Beaked whales, specifically Blainville's (Mesoplodon densirostris) and Cuvier's (Ziphius cavirostris), are known to feed in the Tongue of the Ocean, Bahamas. These whales can be reliably detected and often localized within the Atlantic Undersea Test and Evaluation Center (AUTEC) acoustic sensor system. The AUTEC range is a regularly spaced bottom mounted hydrophone array covering >350 nm2 providing a valuable network to record anthropogenic noise and marine mammal vocalizations. Assessments of the potential risks of noise exposure to beaked whales have historically occurred in the absence of information about the physical and biological environments in which these animals are distributed. In the fall of 2008, we used a downward looking 38 kHz SIMRAD EK60 echosounder to measure prey scattering layers concurrent with fine scale turbulence measurements from an autonomous turbulence profiler. Using an 8 km, 4-leaf clover sampling pattern, we completed a total of 7.5 repeat surveys with concurrently measured physical and biological oceanographic parameters, so as to examine the spatiotemporal scales and relationships among turbulence levels, biological scattering layers, and beaked whale foraging activity. We found a strong correlation among increased prey density and ocean vertical structure relative to increased click densities. Understanding the habitats of these whales and their utilization patterns will improve future models of beaked whale habitat as well as allowing more comprehensive assessments of exposure risk to anthropogenic sound.
Citation: Hazen EL, Nowacek DP, St. Laurent L, Halpin PN, Moretti DJ (2011) The Relationship among Oceanography, Prey Fields, and Beaked Whale Foraging Habitat in the Tongue of the Ocean. PLoS ONE 6(4): e19269. https://doi.org/10.1371/journal.pone.0019269
Editor: Howard Browman, Institute of Marine Research, Norway
Received: January 6, 2011; Accepted: March 25, 2011; Published: April 27, 2011
This is an open-access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.
Funding: The data collection and analysis was funded by the Office of Naval Research as N00014-08-1-1162. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Beaked whales are one of the least understood marine mammal taxa in the world's oceans, particularly the Mesoplodon genera, all of which are listed as data deficient under the IUCN red list . Due to their incredibly long and deep dives (e.g. > 50 minutes and 1000 meters), they have been extraordinarily difficult to study using typical visual techniques , . Moreover, while Mesoplodon species are found in most of world's oceans, they are often distributed offshore necessitating the use of large ships to study their behavior and distribution . Testament to the difficulty in studying them, ecological studies have been able to focus on little more than data from stranded animals, e.g. diet from stomach contents . Beaked whale species also appear to be particularly sensitive to mid-frequency sonar as a number of mass strandings have occurred coincident with naval exercises . The lack of information on beaked whale ecology has made assessing the potential risk from anthropogenic activity much more difficult. Recent research using short-duration tags around oceanic islands (i.e., where deep waters are close to shore) has provided valuable and insightful data on diving behavior, beaked whale echolocation, and identifying beaked whale prey , , .
Vessel surveys along the western side of Abaco Island, Bahamas and in the Tongue of the Ocean have established what may be a resident population of Blainville's beaked whales inhabits these waters , . The Tongue of the Ocean (TOTO) is a deep-water basin approximately 204 kilometers long and 36 kilometers wide and varying in depth from 1280–2010 meters. The semi-enclosed nature of the TOTO makes it an ideal study site as it contains bathymetric features that include known habitat for Blainville's beaked whales. The TOTO is especially conducive for studying beaked whales because it is home to a large bottom-mounted hydrophone array that is part of the Atlantic Undersea Test and Evaluation Center (AUTEC). Moored hydrophones have been commonly used to understand patterns of vocal animals across a range of temporal scales (e.g. daily, monthly, seasonal ). Blainville's beaked whales have very regular and predictable inter-click intervals making species identification possible from an acoustic recording , . Foraging click trains have been detected successfully using a combination of manual and automated methods from moored hydrophones in the TOTO , .
Deep scattering layers (DSLs) serve as an important prey resource for top predators, particularly in oligotrophic oceanic habitats , ,  and their vertical migration may serve as an important source of mixing in the ocean . The composition of species in deep scattering layers is diverse and can change both temporally and spatially requiring multiple samples and sampling gear , , , . The functional groups comprising DSLs include myctophid fishes and squid, both of which are known to be prey items of Blainville's beaked whales , . Traditional net tows give limited information on depth distributions and average biomass throughout the length of a trawl but allow measurements and species identification of sampled organisms. In contrast, fisheries acoustics offer a minimally invasive approach to measure sound scattering organisms and to provide an acoustic density of prey. When possible, trawl samples can be used to ground truth the acoustically detected organisms including length frequency distributions and species compositions . Acoustic data in the absence of trawls can still provide a relative measure of prey distribution ,  that may help inform models of top predator distribution.
Habitat models are a valuable tool that can help identify factors structuring organismal distribution, abundance, and even behavior . In addition, habitat models are a valuable tool for identifying critical habitat and assembling spatial management strategies for highly mobile species . Habitat models for beaked whales have been created in the Bahamas, eastern tropical Pacific, and Hawaiian islands though most have focused on depth as the primary factor in determining preferable habitat , . Some recent models have used oceanographic parameters from in situ sampling , . Previous researchers have modeled marine mammal acoustic recordings as a function of environmental correlates for rorqual whales and dolphins however most of these recordings document social behaviors rather than foraging events , , . A multi-directional bottom-mounted hydrophone was deployed at Cross seamount southwest of the Kona coast and found a high number of beaked whale foraging clicks that were coincident with increased prey density .
The primary goals of this research were to (1) obtain fine scale spatial and temporal measurements of both prey and beaked whale foraging effort, (2) analyze spatial and temporal scale dependence in patterns of the deep scattering layer, (3) incorporate prey and oceanographic data in models of Mesoplodon habitat in the TOTO. By combining measurements of prey, acoustic recordings of Mesoplodon foraging behavior, and physical oceanography, we were able to examine trophic interactions and identify habitat characteristics that can be used to further assess and protect this deep diving and poorly understood species.
