and Medicine. 2011. Hydroacoustic Impacts on Fish from Pile Installation. Washington, DC: The National Academies Press. doi: 10.17226/14596.
×
Page 25
Page 26
Suggested Citation:"Hydroacoustic Impacts on Fish from Pile Installation." National Academies of Sciences, Engineering, and Medicine. 2011. Hydroacoustic Impacts on Fish from Pile Installation. Washington, DC: The National Academies Press. doi: 10.17226/14596.
×
Page 26

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Research Results Digest 363 Responsible Senior Program Officer: Christopher J. Hedges October 2011 HYDROACOUSTIC IMPACTS ON FISH FROM PILE INSTALLATION Bridges, ferry terminals, and other structures commonly have driven-pile foundations, and pile driving can cause effects on fish ranging from altered behavior, hearing loss, and tissue injuries to immediate mortality. The objec- tive of NCHRP Project 25-28 was to develop guidelines for the prediction and mitigation of the negative impacts on fish from underwater sound pressure during pile and casing installation and removal. The research was con- ducted by a team comprised of researchers from the University of Mary- land and Battelle–Pacific Northwest Division. The Principal Investigators were Michele B. Halvorsen, Thomas J. Carlson, and Arthur N. Popper, assisted by Brandon M. Casper and Christa M. Woodley. SUMMARY Introduction and Background As more pile driving activity occurs, there is an increased concern about its potential effects on fishes and other aquatic organisms. The possibility of effects on fishes rises as more offshore wind farms are installed around the United States and other nations in addition to the ongoing infrastruc- ture and industrial maintenance and devel- opment activities, such as those performed by transportation agencies throughout the United States. Effects on fishes potentially associated with pile driving include damage to body tissues that could result in death, as well as impacts on behavior that could cause fishes to leave sites of biological importance (e.g., feeding, spawning). The goal of this study was to provide quantitative data that may be used to define criterion levels for tissue damage onset and then use these criteria in design of future pile driving projects with options for pro- tection of animals. Regulations for pile driv- ing on the U.S. west coast currently utilize a dual interim criteria approach for onset of physiological effects. These criteria include a cumulative sound exposure level (SELcum) of 187 dB re 1 μPa2s for fishes more than 2 grams and 183 dB re 1 μPa2s for fishes less than 2 grams, and a single-strike peak level (SPLpeak) of 206 dB re 1 μPa for all sizes of fishes (Stadler and Woodbury, 2009). If either the SELcum or SPLpeak are exceeded, mitigation protocols should be applied. Field research on effects of pile driving is difficult to execute, and researchers have not had control over the pile driving expo- sures (frequency of strikes, intensity, dura- tion, and other parameters) nor, in many cases, the physiological state of test fishes during exposure. Thus, it was critical in this study to design an experiment in which researchers had control of all experimental parameters in order to investigate which variables play a role in tissue damage caused by barotrauma. Barotrauma occurs when there is a rapid change in pressure that directly affects the body gases. Free gas in the swim bladder, blood, and tissue of fishes along with gas in solution in blood and other fluids can respectively experience a change in volume and state (e.g. expansion and contraction and/or bubble formation and C O N T E N T S Summary, 1 Chapter 1 Background, 4 Potential Effects of Pile Driving Sounds on Fish, 5 Study Rationale, 5 Study Goal and Objectives, 7 Previous Studies Using Similar Equipment, 7 Chapter 2 Methods, 7 Study Fish, 7 Fish Maintenance, 8 Sound Exposure Apparatus and Methods, 8 Sounds, 8 HICI-FT Sound Control Operation, 9 General Experimental Procedures, 10 Barotrauma Analysis and Characterization, 10 Response Variable Derivation, 10 Statistical Analysis, 12 Chapter 3 Results, 12 Sound Exposures, 12 Barotrauma, 12 Chapter 4 Discussion, 17 Overview, 17 Rejection of the Equal Energy Hypothesis for Pile Driving, 18 Relationship Between Pile Driving Exposure and Biological Response, 18 Chapter 5 Application to Pile Driving Projects, 19 Background, 19 Derivation of Exposure Criteria, 20 Chapter 6 Future Research, 21 References, 23 Acronyms, 24 Appendices A Through H, 24 NATIONAL COOPERATIVE HIGHWAY RESEARCH PROGRAM

2absorption) during rapid pressure changes, which can lead to tissue damage, organ failure, and changes in behavior. Experimental Approach To examine the effects of pile driving on fishes, a High Intensity Controlled Impedance Fluid-filled wave Tube (HICI-FT) was developed that enabled replication of aquatic far-field, plane-wave acoustic conditions in the laboratory. The HICI-FT is con- structed of a thick stainless steel tube that has a mov- ing coil shaker at either end of the tube for sound stimulation. The HICI-FT system enabled presenta- tion of pile driving sounds in the laboratory and pro- vided control of the parameters that affect pile driving signals. Juvenile Chinook salmon (Oncorhynchus tshawytscha), a federally protected species that is of great concern around pile driving activities on the U.S. west coast, were exposed to pile driving signals that had been recorded in the field during actual pile driving installations. One objective of the project was to prove a cor- relation between the SELcum with the response level of barotrauma injury. Another objective was to test the validity of the “equal energy” hypothesis that has been implicitly accepted for management of activi- ties that generate impulsive sounds. The equal energy hypothesis states that the relevant metric for risk of injury to fish is the SELcum while other metrics are not relevant [e.g., single-strike SEL (SELss) and/or the number of strikes]. In other words, the equal energy hypothesis predicts that no matter how a damaging SELcum is reached (e.g., a few strikes or many strikes), the effects on fishes would be the same. To test this hypothesis, experiments paired sound expo- sures such that there were two treatments with the same SELcum, while the SELss and number of strikes changed. Findings Examination of barotrauma injuries showed that not all injuries had the same physiological signifi- cance for the fish following exposure. Table S-1 dis- plays a rank, weight, and categorization for each injury based on physiological effect. These data were used in the computation of a response weighted index (RWI). Injuries were categorized as Mild, Moderate, or Mortal. The distributions of results from experimental treatments of 1,920 and 960 pile driving strikes showed a statistically significant correlation between RWI and SELcum. Additional statistical analysis showed that as SELcum increased, there was an increase in RWI values. The increase in RWI was the result of the number of injuries each exposed fish experienced as well as the physiological significance of those injuries. Results also showed that fish exposed to 960 strikes had a significantly higher RWI value (p = 0.0145) than fish exposed to 1,920 strikes at the same value of SELcum. In other words, for the same values of SELcum, higher levels of SELss resulted in increases in the number and severity of injuries observed. These injury trends, when quantified using our assessment model, resulted in significantly higher RWI values for 960 vs. 1,920 strikes. This result is understandable if the energy in a strike and the accumulated number of strikes are viewed as factors in producing the RWI. Conclusions The findings of this study demonstrate that the equal energy hypothesis does not apply to effects of pile driving, thereby showing that a single metric of total energy, SELcum, is not sufficient to determine criteria. Other metrics are necessary and should be taken into consideration. Those metrics include, but are not necessarily limited to, SELcum, SELss, and total number of strikes. Interpreting the contour plot (Figure S-1) for application, an RWI of 1 would be a single Mild injury, and an RWI of 2 would be any two Mild injuries (see Table S-1 for injury descriptions). An RWI of 1 or 2 can only be achieved by 1 or 2 Mild injuries. Because it is clear that there are no life- threatening effects from these Mild injuries, an RWI of 2 is an acceptable level of effect and one that is sub-onset of injury. In contrast, an RWI of 3 could be any three Mild injuries or a single Moderate injury. The RWI contours in Figure S-1 along with Table S-1 would first be used to determine an acceptable level of injury; i.e., an RWI of 2. Second, the SELcum con- tours and x axis would then be used to determine which SELss and SELcum, in combination with number of strikes, define the acceptable limits for exposure.

3Table S-1 Observed barotrauma injuries by mathematical weight, category, injury, physiological rank, and brief biological significance statement. Trauma Physiol. Wt Category Injury Description Rank Biological Significance of Injury 5 5 5 5 5 3 3 3 3 3 3 3 3 3 1 1 1 1 1 1 1 1 Mortal Mortal Mortal Mortal Mortal Moderate Moderate Moderate Moderate Moderate Moderate Moderate Moderate Moderate Mild Mild Mild Mild Mild Mild Mild Mild Dead within 1 hr Pericardial (heart) hemorrhage Hepatic (liver) hemorrhage Renal (kidney) hemorrhage Ruptured swim bladder Intestinal hemorrhage Burst capillaries along body wall Pericardial (heart) hematoma Intestinal hematoma Renal (kidney) hematoma Body muscles hematoma Swim bladder hematoma Fat hematoma Ovaries/testes hematoma Blood spots on vent Dorsal fin hematoma Caudal fin hematoma Pelvic fin hematoma Pectoral fin hematoma Anal fin hematoma Fully deflated swim bladder (no ruptures) Partially deflated swim bladder (no ruptures) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 Dead Discrete organ, main body blood pump, bleeding from heart; decreased blood pressure Discrete organ; bleeding from liver; decreased blood pressure Non-discrete spongy organ, held in place with membrane, bleeding; decreased blood pressure Lost ability to maintain buoyancy, sank to bottom; may affect hearing Blood filling the abdominal cavity; decreasing blood pressure Decreased ability to get blood to muscle; decreased blood pressure Could decrease efficacy of heart Major portal system, decreased amount of blood flow to the rest of body. Large amount of blood pooling in more severe cases Could affect swimming ability Could affect ability to regulate buoyancy; could potentially affect hearing Related to swim bladder, caused from swim bladder Potential short-term damage but potential long- term consequences for reproductive success Dilated capillaries near skin, respiratory acidosis, stress with a predisposition, or severe damage Dilated capillaries near skin, respiratory acidosis, stress with a predisposition, or severe damage Dilated capillaries near skin, respiratory acidosis, or stress with a predisposition, or severe damage Fin is near intestinal portal system Fin is near the heart portal system Dilated capillaries near skin, caused by respi- ratory acidosis, stress with a predisposition, or severe damage Negatively buoyant, which could be beneficial for less barotrauma, quick recovery by surface air gulp Negatively buoyant, which could be beneficial for less barotrauma, quick recovery by surface air gulp

