Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

A week in the life of a pygmy blue whale: migratory dive depth overlaps with large vessel drafts

A week in the life of a pygmy blue whale: migratory dive depth overlaps with large vessel drafts Background: The use of multi‑ sensor tags is increasingly providing insights into the behavior of whales. However, due to limitations in tag attachment duration and the transmission bandwidth of the Argos system, little is known about fine ‑ scale diving behavior over time or the reliability of assigning behavioral states based on horizontal move‑ ment data for whale species. How whales use the water column while migrating has not been closely examined, yet the strategy used is likely to influence the vulnerability of whales to ship strike. Here we present information from a rare week long multi‑ sensor tag deployment on a pygmy blue whale (Balaenoptera musculus brevicauda) that pro‑ vided a great opportunity to examine the fine ‑ scale diving behavior of a migrating whale and to compare the occur‑ rence of feeding lunges with assigned behavioral states. Results: The depth of migratory dives was highly consistent over time and unrelated to local bathymetry. The mean depth of migratory dives (~13 m when corrected for the tag position on the whale) was just below the threshold depth (12 m) that blue whales are predicted to travel below to remove the influence of wave drag at the surface. The whale spent 94 % of observed time and completed 99 % of observed migratory dives within the range of large container ship drafts (<24 m). Areas of high residence identified using the horizontal movement data (FastLoc GPS) did not reflect where lunge feeding occurred. Conclusions: The lack of correspondence between areas of high residence inferred from horizontal movement data and the locations where the whale performed feeding lunges highlights the need for further research to determine whether movement models can accurately detect whale feeding areas or only areas of prey searching. While migrat‑ ing, the whale made dives to a depth that is likely to allow it to avoid wave drag and maximize horizontal movement. Although this strategy may reduce energy expenditure during migration, it also placed the whale at greater risk of ship strike for a much longer period than currently thought. If other whales have similar diving behavior to this animal during migration, many whale species may spend much longer periods than currently estimated within the parts of the water column where the risk of ship strike is high. Keywords: Area‑ restricted search, Accelerometer, Balaenoptera musculus brevicauda, Biotelemetry, Diving, Feeding, Migration, Residence time, Ship strike, Wave drag Background by baleen whales that make annual migrations between Annual migrations involve a large input of energy from high-latitude feeding grounds and low-latitude breeding an animal in order to travel successfully between alter- grounds [2–4]. The large majority of research effort on nately favorable environments [1]. Some of the long- whale behavior has focused on their feeding and breed- est migrations of any mammal species are completed ing grounds, with comparatively less known about their behavior while migrating. In order to make migration successful, whales should have evolved ways to reduce *Correspondence: kylieo@alaskasealife.org energy expenditure while migrating between feeding Alaska SeaLife Center, Seward, AK 99664, USA Full list of author information is available at the end of the article © 2016 The Author(s). This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/ publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Owen et al. Anim Biotelemetry (2016) 4:17 Page 2 of 11 and breeding sites. Morphological adaptations such as bouts reached a depth just deeper than that predicted a streamlined body shape to reduce drag, and the use of to be required to remove wave drag based on their body oscillatory propulsion allowing for thrust generation on diameter [13]. Studies have shown that cetaceans regu- both the upward and downward stroke reduce energy larly intersperse shallow dives between their foraging expenditure and increase propulsion efficiency [5]. In dives [14–16], and while it has been recognized that addition to these morphological adaptations, behavioral continuous swimming along the surface of the ocean is adaptations have also been shown in many marine mam- rarely completed by aquatic mammals [17, 18], there have mals, such as the use of energy-efficient stroke and glide been limited data available to determine whether the dive behavior while diving [6]. depths of migrating whales are optimized, i.e., consist- When swimming close to the surface, the production ently just below the depth threshold where wave drag has of wave drag can increase the drag force felt by an animal minimal influence or whether whales utilize other parts by 2.5–5 times that of an animal moving at depth [7–9]. of the water column while migrating. If whales do opti- The impact of wave drag decreases with depth until it mize their travel depths while migrating, then based on becomes negligible at depths of greater than three times measures of body diameter, it is likely that many species the body diameter of the animal [10]. When swimming of whales may be spending a significant amount of time deeper than this threshold at a given speed, the drag within the depth range where the risk of ship strike is forces acting on a whale remain constant with increasing high. depth [10]. Given that whales are an air-breathing mam- The use of biotelemetry and biologging devices to mal, the most efficient and energy-conserving place for a record the behavior of whales has been rapidly expand- whale to swim might be just below this depth threshold, ing in recent years. The majority of data describing the removing the influence of wave drag, while remaining in diving depths of whales come from feeding grounds, close proximity to the surface to reduce the travel dis- with whales shown to target the shallowest section of the tance to their air supply. As a result, the depth that dif- densest patch of prey [19], and increase the amount of ferent species of whales are predicted to swim at while feeding lunges completed per dive when prey is deeper to migrating should vary between species based on varia- account for the increase in travel time required to reach tion in body diameter (Table 1). the prey [19–21]. However, when considering migratory However, an alternate