We conducted fisheries acoustic and hydrographic surveys aboard the R/V Revelle (84 m beam length) to examine the distribution of the DSL in the TOTO, Bahamas from September 12th–October 2nd, 2008. The surveys consisted of three east-west cross-basin transects, two N-S along-basin transects, and 8 cloverleaf patterns (Figure 1). Each clover pattern consisted four 5 km transects to examine scale dependence and isotropy along each transect as well as temporal patterns at the intersection of each clover. Each clover was centered on or near bottom-mounted hydrophones to correlate with beaked whale detections.
Acoustic data (38 kHz) were primarily collected at night with one of the clovers and two of the transects surveyed during the day. The SIMRAD EK60 38-DD echosounder (7° beam width) was mounted in the transducer tube and was calibrated before the cruise using standard calibration procedures . All acoustic data were collected using 2 kW transmit power with a 2048 s pulse width, and were processed in 500 m×10 m bins for exploratory analysis. Based on geostatistical techniques described below, all acoustic data were re-binned in 1 km×200 m grid cells for model input. Data were processed using a −90 dB threshold from the beginning of the far field until noise levels became apparent (5 to 1000 m). The data were manually scrutinized for noise spikes or regions of double echoes with bad data values excluded from analysis. Bottom echoes were excluded using an 8 ping filter to smooth the detected bottom with a backstep of 0.5 meters above the detected bottom when shallower than 1000 m. Scattering layer density was summarized using Sv a logarithmic relative density of acoustic scattering organisms (dB) and converted into NASC a linear measure of acoustic density (m2·nmi−2) for arithmetic operations and analyses. In addition, a single target detection algorithm with a −70 dB threshold was used to detect individual scattering organisms in the water column . We were unable to tow nets to sample species distributions of the scattering layer but instead used the acoustic backscatter intensity to represent relative density of potential prey.
Physical data were collected using an autonomous deep microstructure profiler (DMP) for turbulence measurements, a conductivity-temperature-depth sensor (CTD), and expendable bathy-thermographs (XBTs). The CTD had salinity and temperature probes on board and was lowered up to 1000 meters in depth or within 50 meters of the bottom. The DMP was released and retrieved three times per clover and per transect to measure turbulence, diffusivity and water column environmental data with depth. Sensors on the profiler included a standard CTD, as well as dual airfoil probes to measure microstructure shear. These data were used to calculate turbulence dissipation rates and diffusivity in the water column (cm2·s−1, e.g. St. Laurent and Thurnherr 2007). We conducted 13 total CTD and XBT casts in addition to 36 DMP casts providing good spatial coverage of measured temperature and salinity data with depth over our study area.
Acoustic data from the 82 bottom mounted hydrophones of the AUTEC range were simultaneously recorded digitally at a 96 kHz sampling rate. Hydrophones were approximate 1 nautical mile apart except for two finer scale 7-hydrophone arrays. A multi-stage FFT based energy detector has been successfully used for detection of clicks from a variety of echolocating odontocetes including Blainville's beaked whales , . We used an automatic click detector with an inter-click interval criterion of (0.15–1 s) to quantify Blainville's beaked whale clicks, which were subsequently manually scrutinized , . The total hours of detected foraging click trains recorded during our survey were summed for each hydrophone to get a relative measure of foraging effort, which became the response variable for our habitat models. Previous research , ,  has provided solid evidence of the links between foraging effort and echolocation, including ‘regular’ clicks and prey capture or ‘feeding buzzes’. These results allowed us to reliably assign acoustic data to foraging vs. non-foraging behavior for whales on AUTEC.
Spatial associations between foraging click rates, oceanography, and prey data were performed using ArcGIS 9.3 . Although data were collected north of TOTO and towards Abaco bank, the hydrophone coverage restricted analysis to waters east of Andros Island and west of the Nassau bank in the Bahamas. All data were projected in Universal Transverse Mercator zone 18 and analyzed in 1 km×1 km horizontal grid cells. This resolution was chosen to minimize data gaps between hydrophones in the TOTO while ensuring adequate resolution for fisheries acoustic and oceanographic data. Bathymetric data were obtained from the Naval Undersea Warfare Center's hydrophone locations as the acoustic data were processed at each of these features. Semivariograms of prey density were used to identify the key spatial scales horizontally and vertically as an input for the geostatistical analysis . To account for autocorrelation in our predictor variables and create uniform surfaces, each dataset was interpolated using a universal kriging function , . This approach allowed us to visually identify hotspots and coldspots in each variable. Each hydrophone location was used to sample the acoustic surface, bottom depth, and microstructure diffusivity as inputs for the models of click density as a function of prey and environment.
Models of foraging intensity as a function of prey metrics, as a function of environmental variables, and as a function of all variables were fitted in R 2.10 . Our response variable was total duration of clicking at each hydrophone concomitant with our cruise, and predictor variables included bottom depth (m), salinity, temperature (°C), diffusivity (cm2/s), backscattering density (dB), single target size (dB), and number of single targets in 1 km×1 km×200 m deep bins. Three model forms were fit to examine factors influencing acoustically measured foraging effort:
We used an iterative regression technique starting with a full model and selectively removing individual terms for both generalized linear (GLM) and generalized additive models (GAM) to fit the data and found the relationships to be largely linear with no additional deviance explained from the additive models. The best performing generalized model was chosen as a function of Akaike's Information Criterion (AIC, ) value for prey models, environmental models, and combined models.