4Potential Criterion Based on the study results and findings, an option is to use an RWI biological response criterion of 2, which establishes a new biological response criterion for barotrauma. The corresponding acceptable expo- sure bounds include impulsive sounds that are less than or equal to 179 dB re 1 μPa2s SELss for 1,920 strikes and less than or equal to 181 dB re 1 μPa2s SELss for 960 strikes, combined with a SELcum of no more than 211 dB. These impulsive sound exposure criteria would allow up to two Mild injuries and would not allow for a single Moderate injury. An RWI biological response criterion of 2 is conservative, while it also raises the current interim exposure criteria to a higher level for impulsive sounds, which take into account more acoustic met- rics. These options are for juvenile Chinook salmon of an average standard length of 103 mm and mean weight of 11.8 grams. Application of these data to other species could be done with care, but specific options for other species are beyond the bounds of this study. CHAPTER 1 BACKGROUND Pile driving is becoming increasingly important in the United States and throughout the world for construction projects both near shore and offshore, and includes construction of bridges, docks, liquid natural gas piers, and the like. Although such con- struction typically has been limited to relatively shallow waters near shore and in rivers and streams, more recent pile driving efforts now include deeper water offshore wind farm construction. Because of the increase in pile driving, there is growing concern that the sounds produced by pile driving activity have the potential to harm or kill fishes and/or result in behavioral changes that could affect the survival of populations or even of species. Despite this increased concern, a recent critical review of the literature detailing the known effects of pile driving on fishes has revealed a significant dearth of information (Popper and Hastings, 2009). Most of the work on effects of pile driving has sig- nificant experimental problems that may include experimental design, inadequate use of controls, Dotted Lines = SELcum = SELss + 10Log10(Num.Strikes) Black Lines = RWI = exp(-30.050 + 0.149*SELcum - 0.000171*Num.Strikes)-1 SELss N um be r o f S tri ke s 170 200 202 204 206 208 210 212 214 216 1 2 3 4 188186184182180178176174172 2200 1800 1400 1000 600 5 6 7 8 9 10 Figure S-1 Treatment RWI and SELcum by SELss and number of strikes for all treatments. A contour plot of RWI (the darker curved lines labeled 1-10) illustrates value increases as SELss increase. The dashed lines represent the SELcum curves. The upper horizontal line indicates the 1,920 strike-line, and the bottom horizontal line indicates the 960 strike-line. The darker curved RWI linear contours are the result of testing at only 1,920 and 960 strikes. It is not known whether the functional relationship shown would persist if additional levels of strike numbers were tested. (Note: the two unlabeled curved lines in upper right corner of plots are 218 and 220 dB.)

and/or inappropriate data interpretation (Popper and Hastings, 2009). The lack of data is related to the considerable difficulty in doing experiments on effects of pile driving under conditions in which the investigators could not control the stimulus. In most pile driving studies to date, fishes were exposed to actual pile driving operations. However, the frequency, mag- nitude, and other aspects of the pile driving were controlled by the construction engineers and not the investigators. Consequently, investigators did not have control of any factors needed to understand and quantify the effects of pile driving on fishes (Popper and Hastings, 2009). The ideal pile driving experiments would enable the investigators to fully control the pile driving operation and define parameters such as number of strikes, intervals between strikes, and sound inten- sity. However, this is generally not feasible in the field. At the same time, it is imperative that quantifi- able data be obtained on the effects of pile driving on fishes so that scientists, industry representatives, and regulators can make science-based assessments of potential harm to fishes from a specific pile driv- ing operation. One suggested approach to circumvent the issue of control of signal parameters has been to replicate pile driving sounds in the laboratory. However, this has not been possible until now because the sounds need to be far more intense than those producible by even the best of underwater projectors. Further, even if such sounds could be produced in the laboratory, they might be sufficiently loud as to prevent humans from being anywhere near the experiment for fear of personal injury. Most important, any pile driving sounds used in the laboratory must be accurate rep- resentations of actual pile driving strikes and not just sounds that are very loud. Potential Effects of Pile Driving Sounds on Fish Pile driving impulsive sound may produce sev- eral types of effects on fishes. The one addressed here is referred to as barotrauma, or damage result- ing from rapid change in pressure that directly affects the body gases and thus affects body tissues. More specifically, two changes of gases in the body of fish can lead to injury. The first is when free gas in the swim bladder, or in bubbles in the blood and tissues of fishes expands and contracts during rapid pressure changes, leading to tissue damage. The sec- ond is when the solubility of gas in the blood and other fluids changes with pressure, thereby increas- ing when pressure increases and decreasing when pressure decreases. The swim bladder in the abdom- inal cavity of most fish species is critical for buoy- ancy control (as well as hearing and sound production in some species). Changes in external pressure may cause rapid and substantial changes in the volume of the swim bladder, causing its walls to move exces- sively and/or rupture. A ruptured swim bladder com- promises the fishes’ swimming performance, thereby increasing the risk for further injury or predation be- cause it cannot maintain buoyancy. A swim bladder that changes in size rapidly (whether it bursts or not) can result in damage to nearby tissues. In addition to the presence of a swim bladder in most species, fishes have gasses dissolved in their blood and body tissues. At decompression, the amount of gas that can remain in solution decreases. When gas leaves solution, it forms bubbles in the blood and body tissues. The presence of these bubbles increases the pressure in the vessels and, in the case of veins in particular, can cause their rupture. Gas bubbles in a fish’s circulatory system can disrupt the function or damage vital organs such as the heart, gills, kidney, gonads, and brain. The most severe effects, such as bubbles in the gills or heart, may result in immedi- ate death at exposure from pile driving sounds. Even if an injury is not immediately Mortal, there may be delayed mortality resulting from injury processes such as hemorrhaging, or there may be indirect mor- tality resulting from predation if fish performance is decreased. Study Rationale Origin of Current Interim Criteria To date, the only regulation of the sound levels from pile driving activities are interim physiological injury onset criteria being used for pile driving pro- jects on the U.S. west coast. These levels were estab- lished in 2008 by a group of state agencies on the west coast working in collaboration with the National Marine Fisheries Service (NMFS) and the U.S. Fish and Wildlife Service offices in that region (Wood- bury and Stadler, 2008; Stadler and Woodbury, 2009). The interim criteria were peak sound pressure (SPLpeak) of 206 dB re 1 μPa and a cumulative sound exposure level (SELcum) of 187 dB re 1 μPa2s for 5

6Recent Studies Relevant to Interim Criteria Subsequently to setting the U.S. west coast interim criteria, Ruggerone et al. (2008) investi- gated the effects of pile driving exposure on caged yearling Coho salmon (Oncorhynchus kisutch) mea- suring approximately 90–121 mm in fork length (FL, weight not given). Fish were placed in cages near the piles being driven and exposed to sound from 1,627 strikes over a 4.3-hour period. Peak sound pressure levels were as high as 208 dB re 1 μPa, and SELss reached 179 dB re 1 μPa2s, leading to a SELcum of approximately 207 dB re 1 μPa2s. The study used controls, and there were no reported effects on body tissues. However, this study did not permit test fish the opportunity to fill their swim bladders prior to exposure; therefore, the study results are not applicable to coho salmon in the wild. Absence of acclimation to neutral buoyancy prior to exposure to changes in pressure effectively removes changes in the volume of the swim bladder as a source of barotrauma either to the swim bladder or other tis- sues and organs that may be affected by changes in swim bladder volume (i.e., fish would be protected from barotrauma damage). In a study at Mad River, California, juvenile steelhead salmon (Oncorhynchus mykiss), measur- ing 55–117 mm FL and weighing 1.49–17.43 g, were exposed to pile driving signals at different dis- tances (ranging from about 35 m to 150 m) from the source (Caltrans, 2010a, 2010b). The juvenile salmon were exposed to peak SPLs ranging from 169 to 188 dB re 1 μPa and SELcum ranging from 179 to 194 dB re 1 μPa2s. The Mad River study was well designed and had appropriate controls, properly performed pathology, and appropriate recordings of received sound levels. On-site necropsies and histopathology results showed no mortality and no tissue damage that could be related to pile driving to fish exposed to SELcum as high as 194 dB 1 μPa2s, and no statistically signifi- cant differences between experimental and control animals were detected (Caltrans, 2010b). Higher sound levels were not used, but considering that there were no differences in tissue effects between exposed and control, it is reasonable to suggest that injury onset is at sound levels above 194 dB re 1 μPa2s SELcum, and likely well above this level. In yet another study, Houghton et al. (2010) exposed 133 caged juvenile coho salmon (Onco- rhynchus kisutch) measuring approximately fishes above 2 g and a SELcum of 183 dB re 1 μPa2s for fishes below 2 g. However, the agreement specif- ically designated the criteria as interim, and the agencies were committed to “review the science periodically and revise the threshold and cumulative levels as needed to reflect current information” (Stadler and Woodbury, 2009). The NMFS also recognized that there is a “reset- ting” of SELcum after 12 hours of non-exposure (Stadler and Woodbury, 2009). Thus, the SELcum for a fish during a pile driving operation is reset to 0 for the next set of exposures if there is a 12-hour period between the end of the first pile driving exposure and the start of the next. This “resetting” was specific for recovery from temporary effects to the hearing of exposed fish, not barotrauma. In preparation for the 2008 decision, the Califor- nia Department of Transportation (Caltrans) asked a group of internationally known investigators to review the literature on effects of sound on fishes and to make recommendations on possible criteria. This resulted in two memos (Popper et al., 2006; Carlson et al., 2007) that examined the best available science and then proposed interim criteria based on those data. In the 2006 memo, Popper et al., (2006) developed a strong case for using a single-strike sound exposure level (SELss) of 187 dB re 1 μPa2s and a SPLpeak of 208 dB re 1 μPa. This was the first attempt to use dual-criteria to protect fishes from physiological injury resulting from exposure to pile driving. The dual-criteria was adopted by the authors, and later by NMFS, with the idea that the SELss value confines the total acoustic energy fishes may experi- ence by exposure to a single impulsive sound, while the peak sound pressure level protects fishes from an especially strong excursion in pressure within the sound impulse. Carlson et al. (2007) used additional data to that available to Popper et al. (2006) to propose SELcum values for onset of tissue damage that depended on fish mass. Carlson et al. (2007) suggested that for fishes above 2 g (small larvae), the SELcum value for non-auditory tissue damage should be 190 dB re 1 μPa2s, and for fishes below 1 g, they suggested an SELcum of 183 dB re 1 μPa2s. Carlson and his col- leagues made the important point that as fishes get larger the exposure value must be increased further. Most pertinently, Carlson et al. (2007) recom- mended a conservative value of 197 dB SELcum for fishes above 8 g, and a value above 213 dB SELcum for fishes over 200 g.