By examining the distribution of acoustic scatterers, we were able to determine ideal binning parameters and analysis approaches. The most apparent acoustic feature in the TOTO was a deep scattering layer with a mean width of approximately 200 meters centered at ca. 500 meters depth. Mean volumetric backscatter at 400–600 meters was greatest at the southern edge of our study site and on the western edge of the basin (Svmean = −71.03 dB) and weakest at eastern edge of the basin (Svmean = −76.3 dB). The acoustic density of scatterers was significantly dispersed in the horizontal (x–y) dimension throughout our study area (Moran's i = −0.02, p<0.01). This correlates with our qualitative observations that the primary scatters were distributed in layers rather than discrete patches. In contrast, the numbers and sizes of single targets per m2 were patchily distributed in the horizontal dimension (Moran's i = 0.07, 0.05 respectively, p<0.01). Semi-variograms for acoustic density in the x–y dimensions corroborated these findings with no discernable sill in the variance (Figure 2). For density as a function of depth (the z-axis), semivariograms showed a consistent sill at 200 m corresponding to the mean width of the deep scattering layer (Figure 2). All acoustic and oceanographic data were subsequently binned in 200-meter increments from 0 to 1000 meters depth and in 1 km horizontal distance.
Lag distance is on the x-axis with semivariance on the y-axis.
Patterns in the distribution of detected single targets were similar to overall scattering volume. Acoustically detected single targets ranged between 1–10 targets per 2500 m2 and had target strengths of −50 dB to −35 dB in size (Figure 3). The number of targets was greatest at 200 meters in depth with just below 10 targets per 2500 m2. The largest acoustic scatterers were between 500–600 meters in depth and had mean target strengths of −35 dB, and the depth of these larger scatterers often overlapped the bottom of the primary scattering layer. In contrast to overall backscattering density patterns, we found both the size and the number of single targets per m2 was not significantly different along the southern and western edge of the basin compared to the northern and eastern portion of the bay (Svmean = −40.7 and −40.8 dB; mean number of targets = 1.18 and 1.16 resepectively). However, without visual confirmation such as cameras or net tows, we are unable to differentiate among various scatterers including the various prey items available to beaked whales.
There was a peak in 9 targets/2500 m2 at 200–250 m, a peak in target size of −35 dB at 550–600 m, and a peak in mean scattering volume at 500 m.
The temporal analysis at the center of each clover showed a maximum change in mean DSL scattering volume of 5.6 dB·m−3 over a period of 5.5 hours (Figure 4). By fitting a line through all of the clover center points, we calculated a mean change of 1 dB·m−3 over 3.2 hours suggesting a relatively static scattering layer over time. While we repeatedly observed a diel vertical migration at sunrise and sunset, the scattering volume at depth did not change significantly within or among days as indicated above. We also found no significant effect of time on whale relative foraging effort (t-test, mean of 53.3% of click duration during daylight, 18.1% sd) thus time of day was not included in our final analyses.
The average change in backscatter was 1.9 dB over 550 minutes.
There were a total of 1857 detected Blainville's beaked whale foraging events (click trains) with durations of up to 69 minutes per event. While it is extremely difficult to detect the number of animals involved in a click train, we used the total amount of time click trains were detected as a proxy for foraging effort. We found the western and southwestern hydrophones had the greatest durations of foraging clicks (Figure 5). There was significant spatial variability in foraging patterns as the total duration of clicks at each hydrophone varied from 0 minutes to over 1500 minutes for the duration of our cruise.
Circles are locations of CTD and MP casts. The western and southwestern areas of the study area had higher foraging effort and greater mean scattering volume.
GLM and GAMs were fit with relative foraging effort as the response against environmental and prey predictor variables. Relative foraging effort was normally distributed with constant variance. In addition, additive models did not offer increased explanatory power over linear models. Iteratively fit GLMs resulted in seven significant variables (Table 1). Bottom depth, salinity from 600–800 meters, and temperature from 200–400 meters had significant and negative effects on foraging effort while backscatter from 400–600 meters, number of single targets from 800–1000 meters, and salinity from 400–600 meters had significant and positive effects on foraging effort (Table 2). The full model explained 54% of the variance in foraging effort of beaked whales (Figure 6). The contrast between salinity at 400–600 meters and 600–800 meters suggests a halocline might have structured foraging effort. Reduced models including only prey variables explained only 34% of the variance in beaked whale foraging. The prey-only models resulted in positive relationships between foraging effort and volumetric backscatter between 400–600 m and single targets between 600–800 m but negative relationships with number of targets between 200–400 m and backscatter between 800–1000 m. Environment only models explained 43% of the variance and had positive relationships with surface salinity (0–200 m) and temperature (400–600 m) while the model showed a negative relationship between foraging effort, bottom depth, and temperature between 200–400 m. These models show good predictive power for foraging effort using environment or prey variables, but the model with the best explanatory power included both prey and environmental predictors.
Research on oceanic top predators such as deep-diving beaked whales requires a suite of technologically advanced tools and analyses. The hydrophone range at AUTEC allowed an unprecedented measurement of the foraging habitat and behavior of these elusive predators. Distribution and density in the deep scattering layer were measured using fisheries acoustics allowing for high-resolution sampling of potential prey with depth. While species identification is near impossible with a single frequency echosounder particularly in diverse assemblages such as the deep scattering layers , we were able to measure relative density of a multispecies aggregation and single targets that include deep sea fish and squid, both primary prey species for Blainville's beaked whales , . Physical oceanographic profiles were taken from CTD casts and microstructure turbulence profiles to examine depth-specific patterns in the pelagia that may serve to aggregate prey or assess suitable beaked whale foraging habitat. We were able to identify discrete foraging events in the TOTO and relate foraging effort to oceanography and prey by combining these advanced measurement techniques in a regression-modeling framework.
Deep scattering layers are common in the world's oceans and serve as an important forage base in often oligotrophic open ocean systems , . We found a single DSL in the TOTO that ranged from 400–600 meters in depth, had a mean Sv (−73.2 dB·m−3), and showed minimal change over time (e.g. vertical migration). Diel vertical migration is a common feature of DSLs where organisms stay below the photic zone during the day to avoid predation and migrate to the surface at night to feed , , . Static, non-migratory layers may be composed of different organisms than migratory layers or may be the same organisms at different life stages or facing different energetic requirements , , . Although no trawl studies have been done examining the DSL in the TOTO, DSLs in the broader Atlantic have been identified as primarily myctophids and pelagic crustaceans by weight with occasional catches of large cephalopods , , . Due to the high diversity of organisms within the DSL and difficulty in sampling at depth, identifying DSL distribution to species is difficult but relative density offers a valuable snapshot of potential beaked whale prey.
Blainville's beaked whales are known to feed primarily on cephalopods and secondarily on myctophids, two of the main components found in deep scattering layers , . Analyzed dive records from tagged Beaked whales showed dives up to 1590 m  with foraging effort beginning at a mean depth of 400 m, continuing through the bottom of the dive and into the ascent, and ceasing clicking at a mean depth of 790 m , . These depths agreed with our findings from generalized linear foraging models where both DSL density between 400–600 meters and single targets greater than 800 m were important in predicting foraging effort. Analyses of the returned echoes on acoustic tags from Mesoplodon also showed a higher capture rate at deeper depths and on larger targets (presumably squid) compared to smaller and shallower targets (presumably myctophids as part of the DSL , . Additionally, including deep single targets improved our foraging models further supporting previous findings of beaked whale foraging behavior , .
Blainsville's beaked whales preferentially used the western edge of the TOTO for foraging and showed no time of day effect on foraging effort. The lack of a diel pattern in vertical distribution of backscatter was echoed by a lack of diel patterns in beaked whale click rates. As beaked whales often forage deep in the water column and below the photic zone , , it is not surprising that time of day had no significant effect (but see ). These results also agree with diel analysis of tag-derived dive behaviors where six Blainville's beaked whales showed similar foraging patterns between day and night, although there was greater time spent at the surface and fewer “bounce” dives at night , . Bounce dives, where beaked whales perform a series of increasingly shallower dives after a single deep dive, were suggested as a method of predator avoidance during day rather than searching for prey.
Generalized linear models showed both environmental and prey predictor variables were important factors influencing Blainville's beaked whale foraging effort. Previous habitat models in the TOTO found that adult Blainville's beaked whales were found at shallower bottom depths (<1000 m) and at higher gradients in depth , . We found a similar pattern with depth serving as an important predictor in both reduced and full models. Tag data from Hawaii suggested beaked whales use deeper waters (>1150 m) for foraging dives, but localized upwelling in the Bahamas may result in higher prey and in turn beaked whale sightings , . Vertical water column structure also provided predictive value of beaked whale foraging effort particularly salinity and temperature between 200–600 meters. The gradient between salinity at 400 and 600 meters in the full model and temperature at 200 and 400 meters in the environment model may be caused by a pycnocline that is correlated with DSL distribution. Ferguson et al.  found the opposite trend with depth (<2000 meters) in the eastern tropical Pacific, but a similar trend for thermocline strength having a positive relationship with Mesoplodon sightings. While vertical profiles of oceanographic parameters are not always available for marine predator habitat models, cetacean habitat models in the California current were equally successful using remotely sensed surface variables as in situ measures , although the vertical structure in the TOTO may differ greatly from the upwelling-driven California current.
Including measures of environmental variables and prey variables explained more variation in beaked whale foraging than either model alone. When possible, both data sources should be included in cetacean foraging models but it is important to ensure both prey and predator measures are collected at consistent and appropriate spatial and temporal scales (e.g. ). Because the reduced model including only environmental variables had greater predictive power than the model with only prey variables, beaked whales were possibly using environmental features (e.g. bottom depth, temperature) to find their prey at the broader scales measured in this study. An alternative hypothesis is that acoustically similar scattering layers may differ in the amount of prey that is of interest to beaked whales requiring either a multi-frequency approach (e.g. ) or additional sampling gear necessary to tease apart the composition of the deep scattering layer. From analyses of fine-scale foraging behavior (e.g. individual foraging events from tag records) relative to environmental and prey correlates, predictive value from prey data might be equally if not more important than environmental variables as shown in models of species such as spinner dolphins (Stenella longirostris) in Hawaii , humpback whales (Megaptera novaeangliae) in the Gulf of Maine  and fin (Balaenoptera physalus), minke (Balaenoptera acutorostrata) and sei whales (Balaenoptera borealis) in Greenland . Because animals interact with their environment at a variety of scales, explicit attention to spatial and temporal scales of ecological processes, sampling methods, and data processing are necessary when making habitat inferences.
Habitat models of both sightings and individual movement data have commonly used hydrographic, bathymetric, and less commonly prey variables to predict the distribution and behavior of marine top predators across a number of spatial and temporal scales. For four dolphin species in the eastern tropical Pacific, thermocline strength, thermocline depth, and surface chlorophyll were all important in predicting habitat . Dolphin sightings were analyzed at a range of scales (2–120 km) relative to oceanographic features and no scale dependence was found  suggesting that dolphins were most likely using their oceanographic environment at broader scales than were measured. Torres et al.  found prey items to be uninformative for foraging bottlenose dolphins in the Florida keys, but used intermittent trawls to describe the prey community potentially leading to a scale mismatch between sampling of prey and predator foraging behavior. While we included diffusivity in our models, it was calculated using measured turbulence levels in the water column often varying at much finer scales than 1 km horizontally or 200 m vertical. It was not surprising that diffusivity did not have a significant effect on beaked whale foraging, but finer scale analyses between acoustically detected organisms and diffusivity may better explain the physical and biological structure in the DSL. Our results suggest beaked whales may use multiple spatial scales to forage on various prey types, potentially tracking broad scale features and using echolocation to find the local maxima in prey density.
Understanding the habitat of marine mammals is important for predicting their distribution, particularly with respect to anthropogenic impacts. Generalized additive models were used to identify suitable sites for harbor porpoise (Phocoena phocoena) marine protected areas as a function of bathymetric and hydrographic variables in Scotland highlighting the utility of habitat models in marine conservation . Bottlenose dolphin (Tursiops truncatus) habitat use within the Moray Firth in Scotland was analyzed to understand how marine mammal species use and existing marine reserve at multiple spatial scales . These studies illustrate how habitat modeling can be a powerful tool that can inform spatially adaptive management of pelagic predators. In the TOTO, our findings suggest increased use of the western edge of the basin compared to the east, which may correspond to an increased risk from sonar exercises.