790–121 mm FL (weight not given) to sheet pile driving in the Port of Anchorage. In this study, fish were exposed to as many as 2,781 pile driving strikes at distances ranging from 1 to 50 m from the source. Acoustic monitoring during the tests measured peak SPLs as high as 195 dB re 1 μPa and SELss values as high as 166 dB 1 μPa2s, with SELcum as high as 191 dB re 1 μPa2s. No mortalities were observed and no tissue damage was reported as late as 48 hrs post-exposure. As in the case of the Ruggerone et al. (2008) study, the test fish in this study were not given the opportunity to fill their swim bladders prior to exposure. The results of this study are only applica- ble to unacclimated (negatively buoyant) salmonids for this reason. Study Goal and Objectives The goals of this study were to assess the effects of exposure to high-intensity pile driving sounds on Chinook salmon physiology and to develop an understanding of the sound exposure(s) that result in the onset of physiological impact on fish. An additional goal was to provide quantitative data that can be used to define criterion levels for tissue damage onset for use in design of pile driv- ing projects and in identifying options for protec- tion of animals. Initially, an additional goal was to measure the effects of exposure to sounds on hearing capabili- ties. However, due to technical difficulties, it was agreed by the investigators, the NCHRP advisors, and outside experts, that hearing measures would not be continued or included in this report (see Appendix H for details of those studies). Previous Studies Using Similar Equipment The original approach of designing a rigid tube for acoustical studies was an attempt to measure hearing thresholds of a fish in the laboratory with an ideal plane-wave sound field (Hawkins and MacLennan, 1976). The Hawkins and MacLennan (1976) study used a steel tube 80 cm long with a 13-cm wall thickness and 27-cm inner diameter fit- ted with a sound projector at each end. The hearing thresholds of the plaice, Pleuronectes platessa, were measured in terms of sound pressure and particle motion. A similar chamber was designed several decades later (Rogers and Lewis, 1999), again with a rigid steel tube but with mechanical shaker-driven pistons at either end. This chamber was used to study the effects of sound exposure on the lungs of mice (Dalecki, 2002) as well as the effects of sonar expo- sure on the vestibular system of guinea pigs. In both cases, it was demonstrated that the traveling wave sound field to which the animals were exposed was equivalent to a far-field scenario and that the response of organs and tissues were comparable to studies done in far-field conditions. CHAPTER 2 METHODS The sound exposure paradigms used in this study were designed, in part, to test the validity of the “equal energy” hypothesis that has been widely cited for management of activities that generate impulsive sounds. In addition, the paradigm was designed to obtain data necessary to derive a stress- response function. The equal energy hypothesis states that the risk of injury to fishes is a function of the SELcum, and there is no need to consider other sound metrics, such as SELss and/or the number of strikes. In other words, the “equal energy” hypothe- sis predicts that no matter how a SELcum value is reached (e.g., a few strikes with higher energy per strike, many strikes with lower energy per strike), the effects on fishes would be the same. The methodology used in the experiments to test this hypothesis is documented in this section in brief, with details provided in referenced Appen- dices. Each aspect of the study is addressed: • Fish source and fish maintenance; • The sound exposure device, including design and operation; and • Barotrauma assessment and characterization. In each case, details are provided for the meth- ods and approaches used to conduct the study. Study Fish This study used juvenile Chinook salmon with an average standard length of 103 mm ± 8.75 mm (standard deviation) and an average weight of 11.8 g ± 3.47 g (Figure 1). The Chinook salmon were pro- vided by Pacific Northwest National Laboratory from the Priest Rapids Hatchery in Mattawa, Wash- ington. Additional details on study fish are found in Appendix F.

8Fish Maintenance Fish were kept in a dedicated aquarium room of the laboratory in the Biology/Psychology build- ing at the University of Maryland. This room met all federal standards for animal care. The care and maintenance of the room, as well as the conduct of all experiments described in this report, were done under protocols reviewed and approved by the Insti- tutional Animal Care and Use Committee (IACUC) of the University of Maryland (see Appendix D). Fish were held under authority of the Maryland Department of Natural Resources (Natural Resources Articles 4-602 and 4-11A-02). Details of fish mainte- nance are provided in Appendix F which can be found online at http://apps.trb.org/cmsfeed/TRBNetProject Display.asp?Project ID=763. Sound Exposure Apparatus and Methods Sound exposure was conducted in a system called the High Intensity Controlled Impedance Fluid-filled wave Tube (HICI-FT). The HICI-FT is a specially designed wave tube that used large shak- ers to produce sounds that accurately reproduce actual pile driving sounds. It enabled presentation of actual pile driving sounds in the laboratory and allowed for control of the number, duration, and other aspects of the pile driving sounds. Thus, it was possible to present stimuli at different cumulative sound levels, single-strike levels, and total number of strikes, using eight different pile driving signals. Essentially, the HICI-FT enabled the investigators to provide the first quantified data on effects of pile driving signals on fish physiology. Details of the design, operation, and control of the HICI-FT are provided in Appendix G. Sound presentation was controlled using Lab- VIEW (National Instruments Corporation, Austin, Texas). In the HICI-FT, the presented stimuli were captured during experiments with a hydrophone (Brüel & Kjær Sound & Vibration Measurement A/S, Naerum, Denmark, Model 8103), and digitized by LabVIEW. In addition, Dazzle MovieStar software (http://dazzle-moviestar.software.informer.com/) cap- tured digital images of a 45° region inside the HICI- FT during experiments. Digital images allowed the observer an occasional glimpse of a fish if it swam into the angle of view. (Note these observations were pri- marily to check on fish survival and cannot be used for behavioral studies.) The HICI-FT chamber (Figures 2 and 3) was a circular tube 0.45 m long with a 0.25-m internal diameter and 3.81-cm-thick stainless steel walls filled with water. At either end of the tube was a rigid lightweight circular piston held by a membra- nous seal in the center of the steel end cap. Each pis- ton was connected to a linear electrodynamic motor (moving coil shaker) anchored to the end caps. The motors of the shakers were driven separately with signals appropriate to create the desired pressure and velocity fields within the tube for the sound expo- sure of the fish. Sounds The HICI-FT chamber was designed to produce propagating plane waves with a peak sound pressure level (SPL) of at least 215 dB re 1 μPa. The HICI- FT was able to generate pressure and particle motion levels that were very similar to those produced by pile driving activity. The pile driving signals used in this study were analogues of field recordings of both pressure and particle motion taken at a range of 10 m from a steel shell pile driven using a diesel hammer at the Eagle Harbor Maintenance Facility (MacGillivray and Racca, 2005). The actual sound exposure paradigms used in the experiments described here were designed to mimic actual pile driving activities. Thus, the experimental characteristics of each sound exposure matched real-life activity, such as the time and frequency domain characteristics of each pile strike, inter-strike-interval, and number of strikes. The signals used in the experiments consisted of eight different pile driving strikes, which were nor- malized to the same SEL and compiled into a single file that contained 12 repetitions of each of the eight signals, for a total of 96 strikes. MATLAB (Math- Works, Inc., Natick, Massachusetts) was used each day to generate a randomization of the 96-strike file. This file then was used by LabVIEW for the day and repeated 10 times for a 960-strike presentation or 20 times for a 1,920-strike presentation. Therefore, Figure 1 Juvenile Chinook salmon used in this study. Note: caudal fin clipped for identification purposes.

9fish received a pseudorandom presentation of pile strikes, and every day was different. HICI-FT Sound Control Operation Signal generation and data acquisition for the HICI-FT was controlled using a Dell laptop com- puter and a 12-bit analog/digital (A/D and D/A) converter (National Instruments Corporation, Model PCI-MI0 16E1). Analog drive signals were gener- ated on two separate channels of the A/D converter. These were filtered and attenuated using anti-aliasing filters and programmable attenuators (Tucker-Davis Technologies [TDT], Alachua, Florida, Model PA4). The attenuators provided the system with 100 dB of dynamic range beyond the 12-bit resolution of the A/D converter. The attenuators were controlled through a serial interface with the PC. The outputs of the attenuators were amplified by an amplifier (Crown International, Elkhart, Indiana, Model XTi 4000), one to each shaker (Vibration Test Systems, VG-150 Vibration Generator, Model VTS 150). All information regarding each experiment was recorded on data sheets. An example of the data sheet is in Appendix C. Figure 3 View inside the HICI-FT chamber (vertical position). A Chinook salmon can be seen toward the bottom of the tube. The long silver cylinder towards the bottom of the tube on the left is the light source and digital camera. Just above this on the left is the hydro- phone (black, smaller tube). The wire in the center is for an accelerometer (white cylinder attached to the bottom of the tube). The large black device (lower right) measured temperature and total dissolved gas of the water in the tube after each experiment. The bottom of the tube is the faceplate for the piston coupled to the lower shaker. Figure 2 The HICI-FT as described in the text and in Appendix G. The section labeled top tank is an acrylic water-filled chamber in which the fish were placed prior to exposure. The HICI-FT is shown in the horizontal position used during sound exposure.