Little is known about the distribution much less the ecology of beaked whales due to long, deep dives, and short surface intervals , . Most of the diet and life history data originate from stranded animals leaving a gap in our knowledge of their foraging behavior . Our study provides an initial analysis of Blainville's beaked whale foraging habitat through the use of advanced acoustic technology and modeling. Even though we focus on a limited area and time snapshot, we found both prey metrics (number of single targets and density of the DSL) and environmental features (salinity, temperature, bottom depth) influenced foraging effort. While further work is necessary to determine whether this foraging habitat model could be extrapolated to other seasons or areas, it provides necessary insight into how beaked whales interact with their environment. Ultimately, both prey distribution and environmental features could inform spatial management approaches for these elusive and deep diving species in the Bahamas.
We would like to thank Dr. Anna McGregor, Jay Hooper, Ashley Dilley, and Ken Decoteau for their help with data collection and processing in addition to chief scientists Drs. Ian Boyd and Brandon Southall for directing the numerous scientific teams. In addition, this project could not have been completed without the help of scientists and crew aboard the R/V Revelle.
Conceived and designed the experiments: ELH DPN PNH DJM. Performed the experiments: ELH DPN LSL DJM. Analyzed the data: ELH DPN PNH DJM. Contributed reagents/materials/analysis tools: DPN LSL. Wrote the paper: ELH DPN. Performed statistical analyses: ELH PNH.
- 1. IUCN (2010) IUCN2010IUCN Red List of Threatened Species. 2010.4 ed. IUCN Red List of Threatened Species. 2010.4 ed.
- 2. Tyack PL, Johnson M, Soto NA, Sturlese A, Madsen PT (2006) Extreme diving of beaked whales. Journal of Experimental Biology 209: 4238–4253.PL TyackM. JohnsonNA SotoA. SturlesePT Madsen2006Extreme diving of beaked whales.Journal of Experimental Biology20942384253
- 3. Claridge D (2006) Fine-scale distribution and habitat selection of beaked whales. Aberdeen: University of Aberdeen. 119 p.D. Claridge2006Fine-scale distribution and habitat selection of beaked whalesAberdeenUniversity of Aberdeen119
- 4. MacLeod C, Santos M, Pierce G (2003) Review of data on diets of beaked whales: evidence of niche separation and geographic segregation. J Mar Biol Assoc Uk 83: 651–665.C. MacLeodM. SantosG. Pierce2003Review of data on diets of beaked whales: evidence of niche separation and geographic segregation.J Mar Biol Assoc Uk83651665
- 5. Cox TM, Ragen TJ, Read AJ, Vos E, Baird RW, et al. (2006) Understanding the impacts of anthropogenic sound on beaked whales. Journal of Cetacean Research and Management 7: 177–187.TM CoxTJ RagenAJ ReadE. VosRW Baird2006Understanding the impacts of anthropogenic sound on beaked whales.Journal of Cetacean Research and Management7177187
- 6. Johnson M, Madsen PT, Zimmer WM, de Soto NA, Tyack PL (2004) Beaked whales echolocate on prey. Proc Biol Sci 271: Suppl 6S383–386.M. JohnsonPT MadsenWM ZimmerNA de SotoPL Tyack2004Beaked whales echolocate on prey.Proc Biol Sci271Suppl 6S383386
- 7. Baird R, Webster D, McSweeney D, Ligon A, Schorr G, et al. (2006) Diving behaviour of Cuvier's (Ziphius cavirostris) and Blainville's (Mesoplodon densirostris) beaked whales in Hawai 'i. Canadian Journal of Zoology 84: 1120–1128.R. BairdD. WebsterD. McSweeneyA. LigonG. Schorr2006Diving behaviour of Cuvier's (Ziphius cavirostris) and Blainville's (Mesoplodon densirostris) beaked whales in Hawai 'i.Canadian Journal of Zoology8411201128
- 8. Madsen PT, Wilson M, Johnson M, Hanlon RT, Bocconcelli A, et al. (2007) Clicking for calamari: toothed whales can echolocate squid Loligo pealeii. Aquatic Biology 1: 141–150.PT MadsenM. WilsonM. JohnsonRT HanlonA. Bocconcelli2007Clicking for calamari: toothed whales can echolocate squid Loligo pealeii.Aquatic Biology1141150
- 9. MacLeod C, Zuur A (2005) Habitat utilization by Blainville's beaked whales off Great Abaco, northern Bahamas, in relation to seabed topography. Marine Biology 147: 1–11.C. MacLeodA. Zuur2005Habitat utilization by Blainville's beaked whales off Great Abaco, northern Bahamas, in relation to seabed topography.Marine Biology147111
- 10. Van Parijs S, Clark C, Sousa-Lima R (2009) Management and research applications of real-time and archival passive acoustic sensors over varying temporal and spatial scales. Marine Ecology Progress Series 395: 21–36.S. Van ParijsC. ClarkR. Sousa-Lima2009Management and research applications of real-time and archival passive acoustic sensors over varying temporal and spatial scales.Marine Ecology Progress Series3952136
- 11. Madsen PT, Johnson M, de Soto NA, Zimmer WMX, Tyack P (2005) Biosonar performance of foraging beaked whales (Mesoplodon densirostris). J Exp Biol 208: 181–194.PT MadsenM. JohnsonNA de SotoWMX ZimmerP. Tyack2005Biosonar performance of foraging beaked whales (Mesoplodon densirostris).J Exp Biol208181194
- 12. Moretti D, Morrissey R, DiMarzio N, Ward JD. MorettiR. MorrisseyN. DiMarzioJ. WardVerified passive acoustic detection of beaked whales (Mesoplodon densirostris) using distributed bottom-mounted hydrophones in the tongue of the ocean, Bahamas.; 2006 Jan 1. Verified passive acoustic detection of beaked whales (Mesoplodon densirostris) using distributed bottom-mounted hydrophones in the tongue of the ocean, Bahamas.; 2006 Jan 1.
- 13. Ward J, Morrissey R, Moretti D, DiMarzio N, Jarvis S, et al. (2008) Passive acoustic detection and localization of Mesoplodon Densirostris (Blainville's beaked whale) vocalizations using distributed bottom-mounted hydrophones in conjunction with a digital tag (DTAG) recording. Can Acoust 36: 60–66.J. WardR. MorrisseyD. MorettiN. DiMarzioS. Jarvis2008Passive acoustic detection and localization of Mesoplodon Densirostris (Blainville's beaked whale) vocalizations using distributed bottom-mounted hydrophones in conjunction with a digital tag (DTAG) recording.Can Acoust366066
- 14. Ballance L, Pitman R, Fiedler P (2006) Oceanographic influences on seabirds and cetaceans of the eastern tropical Pacific: A review. Prog Oceanogr 69: 360–390.L. BallanceR. PitmanP. Fiedler2006Oceanographic influences on seabirds and cetaceans of the eastern tropical Pacific: A review.Prog Oceanogr69360390
- 15. Domokos r, Seki m, Polovina j, Hawn d (2007) Oceanographic investigation of the American Samoa albacore (Thunnus alalunga) habitat and longline fishing grounds. Fisheries Oceanography 16: 555–572.r. Domokosm. Sekij. Polovinad. Hawn2007Oceanographic investigation of the American Samoa albacore (Thunnus alalunga) habitat and longline fishing grounds.Fisheries Oceanography16555572
- 16. Hazen EL, Johnston DW (2010) Meridional patterns in the deep scattering layers and top predator distribution in the central equatorial Pacific. Fisheries Oceanography 19: 427–433.EL HazenDW Johnston2010Meridional patterns in the deep scattering layers and top predator distribution in the central equatorial Pacific.Fisheries Oceanography19427433
- 17. Dewar WK, Bingham RJ, Iverson RL, Nowacek DP, St. Laurent LC, et al. (2006) Does the marine biosphere mix the ocean? Journal of Marine Research 64: 541–561.WK DewarRJ BinghamRL IversonDP NowacekLC St. Laurent2006Does the marine biosphere mix the ocean?Journal of Marine Research64541561
- 18. Kalish J, Greenlaw C, Pearcy W, Van Holliday D (1986) The Biological And Acoustical Structure Of Sound Scattering Layers Off Oregon. Deep-Sea Res 33: 631–653.J. KalishC. GreenlawW. PearcyD. Van Holliday1986The Biological And Acoustical Structure Of Sound Scattering Layers Off Oregon.Deep-Sea Res33631653
- 19. Benoit-Bird KJ, Au WWL, Brainard RE, Lammers MO (2001) Diel horizontal migration of the Hawaiian mesopelagic boundary community observed acoustically. Marine Ecology Progress Series 217: 1–14.KJ Benoit-BirdWWL AuRE BrainardMO Lammers2001Diel horizontal migration of the Hawaiian mesopelagic boundary community observed acoustically.Marine Ecology Progress Series217114
- 20. Kloser R, Ryan T, Young J, Lewis M (2009) Acoustic observations of micronekton fish on the scale of an ocean basin: potential and challenges. ICES Journal of Marine Science 66: 998–1006.R. KloserT. RyanJ. YoungM. Lewis2009Acoustic observations of micronekton fish on the scale of an ocean basin: potential and challenges.ICES Journal of Marine Science669981006
- 21. O'driscoll R, Gauthier S, Devine J (2009) Acoustic estimates of mesopelagic fish: as clear as day and night? ICES Journal of Marine Science: Journal du Conseil 66: 1310.R. O'driscollS. GauthierJ. Devine2009Acoustic estimates of mesopelagic fish: as clear as day and night?ICES Journal of Marine Science: Journal du Conseil661310
- 22. Simmonds E, MacLennan D (2005) Fisheries acoustics : theory and practice. New York: Blackwell. 437 p.E. SimmondsD. MacLennan2005Fisheries acoustics : theory and practiceNew YorkBlackwell437
- 23. Hazen EL, Friedlaender A, Thompson M, Ware C, Weinrich MT, et al. (2009) Fine-scale prey aggregations and foraging ecology of humpback whales Megaptera novaeangliae. Mar Ecol Prog Series 395: EL HazenA. FriedlaenderM. ThompsonC. WareMT Weinrich2009Fine-scale prey aggregations and foraging ecology of humpback whales Megaptera novaeangliae.Mar Ecol Prog Series395
- 24. Elith J, Graham C, Anderson R, Dudík M (2006) Novel methods improve prediction of species' distributions from occurrence data. Ecography 29: 129–151.J. ElithC. GrahamR. AndersonM. Dudík2006Novel methods improve prediction of species' distributions from occurrence data.Ecography29129151
- 25. Embling C, Gillibrand P, Gordon J, Shrimpton J (2009) Using habitat models to identify suitable sites for marine protected areas for harbour porpoises (Phocoena phocoena). Biological Conservation 143: 267–279.C. EmblingP. GillibrandJ. GordonJ. Shrimpton2009Using habitat models to identify suitable sites for marine protected areas for harbour porpoises (Phocoena phocoena).Biological Conservation143267279
- 26. Barlow J (2006) Cetacean abundance in Hawaiian waters estimated from a summer/fall survey in 2002. Marine Mammal Science 22: 446–464.J. Barlow2006Cetacean abundance in Hawaiian waters estimated from a summer/fall survey in 2002.Marine Mammal Science22446464
- 27. Ferguson M, Barlow J, Reilly S (2005) Predicting Cuvier's (Ziphius cavirostris) and Mesoplodon beaked whale population density from habitat characteristics in the eastern tropical Pacific Ocean. Journal of Cetacean Research and Management 7: 287–299.M. FergusonJ. BarlowS. Reilly2005Predicting Cuvier's (Ziphius cavirostris) and Mesoplodon beaked whale population density from habitat characteristics in the eastern tropical Pacific Ocean.Journal of Cetacean Research and Management7287299
- 28. Oleson E, Wiggins S, Hildebrand J (2007) Temporal separation of blue whale call types on a southern California feeding ground. Animal Behaviour 74: 881–894.E. OlesonS. WigginsJ. Hildebrand2007Temporal separation of blue whale call types on a southern California feeding ground.Animal Behaviour74881894