General Experimental Procedures Details of experimental procedures are pre- sented in Appendix G. In brief, two Chinook salmon were transferred into the HICI-FT’s acrylic chamber and allowed to swim freely in the chamber for a 20- min acclimation period. After the 20 min, the fish were scored to be negatively, neutrally, or positively buoyant. Fish were allowed to enter the tube, and then they were closed in by lowering and locking the top shaker onto the tube. Control fish were put through the same process as treatment fish but with- out the pile driving sound. When the exposure was completed, fish were transferred to the barotrauma injury assessment. The fish were exposed to one of eleven pile driv- ing treatments (see Chapter 3 Results). These treat- ments varied in SELcum, SELss, and number of strikes, which affected the duration of the exposure. For example, for fish exposed to 1,920 strikes, the expo- sure duration was 48 minutes, while for 960 strikes it was 24 minutes. Barotrauma Analysis and Characterization Following exposure in the HICI-FT, fish were examined to determine if physical injuries were associated with sound exposure. Prior to examina- tion, fish were euthanized in a buffered MS-222 solution of 100 mg/L. The fish were examined for external and internal signs of barotrauma. All poten- tial injuries, and a fuller treatment of this analysis, are provided in Appendix A. The design for the assessment of barotrauma fol- lowed the procedures developed by co-investigator Carlson’s group at the Pacific Northwest National Laboratory (PNNL) for their study on rapid decom- pression in salmonids (Stephenson et al., 2010). That study created a framework that assessed baro- trauma and the effects of fish physiological condi- tions on barotrauma injuries, and then statistically analyzed and modeled those barotrauma injury observations. Response Variable Derivation The barotrauma data set was based on binary vari- ables (0 or 1) that denoted the presence or absence of observed external and internal barotrauma injuries. After thorough review of the entire injury regime, many injury indices from the original barotrauma 10 injury list (Appendix A) were eliminated due to lack of occurrence (i.e., embolisms were removed). A few other injuries were combined into one score because they indicated the same injury (i.e., external sign of pericardial hemorrhage was combined with internal scoring of pericardial hemorrhage). Examination of the injury panel showed that not all injuries had the same physiological significance for the health of the fish following exposure. The physiological cost or effects of trauma and barotrauma are poorly understood in fish, thus a novel model was developed to qualitatively assess barotrauma and was applied to this study and to a concurrent study considering the effects of explo- sive sound from underwater rock blasting (Carlson et al., 2011). The physiological significance of each injury was determined using available literature whether fisheries or mammalian-based (Husum and Strada, 2002; Oyetunji et al., 2010) and pro- posed energetic costs based on an understanding of each type of injury (Woodley and Halvorsen, per- sonal observations; Gaspin et al., 1975; Iwama et al., 1997). Physiological significance of each observed injury was assessed and given a physiological rank and a weight (Table 1). Observed injuries were assigned to trauma categories based on the physio- logical significance for each observed injury, indi- vidually. The injuries were then separated into three trauma categories: Mortal, Moderate, and Mild. The Mortal trauma category included observed injuries that were severe enough to lead to death. The Moderate trauma category included observed injuries likely to adversely impact fish health, but which, when considered individually, were likely recoverable under ideal conditions (i.e., no addi- tional stressors) without being Mortal (Casper et al., in prep.). Finally, Mild trauma category refers to observed injuries that had minimal to no physiolog- ical cost to fish, which quickly recovered under ideal conditions (Casper et al., in prep.). A mathematical weighting was applied to the trauma categories to underscore the contribution of the observed injury to the response weighted index (RWI, see below). Mortal injuries (injuries catego- rized under Mortal trauma) were assigned a weight of 5, Moderate injuries (injuries categorized under Moderate trauma) weighted as 3, and Mild injuries (injuries categorized under Mild trauma) weighted as 1 (Krischer, 1979; Chawda et al., 2004). The weight assignments to the trauma categories were

11 Table 1 Observed barotrauma injuries by mathematical weight, category, injury, physiological rank, and brief biological significance statement. Trauma Physiol. Wt Category Injury Description Rank Biological Significance of Injury 5 5 5 5 5 3 3 3 3 3 3 3 3 3 1 1 1 1 1 1 1 1 Mortal Mortal Mortal Mortal Mortal Moderate Moderate Moderate Moderate Moderate Moderate Moderate Moderate Moderate Mild Mild Mild Mild Mild Mild Mild Mild Dead within 1 hr Pericardial (heart) hemorrhage Hepatic (liver) hemorrhage Renal (kidney) hemorrhage Ruptured swim bladder Intestinal hemorrhage Burst capillaries along body wall Pericardial (heart) hematoma Intestinal hematoma Renal (kidney) hematoma Body muscles hematoma Swim bladder hematoma Fat hematoma Ovaries/testes hematoma Blood spots on vent Dorsal fin hematoma Caudal fin hematoma Pelvic fin hematoma Pectoral fin hematoma Anal fin hematoma Fully deflated swim bladder (no ruptures) Partially deflated swim bladder (no ruptures) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 Dead Discrete organ, main body blood pump, bleed- ing from heart; decreased blood pressure Discrete organ; bleeding from liver; decreased blood pressure Non-discrete spongy organ, held in place with membrane, bleeding; decreased blood pressure Lost ability to maintain buoyancy, sank to bottom; may affect hearing Blood filling the abdominal cavity; decreasing blood pressure Decreased ability to get blood to muscle; decreased blood pressure Could decrease efficacy of heart Major portal system, decreased amount of blood flow to the rest of body. Large amount of blood pooling in more severe cases Could affect swimming ability Could affect ability to regulate buoyancy; could potentially affect hearing Related to swim bladder, caused from swim bladder Potential short-term damage but potential long- term consequences for reproductive success Dilated capillaries near skin, respiratory acidosis, stress with a predisposition, or severe damage Dilated capillaries near skin, respiratory acidosis, stress with a predisposition, or severe damage Dilated capillaries near skin, respiratory acidosis, or stress with a predisposition, or severe damage Fin is near intestinal portal system Fin is near the heart portal system Dilated capillaries near skin, caused by respira- tory acidosis, stress with a predisposition, or severe damage Negatively buoyant, which could be beneficial for less barotrauma, quick recovery by surface air gulp Negatively buoyant, which could be beneficial for less barotrauma, quick recovery by surface air gulp

based on the assessment of physiological signifi- cance, which considered the significance of multi- ple injuries and inspection of data for the occurrence of injury combinations. For example, the occurrence of two injuries categorized as Moderate was assessed to have physiological costs similar to that of one injury categorized as Mortal. An RWI was calculated for each treatment fish and each control fish. The formulas used were: Where RWI = response weighted index, i = injury type index, m = 22, number of injury types (Table 1), Ti = the proportion of the sample of fish exposed to a treatment that experienced injury type i, Wi = the Trauma Category weight (5, 3, or 1) for injury type i, Ci = the proportion of the sample of control fish for a treatment that experienced injury type i. Statistical Analysis The response variable RWI was transformed, as shown in Equation 1.3, before analysis in order to stabilize variance and linearize the response curve. In addition, cumulative energy was expressed as: Analyses of covariance (ANCOVA) were per- formed regressing yi against SELcum and assessing whether number of strikes (960 or 1,920) had an additional effect on fish response beyond that de- scribed by SELcum. Initial analyses were conducted on Treatments 2 through 11 to balance the design. Treatments 2 through 11 were paired while Treat- ment 1 lacked a counterpart. Once a model was SEL SEL log number of strikes Equ cum ss 10= + ( )10 ation 1 4. yi = +( )ln .RWI Equationi 1 1 3 RWI Treatment W T Equationi i i m( ) = ×( )∑ 1 2. RWI Control W C Equationi i i m( ) = ×( )∑ 1 1. selected using the balanced design, Treatment 1 was added to the analysis to refine the results. A com- plete statistical analysis is shown in Appendix E. CHAPTER 3 RESULTS Sound Exposures A primary objective of this project was to test the hypothesis that the magnitude of barotrauma effects increased with increasing exposure to pile driving sounds as measured in terms of cumulative sound exposure level (SELcum). Each experimental animal was exposed to one of the pile driving treat- ments shown in Table 2. The maximum exposure had an average SELss of 187 dB re 1 μPa2s and a SELcum of 219 dB re 1 μPa2s. This level of SELss had an average peak SPL of 213 dB re 1 μPa, which was the maximum amount of energy that could be safely generated by the HICI-FT. Therefore, for Treatment 1’s “pair,” a SELcum of 219 dB re 1 μPa2s could not be generated with 960 strikes since this would have required a peak SPL above 215 dB re 1 μPa. The remaining treatments were conducted in pairs since the same SELcum could be reached with both 1,920 and 960 strikes by having a higher SELss (and thus higher peak SPL) for 960 strikes than for 1,920. In each treatment pair, the goal was to present the same SELcum value but vary the number of pile strikes, which altered the SELss value. Holding the SELcum steady was done to imple- ment treatments that could be used to explore the equal energy hypothesis while providing insight into the relative importance of SELss and SELcum in deter- mining effects of sound on fish. Each treatment pair was aimed at a specific SEL value. However, many variables affect the ability to present the precise signal level, and thus the range of SELcum treatments was continuous, rather than dis- crete points. Each treatment blends with the next SELcum, such that each treatment is ± 1.5 dB of its specified value, i.e., 216 ± 1.5. Because the data are continuous, an RWI was calculated for each individ- ual fish. However, in Table 2, the average RWI is reported for each treatment. Barotrauma Inspection of the log-transformed RWI values show that fish with 960 strikes had a statistically significant higher RWI value than fish exposed to 12

13 1,920 strikes at the same value of SELcum. This can be seen when comparing the ln (RWI+1) regression lines for each 960- and 1,920-strike treatment. In other words, for common values of SELcum, higher values of SELss resulted in significantly higher val- ues of RWI for 960 strikes (Figure 4). The ANCOVA demonstrated that ln (RWI+1) was linearly related to SELcum, and fish exposed to a common value of SELcum using 960 or 1,920 strikes had statistically different RWI values. Using a balanced design for Treatments 2 through 11 (i.e., these had pairs of common values; Treatment 1 did not), common slopes were found for the regression of ln (RWI+1) versus SELcum, but different inter- cepts for fish exposed to either 960 or 1,920 strikes (Table 3). Adding treatment 1 (i.e., SELss = 187, SELcum = 219, number of strikes = 1,920) to the analysis did not change the linearity of the data on the natural log scale or the regression relationships (Table 4, Figure 4). The final model was based on the use of all Treatments 1-11. The RWI values were calculated and plotted for each fish as shown in Figure 5. Distributions of 1,920 and 960 strikes in Figure 5 show an increase of RWI values correlated with an increase of exposure severity (SELcum). The increase in RWI was the result of both the number of injuries Table 2 Study exposure treatments and exposure details. Treatment Avg. Number of Avg. Avg. Peak Duration, Exposed Control Avg. No. SELcum Strikes SELss SPL min Fish, n Fish, n RWI 1 219 1920 187 213 48 44 33 15.318 2 216 1920 183 210 48 36 16 5.971 3 216 960 186 213 24 28 10 6.071 4 213 1920 180 207 48 26 5 2.346 5 213 960 183 210 24 31 7 4.323 6 207 960 177 203 24 24 8 1.042 7 207 1920 174 201 48 43 17 0.581 8 210 960 181 208 24 31 10 4.032 9 210 1920 177 204 48 30 11 3.433 10 203 960 174 201 24 32 11 0.656 11 204 1920 171 199 48 31 12 0.419 All Treatments: Model Ln(RWI+1) ~ SELcum + #strikes SELcum Ln (R W I+1 ) 200 960 Strikes 1920 Strikes 960 Strikes 1920 Strikes 225220215210205 4 3 2 1 0 –1 –2 Figure 4 Scatterplots of SELcum vs. ln (RWI+1) for all treatments. Solid line shows predicted ln (RWI+1) values for 960 strikes and dashed line for 1,920 strikes. Red squares denote the 960 strikes and blue diamonds denote the 1,920 strikes.