- 29. Stafford K, Citta J, Moore S, Daher M, George J (2009) Environmental correlates of blue and fin whale call detections in the North Pacific Ocean from 1997 to 2002. Mar Ecol Prog Ser 395: 37–53.K. StaffordJ. CittaS. MooreM. DaherJ. George2009Environmental correlates of blue and fin whale call detections in the North Pacific Ocean from 1997 to 2002.Mar Ecol Prog Ser3953753
- 30. Soldevilla MS, Henderson EE, Campbell GS, Wiggins SM, Hildebrand JA, et al. (2008) Classification of Risso's and Pacific white-sided dolphins using spectral properties of echolocation clicks. The Journal of the Acoustical Society of America 124: 609–624.MS SoldevillaEE HendersonGS CampbellSM WigginsJA Hildebrand2008Classification of Risso's and Pacific white-sided dolphins using spectral properties of echolocation clicks.The Journal of the Acoustical Society of America124609624
- 31. Johnston D, McDonald M, Polovina J, Domokos R, Wiggins S, et al. (2008) Temporal patterns in the acoustic signals of beaked whales at Cross Seamount. Biology Letters 4: 208–211.D. JohnstonM. McDonaldJ. PolovinaR. DomokosS. Wiggins2008Temporal patterns in the acoustic signals of beaked whales at Cross Seamount.Biology Letters4208211
- 32. Foote K (1982) Optimizing copper spheres for precision calibration of hydroacoustic equipment. The Journal of the Acoustical Society of America 71: 742–747.K. Foote1982Optimizing copper spheres for precision calibration of hydroacoustic equipment.The Journal of the Acoustical Society of America71742747
- 33. Soule M, Barange M, Solli H, Hampton I (1997) Performance of a new phase algorithm for discriminating between single and overlapping echoes in a split-beam echosounder. ICES Journal of Marine Science 54: 934–938.M. SouleM. BarangeH. SolliI. Hampton1997Performance of a new phase algorithm for discriminating between single and overlapping echoes in a split-beam echosounder.ICES Journal of Marine Science54934938
- 34. Moretti D, Ward J, Jarvis S, DiMarzio N, Morrissey R, et al. (2002) Open Ocean Marine Mammal Monitoring Using Widely Spaced Bottom Mounted Hydrophones. Journal of Underwater Acoustics 52: 651–668.D. MorettiJ. WardS. JarvisN. DiMarzioR. Morrissey2002Open Ocean Marine Mammal Monitoring Using Widely Spaced Bottom Mounted Hydrophones.Journal of Underwater Acoustics52651668
- 35. ESRI (2009) ArcGis, the Complete Geographical Information System. 2009 ed. Redlands, California: ESRI. ESRI2009ArcGis, the Complete Geographical Information System. 2009 edRedlands, CaliforniaESRI
- 36. Perry JN, Liebhold AM, Rosenberg MS, Dungan J, Miriti M, et al. (2002) Illustrations and guidelines for selecting statistical methods for quantifying spatial pattern in ecological data. Ecography 25: 578–600.JN PerryAM LiebholdMS RosenbergJ. DunganM. Miriti2002Illustrations and guidelines for selecting statistical methods for quantifying spatial pattern in ecological data.Ecography25578600
- 37. Krige D (1951) A Statistical Approach to Some Basic Mine Valuation Problems on the Witwatersrand. Journal of the Chemical, Metallurgical and Mining Society of South Africa 52: 119–139.D. Krige1951A Statistical Approach to Some Basic Mine Valuation Problems on the Witwatersrand.Journal of the Chemical, Metallurgical and Mining Society of South Africa52119139
- 38. Petitgas P (1993) Geostatistics for fish stock assessments: a review and an acoustic application. ICES Journal of Marine Science: Journal du Conseil 50: 285–298.P. Petitgas1993Geostatistics for fish stock assessments: a review and an acoustic application.ICES Journal of Marine Science: Journal du Conseil50285298
- 39. R Development Core Team (2010) R: A Language and Environment for Statistical Computing. 2010 ed. Vienna, Austria: R Foundation for Statistical Computing. R Development Core Team2010R: A Language and Environment for Statistical Computing. 2010 edVienna, AustriaR Foundation for Statistical Computing
- 40. Akaike H (1974) A new look at the statistical model identification. IEEE Transactions on Automatic Control 19: 716–723.H. Akaike1974A new look at the statistical model identification.IEEE Transactions on Automatic Control19716723
- 41. Horne J (2000) Acoustic approaches to remote species identification: a review. Fish Oceanogr 9: 356–371.J. Horne2000Acoustic approaches to remote species identification: a review.Fish Oceanogr9356371
- 42. Stanton TK, Lavery AC, Johnson MP, Madsen PT, Tyack PL (2008) Classification of broadband echoes from prey of a foraging Blainville's beaked whale. J Acoust Soc Am 123: 1753–1762.TK StantonAC LaveryMP JohnsonPT MadsenPL Tyack2008Classification of broadband echoes from prey of a foraging Blainville's beaked whale.J Acoust Soc Am12317531762
- 43. Isaacs JD, Tont SA, Wick GL (1974) Deep scattering layers: vertical migration as a tactic for finding food. Deep Sea Research and Oceanographic Abstracts 21: 651–656.JD IsaacsSA TontGL Wick1974Deep scattering layers: vertical migration as a tactic for finding food.Deep Sea Research and Oceanographic Abstracts21651656
- 44. Hays G (2003) A review of the adaptive significance and ecosystem consequences of zooplankton diel vertical migrations. Hydrobiologia 503: 163–170.G. Hays2003A review of the adaptive significance and ecosystem consequences of zooplankton diel vertical migrations.Hydrobiologia503163170
- 45. Tont SA (1976) Deep scattering layers: Patterns in the Pacific. CalCOFI Reports 18: 112–117.SA Tont1976Deep scattering layers: Patterns in the Pacific.CalCOFI Reports18112117
- 46. Huntley M, Brooks ER (1982) Effects of age and food availability on diel vertical migration of Calanus pacificus. Marine Biology 71: 23–31.M. HuntleyER Brooks1982Effects of age and food availability on diel vertical migration of Calanus pacificus.Marine Biology712331
- 47. Barham E (1957) E. Barham1957The ecology of sonic scattering layers in the Monterey area. The ecology of sonic scattering layers in the Monterey area.
- 48. Hareide N, Garnes G (2001) The distribution and catch rates of deep water fish along the Mid-Atlantic Ridge from 43 to 61°N Fisheries research 51: 297–310.N. HareideG. Garnes2001The distribution and catch rates of deep water fish along the Mid-Atlantic Ridge from 43 to 61°NFisheries research51297310
- 49. Fock H, Matthiessen B, Zidowitz H, Westernhagen H (2002) Diel and habitat-dependent resource utilisation by deep-sea fishes at the Great Meteor seamount: niche overlap and support for the sound scattering layer interception hypothesis. Marine Ecology Progress Series 244: 219–233.H. FockB. MatthiessenH. ZidowitzH. Westernhagen2002Diel and habitat-dependent resource utilisation by deep-sea fishes at the Great Meteor seamount: niche overlap and support for the sound scattering layer interception hypothesis.Marine Ecology Progress Series244219233
- 50. Sutton T, Porteiro F, Heino M, Byrkjedal I, Langhelle G, et al. (2008) Vertical structure, biomass and topographic association of deep-pelagic fishes in relation to a mid-ocean ridge system. Deep-Sea Research II 55: 161–184.T. SuttonF. PorteiroM. HeinoI. ByrkjedalG. Langhelle2008Vertical structure, biomass and topographic association of deep-pelagic fishes in relation to a mid-ocean ridge system.Deep-Sea Research II55161184
- 51. Clarke MR (1996) Cephalopods as Prey. III. Cetaceans. Philosophical Transactions of the Royal Society of London Series B: Biological Sciences 351: 1053–1065.MR Clarke1996Cephalopods as Prey. III. Cetaceans.Philosophical Transactions of the Royal Society of London Series B: Biological Sciences35110531065
- 52. Baird RW, Webster DL, Schorr GS, McSweeney DJ, Barlow J (2008) Diel variation in beaked whale diving behavior. Mar Mammal Sci 24: 630–642.RW BairdDL WebsterGS SchorrDJ McSweeneyJ. Barlow2008Diel variation in beaked whale diving behavior.Mar Mammal Sci24630642
- 53. Schorr G, Baird R, Hanson M, Webster D, McSweeney D, et al. (2009) Movements of satellite-tagged Blainville's beaked whales off the island of Hawai‘i. Endang Species Res doi 10: 203–213.G. SchorrR. BairdM. HansonD. WebsterD. McSweeney2009Movements of satellite-tagged Blainville's beaked whales off the island of Hawai‘i.Endang Species Resdoi 10203213
- 54. Becker E (2007) Predicting seasonal patterns of California cetacean density based on remotely sensed environmental data. Santa Barbara: University of California, Santa Barbara. 303 p.E. Becker2007Predicting seasonal patterns of California cetacean density based on remotely sensed environmental dataSanta BarbaraUniversity of California, Santa Barbara303
- 55. Torres L, Read A, Halpin P (2008) Fine-scale habitat modeling of a top marine predator: do prey data improve predictive capacity? Ecological Applications 18: 1702–1717.L. TorresA. ReadP. Halpin2008Fine-scale habitat modeling of a top marine predator: do prey data improve predictive capacity?Ecological Applications1817021717
- 56. Gauthier S, Horne J (2004) Potential acoustic discrimination within boreal fish assemblages. ICES Journal of Marine Science 61: 836–845.S. GauthierJ. Horne2004Potential acoustic discrimination within boreal fish assemblages.ICES Journal of Marine Science61836845
- 57. Benoit-Bird K, Au W (2003) Prey dynamics affect foraging by a pelagic predator (Stenella longirostris) over a range of spatial and temporal scales. Behavioral Ecology and Sociobiology 53: 364–373.K. Benoit-BirdW. Au2003Prey dynamics affect foraging by a pelagic predator (Stenella longirostris) over a range of spatial and temporal scales.Behavioral Ecology and Sociobiology53364373
- 58. Laidre K, Heide-Jørgensen M, Heagerty P, Cossio A, Bergstroem B, et al. (2010) Spatial associations between large baleen whales and their prey in West Greenland. Mar Ecol Prog Ser 402: 269–284.K. LaidreM. Heide-JørgensenP. HeagertyA. CossioB. Bergstroem2010Spatial associations between large baleen whales and their prey in West Greenland.Mar Ecol Prog Ser402269284
- 59. Redfern J, Barlow J, Ballance L, Gerrodette T, Becker E (2008) Absence of scale dependence in dolphin-habitat models for the eastern tropical Pacific Ocean. Mar Ecol Prog Ser 363: 1–14.J. RedfernJ. BarlowL. BallanceT. GerrodetteE. Becker2008Absence of scale dependence in dolphin-habitat models for the eastern tropical Pacific Ocean.Mar Ecol Prog Ser363114
- 60. Bailey H, Thompson P (2009) Using marine mammal habitat modelling to identify priority conservation zones within a marine protected area. Marine Ecology Progress Series 378: 279–287.H. BaileyP. Thompson2009Using marine mammal habitat modelling to identify priority conservation zones within a marine protected area.Marine Ecology Progress Series378279287