14 Table 3 Sequential analysis of covariance (ANCOVA) results for Treatments 2–11. Response: ln (RWI+1). Best model contains the covariate SELcum followed by number of strikes. Treatments 2–11 Source df SS MS F-value P-value SELcum 1 103.22 103.22 281.240 <0.0001 Residuals 309 113.41 0.37 Source df SS MS F-value P-value SELcum 1 103.22 103.22 289.679 <0.0001 Strikes 1 3.95 3.95 11.077 0.0010 Interaction 1 0.07 0.07 0.196 0.6583 Residuals 307 109.39 0.36 Best Model Source df SS MS F-value P-value SELcum 1 103.22 103.22 290.437 <0.0001 Strikes 1 3.95 3.95 11.106 0.0010 Residuals 308 109.46 0.36 Table 4 Analysis of covariance (ANCOVA) from the best model from Table 6 applied to all Treatments 1–11. Response: ln (RWI+1). All Treatments Source df SS MS F-value P-value SELcum 1 201.30 201.30 542.039 <0.0001 Strikes 1 2.24 2.24 6.033 0.0145 Residuals 352 130.72 0.37 each exposed fish experienced as well as the biologi- cal significance of those injuries. The increase in the number of injuries per test fish with increase in sever- ity of exposure is shown in Figure 6. Additional insight can be obtained by examining Figure 7. As SELss increases, RWI increases as well. However, the indication is that as SELss increases, the severity of barotrauma response is differentially amplified as the number of strikes increases. There is a higher level of RWI values for the 960 strike treatments, which can be seen when comparing the ln (RWI+1) regressions lines for each 1,920 and 960-strike treatment (Figure 4). Figure 8C summarizes study findings. However, because of the complexity of the response of fish to exposure over the SELcum treatments as they are defined by the SELss and number of strike variables (Figure 8), Figures 8A and 8B, are presented sepa- rately to show the construction of 8C. Figure 8A is the background layer of the plot and represents the sample space for the study. The x- and y-axes are SELss and number of strikes, respectively, while the z-axis is SELcum. The blue dashed contours represent SELcum values, which were generated by calculating the relationship between the number of strikes and the SELss (see Equation 1.4). For exam- ple, a SELcum of 208 dB is produced when there are 960 strikes at a SELss of 178 dB, or when there are 1,920 strikes at a SELss of 175 dB. Figure 8B shows the RWI contour lines in black. Axes are the same as for Figure 8A with treatment RWI in the z-axis for this layer. Note that the top

15 0 5 10 15 20 25 30 35 40 203 R W I 206 SELcum RWI 1920 RWI 960 221218215212209 Figure 5 RWI values for each fish by SELcum for 1,920 and 960 pile strikes. 0 2 4 6 8 0 5 10 15 20 25 30 N um be r o f F ish Number of injuries Treatment (ordered by S ELcum) T1T3T2T5T4T8T9T6T7T10T11 10 12 Figure 6 Frequency of barotrauma injury occurrence per fish. The number of test fish (z-axis) with number of unweighted-barotrauma injuries (y-axis) by each treatment (x-axis). For example, in the most severe exposure (Treatment 1 = T1), 1 fish had 13 injuries, and 10 fish had 8 injuries. Similarly, for the least severe exposure (T11), 6 fish had 1 injury, and 24 fish had 0 injuries. 0 5 10 15 20 25 30 35 40 45 170 R W I RWI 1920 RWI 960 173 SELss 188185182179176 Figure 7 Individual RWI values by SELss.

Figure 8 Panel A is the background layer plotting the SELcum contours (dashed lines) by SELcum, by SELss, and number of strikes within the treatment range. Panel B is a contour plot of ln (RWI+1) (the solid lines labeled 1-10), which illustrates values increase as SELss increase. The upper horizontal line indicates the 1,920 strike-line, and the bottom horizontal line indicates the 960 strike-line. Panel C is the composite of B on top of A, and shows where the RWI contours fall over the SELcum and SELss in relation to number of strikes. See text for further discussion. Dotted Lines = SELcum = SELss + 10Log10(Num.Strikes) SELss N um be r o f S tri ke s 200 202 204 206 208 210 212 214 216 170 188186184182180178176174172 2200 1800 1400 1000 600 Black Lines = RWI = exp(-30.050 + 0.149*SELcum - 0.000171*Num.Strikes)-1 SELss N um be r o f S tri ke s 170 1 2 3 4 5 6 7 8 9 10 188186184182180178176174172 2200 1800 1400 1000 600 Dotted Lines = SELcum = SELss + 10Log10(Num.Strikes) Black Lines = RWI = exp(-30.050 + 0.149*SELcum - 0.000171*Num.Strikes)-1 SELss N um be r o f S tri ke s 170 200 202 204 206 208 210 212 214 216 188186184182180178176174172 2200 1800 1400 1000 600 1 2 3 4 5 6 7 8 9 10 A B C

horizontal line represents 1,920 strikes and the lower horizontal black line represent 960 strikes, the curvi- linear contours between represent the RWI values (1-10) derived using the results of testing at 1,920 and 960 strikes. It is unknown whether the derived relationship plotted would persist if additional levels of strike numbers were tested. However, at both strike number levels the RWI values increase with increase in SELss. The graph of RWI as a func- tion of number of strikes and SELss is interpreted the same way as SELcum in Figure 8A. Thus, an RWI of 2 would be achieved when the SELss is about 181 dB for 960 strikes, and an SELss of 179 for 1,920 strikes. Figure 8C brings both contour layers, SELcum and RWI, together onto one graph, thereby showing their relationship to each other as well as their rela- tionship to SELss and number of strikes. While com- plex, it links a common metric used to manage the exposure of fish to impulsive sound generated by pile driving, SELcum through its constituent parts, SELss and number of strikes, along with the derived RWI in this study. Mathematically, RWI and SELcum are both dependent variables defined by the indepen- dent variables SELss and number of strikes. The composite plot (Figure 8C) identifies the criteria for acceptable pile driving sound exposure given a selected response outcome or RWI value. For example, if a RWI of 2 was selected as the maxi- mum acceptable level of biological response for a pile driving project, during project planning the likely SELss and number of strikes needed to drive each pile could be considered to identify alternative pile driving strategies to avoid exposures that would risk exceedance of the selected RWI value. These expectations would be dynamically managed during the project as actual SELss values were observed. Practically speaking, SELss with the addition of mit- igating actions such as bubble curtains are the pri- mary means available to control the exposures fish experience to pile driving sound. The results of this project suggest that a RWI of 2 or less does not lead to physiological effects that reduce either the imme- diate or long-term performance and energetics of Chinook salmon, or probably other species as well. CHAPTER 4 DISCUSSION Overview The work presented here represents the first study to test the effects of pile driving sounds on fish in a controlled plane-wave acoustic field. Thus, the design enabled systematic exploration of the relationship between potential injury to fish and specific sound characteristics such as number of strikes and sound level, expressed as both single strike (SELss) and cumulative sound exposure lev- els (SELcum). In contrast, all earlier studies on pile driving used caged fishes under conditions in which the investigators were unable to control any aspects of sound presentation (e.g., number of strikes, sound intensity), or provide adequate bio- logical control groups (reviewed in Popper and Hastings, 2009). Many aspects of pile driving could have been explored during these experiments. These included sounds from various types of piles (e.g., steel, con- crete), inter-strike intervals, or total number of strikes. After discussion with the NCHRP advisory group for this project, it was decided to limit param- eters and the variables to allow for collecting mean- ingful and statistically valid data and analyses in a reasonable amount of time. Examination of the lit- erature and talking with scientists, regulators, and industry representatives, found the most important variables were SELss and number of strikes. These two variables can be used to control driving piles, either through management of the energy applied to a pile during each strike or by implementation of mitigating actions such as bubble curtains. Since the study focuses on sounds, it is reasonable to conclude that these sound level metrics could be extrapolated to other impulsive sounds from pile driving, as long as the sound spectra are reasonably similar to strik- ing steel piles. The study’s experimental strategy was designed to evaluate the relationship between SELss and SELcum, number of strikes, and barotrauma damage. To accommodate concerns regarding the variability in impulsive sound amplitude distribution, the expo- sure stimulus included eight different impulsive pile strike sounds, pseudorandomly presented, all with the same SELss differing in details of their amplitude (SPLpeak) characteristics. At the same time, a detailed examination of the variables selected allowed for the quantitative examination of relationships between sound exposure level and injury. Techniques devel- oped during and results collected from this study have been incorporated into follow-up studies exam- ining the recovery from Mild and Moderate injuries and responses in Chinook salmon as well as other fish species; these will be published in subsequent papers (Casper et al., in prep.). 17

Rejection of the Equal Energy Hypothesis for Pile Driving The equal energy hypothesis was suggested to be irrelevant for fishes exposed to pile driving (discussed in Carlson et al., 2007). The significant dif- ference (p = 0.015) between the 1,920- and 960-strike regression lines (Figure 4) strongly supports this sug- gestion, and similar results have been shown for mammals exposed to impulsive sounds (Hamernik et al., 2003). As a result, the use of a single metric, typically SELcum, is not sufficient to determine regu- latory criteria but remains an important variable to observe during pile driving activity. Other metrics related to SELcum, such as SELss and number of strikes need to be taken into consideration. Relationship Between Pile Driving Exposure and Biological Response This study has shown that the severity of baro- trauma is a function of the energy in each strike (SELss) summed over the total number of strikes needed to drive a pile, SELcum. For each strike that occurs during pile driving activity, the energy deliv- ered to the pile is used to overcome resistance encountered from substrate changes and other factors. Furthermore, for each strike, the energy delivered to the pile is managed by manipulating the stroke (the distance the hammer travels). For a particular pile, generally the greater the energy in the strike on a pile, the higher the energy level in the impulsive sound generated (Carlson and Weiland, 2007). This means that the elemental unit of exposure for fish to pile driving sound is the energy in individual strikes, SELss. This complex relationship between exposure and response is summarized for juvenile Chinook salmon in Figure 8. Data from this study demonstrate that as energy levels of pile driving exposures increase there is a statistically significant increase in the severity of barotrauma injuries. The highest energy levels pre- sented in this study caused injuries that resulted in substantial physiological costs to the fish. The less severe exposures caused fewer barotrauma injuries, and these tended to be Mild injuries, which imposed minimal or no physiological effects on the fish. The survival of fish exposed to pile driving sound that experience injury is dependent on the cumulative effect of those injuries on the perfor- mance and energetics of the fish. The most severe injuries have a clear impact on performance because of the consequences, likely mortality and/or damage to organs. Severe injuries would require consider- able opportunity for recovery, which under most cir- cumstances would not unlikely be available to the fish (e.g., predator-free refuge, ideal flow rates, eas- ily accessed nutrition-rich foraging). In contrast, the Mild physiological injuries seemingly would not affect the performance of fish because life functions would not be compromised and recovery needs would be minimal. The biological response metric, RWI, derived in this study, permits identification of impulsive sound exposure thresholds, expressed in terms of common sound measures, which protect fish from levels of physiological injury that would likely affect their performance and ability to survive. None of the Mild injuries singularly or in combination would be likely to reduce individual performance or affect ecologi- cal endpoints. This is not necessarily the case for Moderate injuries and certainly not the case for Mortal injuries. At the lower end of RWI values, an RWI level of 1 or 2 can only be realized by 1 or 2 Mild injuries respectively. An RWI of 3 can occur with 3 Mild injuries or 1 Moderate injury. While 3 Mild injuries would not likely reduce performance of fish, the same may not be true of 1 Moderate injury. With this in mind, an option would be that an RWI level of 2 be used as an acceptable level of physiological injury for juvenile Chinook salmon exposed to pile driving sound. This level of injury is not overly cautious, but still very protective of fish exposed to pile driving sound. This recommendation is specifically for juvenile Chinook salmon in the range of 93–115 mm (standard length) and an aver- age wet weight of 11.8 g. An RWI of 2 could be carefully extrapolated to include other fish within the salmonid family of similar size. It may also be possible to extrapolate to other species, but that would be beyond the scope of this report. In addi- tion, based on predictions made by Carlson et al. (2007), it is likely that larger fish would show less effects at the same exposure levels as those used here. Thus, at higher exposure levels, it is possible that larger fish would potentially have an RWI of 2 or below. Figure 8 can be used to identify fish responses (RWI) to impulsive pile driving sound, in terms that map directly to the pile driving activity. This mapping will protect fish from having a biological response that is greater than an RWI of 2, for example. If an 18

RWI of 2 were selected as the maximum acceptable physiological response, then pile strikes that can generate impulsive sound with SELss values up to 180 dB re 1 μPa2s could be used, as long as the num- ber of strikes required to drive the pile was less than 960. If the sound generated by each strike contained less SELss energy, then the number of strikes avail- able to drive the pile could be increased, as long as the combination of SELss and number of strikes were to the left of the RWI 2 contour in Figure 8. The data from this study do not permit the RWI curves to be extrapolated beyond the 960 and 1,920 strike bounds of this study. It should be noted that the current west coast interim criteria use a dual approach for decisions on potential onset of physiological effects, either a specific SELcum or a single-strike peak level of sound pressure (SPLpeak). Under those criteria, if the SELcum is reached over multiple strikes, or any one strike exceeds the SPLpeak, mitigation occurs (Popper et al., 2006; Carlson et al., 2007; Stadler and Woodbury, 2009). Furthermore, Carlson and Weiland, (2007) found that SELss and SPLpeak are highly correlated. In their study of pile driving sounds, they deter- mined that SPLpeak could be estimated given SELss using Equation 4.1. The linear fit of SPLpeak to SELss had an r2 of 0.85. Using equation 4.1, all of the impulsive sound metrics, used in previous criteria, can be defined for operating criteria obtained from a selected RWI value and Figure 8. CHAPTER 5 APPLICATION TO PILE DRIVING PROJECTS Background Exposure criteria for fishes exposed to pile driv- ing sound are almost always given as one or two sound metrics such as SELss, SELcum, SPLpeak, or SPLrms. During pile driving operations, monitoring is conducted at one or two points in the sound field. Over time, a convention has evolved to monitor the sound field at 10 m from the pile being driven at one or two depths in the water column, with one of the depths being near the midpoint of water depth. As long as the regulatory criteria are not exceeded, the pile driving activity is determined to be in compli- ance with operating permits. SPL SEL Equationpeak ss= +18 02 1 05 4 1. . . The specified exposure criteria are typically con- servative or precautionary in nature. Pile strikes gen- erate a sound field that is seldom known or modeled. Furthermore, the presence, distribution, and behav- ior of fishes of concern, particularly species within regions at risk of high level sound exposures, are also unknown. Over time, information from monitor- ing activities has accumulated, and it is now possible to review available reports and peer reviewed publi- cations. These reports can be used to provide initial estimates of the observed level of sound likely to be found at various monitoring ranges for a variety of pile types, hammer types, and environmental settings. The same cannot be said for observations of the behavioral effects on fishes from pile driving expo- sure. Because of the lack of behavioral information, regulators tend to preferentially permit operations during times when species of concern are unlikely to be present or only present in small numbers. Pile drivers seem to have accepted restrictive schedules in consideration of the expected high cost and com- plexities of assessing the exposure and impacts to fishes during times when the species of concern may be present. While exposure criteria are defined for a point in the volume of water ensonified by pile driving sounds, both the generated sound field and the risks to exposed fishes are four-dimensional (time and three-dimensional location in the ensonified vol- ume). The energy in the generated sound field is continuously variable from the pile to the range at which the energy falls below the ambient noise. The amount of energy at any point in the field depends upon many factors, and it is transient because of the intermittent nature of pile driving with one pile strike every 1.5 seconds or so. A fish would experience sound exposure on the order of 0.01 seconds in every 1.5 seconds, with the amount of exposure energy highly dependent on the location of the fish in the water column rel- ative to the pile being driven. The instantaneous amount of energy a fish may be exposed to decreases rapidly with distance from the pile. In sound propagation models, the decrease in energy is referred to as transmission loss. Typical models for transmission loss, in decibels relative to a μPa in deep and shallow water, are 20 log R and 15 log R respectively, where R is the range (i.e., distance) from the sound source and log is the logarithm base 10. However, the majority of pile driving projects take place in shallow water. Yet, given a measure of the 19

sound at a short range from the source, such as SPLpeak at 10 m, it is possible to obtain an estimate of the SPLpeak of the sound impulse at greater ranges from the source (Au and Hastings, 2008). The barotrauma risk to a fish is a function of the probability of its location in the sound field and the energy in the incident sound on the fish (for each strike impulse and cumulative over all impulses) while it is in that location. Because of the absence of behavior-based exposure models, it is usually assumed that the fishes of interest are stationary in the sound field over the duration of exposure, which is typically defined as the time required to drive a pile. It is interesting to consider the risk of baro- trauma to fishes from particular exposure criteria, considering the characteristics of a propagating sound field generated by pile driving. Assume that an RWI value of 2 was selected as the maximum acceptable physiological exposure response of juve- nile Chinook salmon to pile driving sound. Refer- ring to Figure 8, it can be seen that if SELss was limited to 180 dB re 1 μPa2s and the number of strikes required to drive the pile was less than 960, the SELcum received by the fish would be less than 210 dB re 1 μPa2s. The exposure criteria by permit would then be SELss ≤ 180 dB re 1 μPa2s and SELcum ≤ 210 dB re 1 μPa2s. The SPLpeak corresponding to the SELss criteria can be estimated, using Equation 4.1, to be 207 dB re 1 μPa. If permitting and monitoring con- vention were followed, this would be the criteria at a range of 10 m from the pile at mid-depth. As the range to the pile decreased from the 10 m monitoring location, two factors would change. First, the amount of energy in the sound exposure would increase and second, the affected region around the pile would decrease. For example, mov- ing from a range of 10 m to a range of 5 m, the energy in a single impulse, SELss, would increase by 2.25 dB (a doubling in energy would be 3 dB), from 180 to 182.25 dB re 1 μPa2s, assuming cylindrical spread- ing. Additionally, the region of volume affected would decrease in proportion to the square of the distance by a factor of 4, assuming no water depth change, from approximately 314H m3 to 78 H m3, where H is water depth. At a 10 m range and 180 dB re 1 μPa2s sound exposure, the expected biological response would be an RWI of 3. An RWI of 3 would be 3 Mild injuries or 1 Moderate injury. Given the example criteria, as the distance to the pile decreases from 10 m to 5 m, sound exposure increases by 2.25 dB, the volume affected decreases by a factor of 4, then the biological response would increase by 1 Mild injury (totaling 4 Mild injuries), or by 1 Moderate injury (totaling 2 Moderate injuries). The non-linearity of biological response, RWI, with increasing severity of exposure is evi- dent in Figure 8. Conversely, for this example, as range from the pile doubles from 10 m to 20 m, SELss would decrease to 177.75 dB from 180 dB re 1 μPa2s, affected volume would increase from 314H m3 to 1,257 H m3, and the expected biological response would decrease to an RWI of approximately 1, the equivalent of 1 Mild injury. Derivation of Exposure Criteria Terms used to express exposure criteria relate directly to the pile driving activity and the maximum acceptable impulsive sound exposure at the identi- fied monitoring location in an operating permit. The entity performing the pile diving is typically expected to monitor the generated sound and to mitigate their activities as needed to avoid exceedance of exposure criteria. Possible mitigating actions for underwater sound could include management of the applied energy such as bubble curtains, which might reduce the amount of energy that propagates away from the immediate vicinity of the pile. The research results reported here permit deriva- tion of exposure criteria by starting with a selected level of biological response that protects the individ- uals in an exposed area from injuries that affect per- formance and/or energetics. The selected biological response level and the results of this study can be used to identify the level of exposure that should not be exceeded, i.e., exposure criteria, to assure pro- tection of the fish of concern. The results of this research are specific to juvenile Chinook salmon. Extension of juvenile Chinook salmon’s biological responses to other species and size groups may be possible with consideration of potential differences in biological responses resulting from species’ physio- logical differences. Two cases for derivation of exposure criteria are considered. The first case is for examples where the pile driving duration is short and/or the number of strikes required to drive a pile are less than 960. This is the most likely scenario for the majority of piles driven. The second case is for occasions where the number of strikes required to drive a pile is more than 960 but less than 1,920. 20

Both cases were developed using an RWI of 2 as an acceptable level. The research team’s findings during this study and those of subsequent and ongo- ing research (Carlson et al., 2011) permit us to con- clude that 2 Mild barotrauma injuries do not reduce fish performance and are a conservative threshold for barotrauma injury from exposure to impulsive sound generated by pile driving. Case 1: Number of Strikes ≤ 960, RWI = 2 The lower bound for exposures in this study at the 960 strike level was a SELss of 174 dB re 1 μPa2s with corresponding SELcum of 203 dB re 1 μPa2s (see Table 2). Exposures with fewer strikes were not considered, and extrapolation of RWI to exposures with less than 960 strikes using the function rela- tionships of Figure 8 is not advised. However, for pile driving durations less than 960 strikes the rela- tionships between RWI, SELss, and SELcum shown in Figure 8 can be used to obtain conservative expo- sure criteria. The exposure criteria would be the SELss and SELcum shown in Figure 8 where the RWI curve intersects the horizontal 960 strike line. For an RWI of 2, these values would be a SELss of approximately 180.25 dB re 1 μPa2s and a SELcum of 210 dB re 1 μPa2s. These criteria would be conservative because if the SELss is not exceeded, the SELcum would always be less than 210 dB re 1 μPa2s. For an exposure maximum of 960 strikes at an RWI of 2, more precise estimates of SELss and SELcum can be found using the equations presented in Figure 8. Case 2: Number of Strikes between 960 and 1,920, RWI = 2 If the expected number of strikes to drive a pile is greater than 960 and the selected biological response threshold is an RWI of 2, then the equa- tions in Figure 8 may be used to identify the SELss and SELcum exposure criteria. The relationships be- tween number of strikes, SELss, and SELcum pre- sented in Figure 8 are shown below as equations 5.1 and 5.2. RWI exp( 30.050 SEL Str cum= − + − ( . . ( 0 149 0 000171 g g ike Equation))) .− 1 5 2 SEL SEL Log Strike Equationcum ss= + ( )10 5 1. In both Equations 5.1 and 5.2, Strike means the num- ber of strikes. Given two of the four variables in these equations, the other two variables can be found. For a pile driving project, the variables most likely to be known a priori will be the RWI and SELss. The reason is that acceptable biological responses will be the initial selected variable in consultation between regulatory agencies and those wanting to drive piles. In addition, the type of pile, hammer, and pile driving conditions will likely be determined dur- ing project planning, and available information will permit estimation of the expected SELss value. Given SELss and RWI, it is most likely accurate to use Figure 8 to estimate SELcum and the number of strikes available to drive a pile. In practice, the expectations intended for fish protection are implicit in the selection of an RWI value and the implication for underwater pile driving sound generation that will require negotiable tradeoffs using Figure 8 and/or Equations 5.1 and 5.2. For example, if the selected RWI is 2 and the expected SELss is greater than 181 dB re 1 μPa2s, a solution for SELcum and number of strikes is not pos- sible within the bounds of the data for this study. Either the RWI value would need to be increased or the pile driver would need to identify strategies that would reduce SELss. Alternatively, the negotiating parties might elect to extrapolate the RWI curves outside the range of experimental data to obtain solutions. Other strategies for exposure criteria derivation could be to permit higher SELss outside of the solution range of experimental data, or limiting SELcum and RWI within the solution range. An example would be permitting SELss of 182 dB re 1 μPa2s but limiting SELcum to 208 dB re 1 μPa2s. Such strategies would most likely pose a high risk of exceedance of RWI; however, an SELcum of 208 would seem adequate to prevent exceedance of an RWI of 2 (Figure 8). The risk of exceeding the RWI threshold would be partic- ularly high if the majority of barotrauma injuries at a particular SELss occurred during the initial stages of exposure when the number of strikes was low. CHAPTER 6 FUTURE RESEARCH This study was designed to explore the relation- ships between the number of pile strikes, SELss, and SELcum, with particular reference to Chinook salmon. The results provide a highly quantified view of the most important variables in pile driving and their 21

effects on fish. Yet, this is just the first critical step in the goal of fully quantifying the effects on fishes from impulsive pile driving sounds and developing a full array of criteria to protect fish. However, to fully understand effects of pile driving, it is impera- tive to get quantitative data on other variables asso- ciated with pile driving. These range from impulsive sounds generated by different pile types to responses by different species from impulsive sounds. Future studies should also be directed at refining knowledge and extrapolating to other sounds, other species, and fish of different sizes. The following are a number of such studies that the research team believes are the most critical next steps: • Temporary threshold shifts in hearing sensi- tivity are called TTS, and are a temporary loss of some hearing. Both TTS and barotrauma tissue injury have been used to assess the response of fish to sound exposures. However, their respective positioning along a biological response continuum has not been determined for any species. It would be useful to learn if the biological responses for each assay lies along a continuum for impulsive sound expo- sure. If the responses fall on a continuum, then a useful comparison would be TTS and baro- trauma sensitivities modeled as measures for the onset of effect and/or injury in fish from exposure to impulsive sound. • Studies on the effects of decompression on fish have shown that the magnitude of the ratio of pressure to which fish are acclimated and the pressure at which fish are exposed is propor- tional to the severity of barotrauma injury. If this ratio extends to pile driving and seismic impulsive sounds, it would introduce depth as a variable into the assessment of the effects of these sounds. The result would be a rapid decrease in the severity of exposure and response from relatively small changes in depth, given that the static pressure in water increases by about 100k Pa per 10 m of depth. Research is needed to determine the rela- tionship between acclimation- and exposure- pressures, and if response severity is the same for impulsive sound exposure as it is for decompression. • The results of this study show that severity of responses increases with increased exposure to impulsive pile driving sounds. However, the scope of the study did not permit investigation into how injuries accrued during exposure. It is possible that injury accumulation was almost complete within the initial exposure- strikes and that the increase in injury accumu- lation rate and their severity was not uniform over the course of exposure. Understanding the growth of effects is essential to more accu- rately estimate biological response. • The use of the results from studies such as the one described here must deal with the ques- tion of applicability across fish species and sizes within species. The most conservative approach is not to extrapolate the results of a study to other species and size groups. How- ever, the reality is that data are few and quite difficult to obtain. Thus, in most situations, it is better to use available data to inform a deci- sion. Indeed, and as discussed in Chapters 4 and 5, it is possible to generalize the results of this study with caution. While the most thorough approach would be to do the same detailed studies as done here on many other species, the time and expense of such studies make that approach prohibitive. However, it would be possible, and of great value, to do a limited set of studies using a defined array of exposures across a range of carefully selected species to represent the diversity of fish struc- ture and physiology, as well as fish under the most regulatory-concern. This, combined with the extensive results reported here, would per- mit a broader understanding of the compre- hensive effects, which could serve as a guide for extrapolation to a wider range of species. • The least significant class of barotrauma injury defined in this study (Mild) includes those injuries that are not likely to impose a physiological burden that will affect the health of a fish. This assessment is the basis for selection of a RWI value of 2 as an accept- able response threshold for derivation of expo- sure criteria to impulsive sound. However, it is also possible that higher levels of injury would not pose physiological risk to fish health. Through experiments that assess the fitness of fish with Moderate injuries, it would be possi- ble to estimate the level of injury fish can expe- rience without impact on overall health. In the absence of such data, regulation to protect fish must be very conservative to assure protection 22

for exposed animals, which is why this study recommends an RWI of 2. • Exploration of assays to detect the presence of specific proteins (biomarkers) in blood offer the potential for non-lethal assessment of injury from exposure to impulsive sound and other exposures. Biomarkers used to detect traumatic brain injury (TBI) in humans have been found to be a candidate biomarker to detect sub-mortal injury in fish exposed to decompression and impact. There is a growing literature that impulsive signals can cause TBI and that the effects do not show up for some time post exposure. Thus, biomarkers could be an important response assessment tool particu- larly for fish that are difficult to necropsy be- cause of value (such as adult breeding females), handling challenges, or size limitations (larval fish or full-grown adult). • There have been suggestions that the cumula- tive effects on fish from the natural pauses that occur during pile driving activity result in a resetting of the injury accumulations. For example, would the effect on fish be the same from a group of 960 continuous strikes com- pared to a group of 96 strikes presented 10 times with a defined time separation? It would be useful for mitigation purposes to determine the injury response levels from these two paradigms and address whether or not natural pauses lead to different injury response levels than continuous activity. • While it has been predicted that effects of pile driving would decrease as fishes get larger (e.g., Carlson et al., 2007), this could not be examined in the results reported here due to the size of the HICI-FT. Understanding such effects would have significant implications for setting criteria for effects of sound on fish. REFERENCES Au, W. L., and M. Hastings. 2008. Principles of Marine Bioacoustics. New York: Springer-Verlag, p. 20–90. California Department of Transportation. 2010a. Effects of Pile Driving Sound on Juvenile Steelhead. Pre- pared by ICF Jones & Stokes, Seattle, Washington. Available at http://www.dot.ca.gov/hq/env/bio/files/ madriver_cagedfsh.pdf. California Department of Transportation. 2010b. Necropsy and Histopathology of Steelhead Trout Exposed to Steel Pile Driving at the Mad River Bridges, U.S. Highway 101, July 2009. March. Prepared by G. D. Marty, DVM, Ph.D., Fish Pathology Services, Abbots- ford, British Columbia, Canada. Carlson, T. J., G. E. Johnson, C. M. Woodley, J. R. Skalski, and A. G. Seaburg. 2011. Compliance Monitoring of Underwater Blasting for Rock Removal at Warrior Point, Columbia River Channel Improvement Project, 2009/2010. PNNL-20388, Pacific Northwest National Laboratory, Richland, Washington. Carlson, T. J., and M. A. Weiland. 2007. Dynamic Pile Driving and Pile Driving Underwater Impulsive Sound. PNWD-3808, Battelle—Pacific Northwest Division, Richland, Washington. Carlson, T. J., M. C. Hastings, and A. N. Popper. 2007. Update on recommendations for revised interim sound exposure criteria for fish during pile driving activities. Available at http://www.dot.ca.gov/hq/ env/bio/files/ct-arlington_memo_12-21-07.pdf (February 2011). Casper, B. M., and D. A. Mann. 2006. Evoked Potential Audiograms of the Nurse Shark (Ginglymostoma cir- ratum) and the Yellow Stingray (Urobatis jamaicen- sis). Environ. Biol. Fish. 76(1):101–110. Chawda, M. N., F. Hildebrand, H. C. Pape, and P. V. Giannoudis. 2004. Predicting outcome after multiple trauma: which scoring system? Injury, Int. J. Care Injured 35:347–358. Dalecki, D., S. Z. Child, and C. H. Raeman. 2002. Lung damage from exposure to low frequency sound. J. Acoust. Soc. Am. 111:2464. Gaspin, J. B. 1975. Experimental investigation of the effects of underwater explosions on swimbladder fish, 1: 1973 Chesapeake Bay Tests. Naval Surface Weapons Center Technical Report NSWC/WOL/ TR 75-58. White Oak Laboratory, Silver Spring, Maryland. Hamernik, R. P., W. Qiu, and B. Davis. 2003. The effects of the amplitude distribution of equal energy expo- sures on noise-induced hearing loss: The kurtosis metric. J. Acoust. Soc. Am. 114:386–395. Hawkins, A. D., and D. N. MacLennan. 1976. An acoustic tank for hearing studies on fish. In Sound Reception in Fish, A. Schuijf and A. D. Hawkins (eds.), pp. 149–169, Elsevier, New York. Husum, H., and G. Strada. 2002. Measuring injury sever- ity. The ISS as good as the NISS for penetrating injuries. Prehosp. Disast. Med. 17:27–32. Houghton, J. P, J. E. Starkes, J. P. Stutes, M. A. Harvey, J. A. Reyff, and D. E. Erikson. 2010. Acoustic mon- itoring of in situ exposures of juvenile coho salmon to pile driving noise at the Port of Anchorage Marine Terminal redevelopment project, Knik Arm, Alaska. In Alaska Marine Science Symposium Book of 23

Abstracts, p. 83. Available at http://doc.nprb.org/web/ symposium/2010/2010_AMSS_Abstract_Book.pdf (February 2011). Iwama, G. K., A. D. Pickering, J. P. Sumpter, and C. B. Schreck (eds.). 1997. Fish stress and health in aqua- culture, pp. 13–67, Cambridge University Press, New York. Krischer, J. P. 1979. Indexes of severity: Conceptual development. Health Serv. Res. 14(1):56–67. MacGillivray, A., and R. Racca. 2005. Sound Pressure and Particle Velocity Measurements from Marine Pile Driving at Eagle Harbor Maintenance Facility, Bainbridge Island, WA. JASCO Research Ltd., Vic- toria, British Columbia, Canada. Oyetunji, T., J. G. Crompton, D. T. Efron, E. R. Haut, D. C. Chang, E. E. Cornwall III, S. P. Baker, and A. H. Haider. 2010. Simplifying physiologic injury severity measurement for predicting trauma out- comes. J Surg Res. 159(2):627–632. Popper, A. N., and M. C. Hastings. 2009. Effects of anthropogenic sources of sound on fishes. J. Fish Biol. 75:455–498. Available at http://www.wsdot. wa.gov/NR/rdonlyres/0B027B4A-F9FF-4C88-8DE0- 39B165E4CD94/61427/BA_AnthroSoundonFish.pdf (February 2011). Popper, A. N., T. J. Carlson, A. D. Hawkins, B. L. Southall, and R. L. Gentry. 2006. Interim criteria for injury of fish exposed to pile driving operations: A white paper. Available at http://www.wsdot.wa. gov/NR/rdonlyres/84A6313A-9297-42C9-BFA6- 750A691E1DB3/0/BA_PileDrivingInterimCriteria. pdf (February 2011). Popper, A. N., M. B. Halvorsen, A. Kane, D. Miller, M. E. Smith, J. Song, P. Stein, and L. E. Wysocki. 2007. The effects of high intensity, low frequency active sonar on rainbow trout. J. Acoust. Soc. Amer. 122 (1):623–635. Rogers, P. H., and T. Lewis. 1999. Test Chambers for LFS Animal Exposures. In Measurement of Lung Vibration from Low Frequency Underwater Sound in an Animal Model and Divers Using NIVAMS, P. H. Rogers (principal investigator). Final Report for Navy Contract N0014-97-1-0949, Georgia Institute of Technology, Atlanta. Available at http:// www.dtic.mil/cgi-bin/GetTRDoc?Location=U2& doc=GetTRDoc.pdf&AD=ADA371266 (February 2011). Ruggerone, G. T., S. E. Goodman, and R. Miner. 2008. Behavioral Response and Survival of Juvenile Coho Salmon to Pile Driving Sounds. Natural Resources Consultants, Inc., Seattle, Washington. Available at http://home.comcast.net/∼ruggerone/FishTerminal PileDriveStudy.pdf (February 2011). Stadler, J. H., and D. P. Woodbury. 2009. Assessing the effects to fishes from pile driving: Application of new hydroacoustic criteria. Inter-Noise 2009, Ottawa, Ontario, Canada. Available at ftp://167.131.109.8/ techserv/Geo-Environmental/Biology/Hydroacoustic/ References/Literature%20references/Stadler%20and %20Woodbury%202009.%20%20Assessing%20the %20effects%20to%20fishes%20from%20pile%20 driving.pdf (February 2011). Stephenson, J. R., A. J. Gingerich, R. S. Brown, B. D. Pflugrath, Z. Deng, T. J. Carlson, M. J. Langeslay, M. L. Ahmann, R. L. Johnson, and A. G. Seaburg. 2010. Assessing barotrauma in neutrally and nega- tively buoyant juvenile salmonids exposed to sim- ulated hydro-turbine passage using a mobile aquatic barotrauma laboratory. Fisheries Res. 106: 271–278. Woodbury, D., and J. Stadler. 2008. A proposed method to assess physical injury to fishes from underwater sound produced during pile driving. Bioacoustics 17(1–3):289–291. ACRONYMS A/D analog/digital AEP auditory evoked potentials ANCOVA analysis of covariance BOEMRE Bureau of Ocean Energy Management, Regulation and Enforcement Caltrans California Department of Transportation EVREST evoked response study tool (software/hardware system) HAT hearing assessment tube HICI-FT High Intensity Controlled Impedance Fluid-filled wave Tube IACUC Institutional Animal Care and Use Committee ln natural log MS-222 tricaine methanesulfonate MSL Marine Sciences Laboratory (Battelle) NMFS National Marine Fisheries Service PVC polyvinyl chloride RWI response weighted index SEL sound exposure level SELcum cumulative sound exposure level SELss single-strike sound exposure level SPL sound pressure level SPLpeak peak sound pressure level SPLrms root mean square sound pressure level TBI traumatic brain injury TTS temporary threshold shift APPENDICES A THROUGH H Appendices A through H can be found on the TRB website at http://apps.trb.org/cmsfeed/TRB NetProjectDisplay.asp?ProjectID=763. 24

AUTHOR ACKNOWLEDGMENTS The following individuals contributed to NCHRP Project 25-28: Michele B. Halvorsen, Battelle— Pacific Northwest Division, Sequim, WA, and the University of Maryland, College Park, MD; Bran- don M. Casper, the University of Maryland, College Park, MD; Christa M. Woodley, Battelle—Pacific Northwest Division, Sequim, WA; Thomas J. Carl- son, the University of Maryland, College Park, MD; and Arthur N. Popper, the University of Maryland, College Park, MD. Additional support for this project came from the Bureau of Ocean Energy Management, Review, and Enforcement (BOEMRE–formerly Minerals Man- agement Service) and from the California Depart- ment of Transportation (Caltrans). In particular, the research team would like to thank the project man- ager at BOEMRE, Elizabeth Burkhard, for her guid- ance and insight and David Buehler of ICF for his managing the project for Caltrans and for his expert advice. The research team would also like to express grat- itude to the external group of experts for their advice and guidance on the design of the hearing assessment tube and insight into various questions that arose dur- ing the project: William Ellison, Marine Acoustics Inc.; Richard Fay, Loyola University of Chicago; Mardi Hastings, Georgia Institute of Technology; A.D. Hawkins, Loughine Ltd.; David Mann, Univer- sity of South Florida; James Martin, Georgia Institute of Technology; Peter Rogers, Georgia Institute of Technology; and David Zeddies, Marine Acoustics Inc. Special thanks to James Martin for building and personally delivering the HICI-FT to the University of Maryland, and for countless hours on the phone trouble shooting and advising. The research team would also like to thank those who have contributed extensive technical assistance with data collection at the University of Maryland: Dr. Jiakun Song, David Sanderson-Kilchenstein, Frazer Matthews, and Sara Therrien. The research team is very grateful to Rich Brown, John Stephenson, Timothy Linley, and Piper Benjamin of the Pacific Northwest National Laboratory in Richland, WA, for husbandry and the tasks of preparing, sorting, and shipping the salmon. The research team would also like to extend sin- cere appreciation to facilities management personnel at the University of Maryland for their care, exper- tise, and speed at handling the demands this experi- ment required of plumbing, HVAC, and electricity. The research team would like to specifically thank: Del Propst, Brian Brevig, Steve Russo, Dave Dalo, and Richard Tucci. 25

Transportation Research Board 500 Fifth Street, NW Washington, DC 20001 These digests are issued in order to increase awareness of research results emanating from projects in the Cooperative Research Programs (CRP). Persons wanting to pursue the project subject matter in greater depth should contact the CRP Staff, Transportation Research Board of the National Academies, 500 Fifth Street, NW, Washington, DC 20001. COPYRIGHT INFORMATION Authors herein are responsible for the authenticity of their materials and for obtaining written permissions from publishers or persons who own the copyright to any previously published or copyrighted material used herein. Cooperative Research Programs (CRP) grants permission to reproduce material in this publication for classroom and not-for-profit purposes. Permission is given with the understanding that none of the material will be used to imply TRB, AASHTO, FAA, FHWA, FMCSA, FTA, or Transit Development Corporation endorsement of a particular product, method, or practice. It is expected that those reproducing the material in this document for educational and not-for-profit uses will give appropriate acknowledgment of the source of any reprinted or reproduced material. For other uses of the material, request permission from CRP. Subscriber Categories: Construction • Environment • Highways ISBN 978-0-309-21360-8 9 780309 213608 9 0 0 0 0

Hydroacoustic Impacts on Fish from Pile Installation Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

TRB’s National Cooperative Highway Research Program (NCHRP) Research Results Digest 363: Hydroacoustic Impacts on Fish from Pile Installation explores the prediction and mitigation of the negative impacts on fish from underwater sound pressure during pile and casing installation and removal.

Appendices A through H for RRD 363 are available online.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!