or combined strategy is that behavior, the vast majority of studies focus on the hori- whales also use gliding descents during migration, con- zontal movement of whales in an attempt to delineate verting potential energy from their negative buoyancy population connectivity and the location of unknown to kinetic energy as a result of lift generated from their breeding or feeding areas, e.g., [22–25]. Very little infor- pectoral fins [11]. However, in a theoretical model for mation is available on the diving behavior of whales while elephant seals (Mirounga angustirostris) it has been sug- migrating. In addition, models of horizontal movement gested that there is little to no energetic advantage of are also used to infer potential behavioral states, based on gliding descents over horizontal swimming at a depth of the assumption that periods of slow travel speed and high three times their body diameter when transiting to for- turning angles represent area-restricted search (ARS) aging areas [12]. In addition, one study of North Atlan- behavior and likely foraging areas [26–29]. However, it tic right whales (Eubalaena glacialis) on their feeding is unknown whether these areas accurately reflect where grounds suggested that traveling dives between foraging lunge feeding behavior occurs for whale species. Given Table 1 The depth threshold that various baleen whale species should swim below in order to avoid the impact of wave drag close to the surface Species Length N Girth max Diameter Wave drag depth Wave drag (m) (m) (m) threshold (m) depth-rounded threshold (m) Blue whale (Balaenoptera musculus) 24.72 3 12.19 3.88 11.64 12 Humpback whale (Megaptera novaeangliae) 13.5 29 10.08 3.21 9.63 10 Gray whale (Eschrichtius robustus) 12.29 5 6.86 2.18 6.54 7 North Atlantic right whale (Eubalaena glacialis) 15.02 10 10.3 3.28 9.84 10 Antarctic minke whale 8.36 2872 4.28 1.36 4.08 4 (Balaenoptera bonaerensis) Data on the girth of whale species were taken from [73] (Antarctic minke whales) and [74] (all other species). Diameter is calculated from girth by assuming a circular shape. The wave drag depth-rounded threshold is calculated to represent the level of accuracy of the pressure sensor in the biotelemetry device used in this study Owen et al. Anim Biotelemetry (2016) 4:17 Page 3 of 11 the increase in evidence of whales feeding while migrat- designed to only penetrate 4.5  cm into the tissue, and ing [30–33], and suggestions that some species such as there was only one row of three backward facing barbs, blue whales (Balaenoptera musculus) may feed year- each 10  mm long by 4  mm wide. The tag contained a round [27, 34], understanding whether ARS behavior is three-axis accelerometer and a three-axis magnetom- likely to indicate where lunge feeding occurs is necessary. eter, each sampled at 16  Hz, a pressure sensor (accu- The ability to describe migratory behavior in detail has racy  ±  1  m) and a wet–dry sensor sampled at 4  Hz and been largely constrained due to limitations in tag attach- a light level sensor and fast-response thermistor sam- ment methods and the transmission bandwidth of the pled at 2  Hz. The tag included a FastLoc GPS receiver available telemetry systems. In order to provide a long (Wildtrack Telemetry Systems Ltd, Leeds, UK) and was temporal scale, the amount of data that are collected satellite-linked, programmed to transmit diving and posi- by a tag needs to be reduced. In addition, the high-res- tion information via the Argos system. However, the olution data required to determine when lunge feeding high-resolution data were obtained upon recovery of the behavior occurs and to examine fine-scale diving behav - floating tag. ior are too large to be transmitted, requiring retrieval of the tag, which usually limits the sampling duration. For Identifying lunge feeding behavior cetaceans, the sampling duration is typically 1 day or less In order to determine when feeding lunges occurred because high-resolution, multi-sensor tags are normally throughout the deployment, the minimum specific attached with suction cups (e.g., [35]). Here we present acceleration (MSA) was calculated using the norm of data from a rare week long deployment of a multi-sen- the acceleration vector in all three axes (x, y and z) and sor tag attached with small barbed darts to a pygmy blue subtracting the influence of gravity [37]. The rate of whale (Balaenoptera musculus brevicauda) that provided change in the MSA signal, ‘jerk,’ was also calculated [37]. highly accurate information on the horizontal movement Although these two parameters are calculated from the of the whale (using FastLoc GPS), as well as the dive pro- same sensor, jerk is likely to provide a stronger signal of file and fine-scale movement data collected by a three- rapid changes in the forces acting on a whale at the tag axis accelerometer. Although our sample size is only one location, such as those that occur when the mouth opens animal, it is rare for comparisons of the behavioral states during a feeding lunge [37]. Peaks in MSA have been inferred from horizontal movement data and the actual shown to be a reliable indicator of lunge feeding behav- location of lunge feeding behavior to be able to be made ior by humpback whales (Megaptera novaeangliae), for whale species, due to a lack of accelerometer data at with MSA found to be two to seven times greater dur- the broad temporal scales required to designate behav- ing lunges than during descents or ascents [37]. However, ioral states. Our aim was to describe the behavior of the inspection of the data recorded in this study revealed that whale over the week, in order to (1) determine whether the MSA during descents was often higher than that of the depth of migratory dive behavior matches the predic- lunges (Fig.  1a), making automatic detection difficult. tions made of optimal swimming depth, (2) determine Instances of lunges could be easily identified by eye using whether areas identified as ARS using a behavioral clas - the MSA and jerk signal, with lunges having a distinct sification method from horizontal movement data match rise in MSA and jerk as a result of the gradual increase in areas where lunge feeding behavior occurred and (3) acceleration, followed by rapid deceleration and drop in investigate the environmental factors that may result in both MSA and jerk when the mouth opens (Fig. 1a) [38]. a whale completing exploratory dives and increasing ARS In contrast, periods of fluking had a diamond- or round- behavior while migrating. shaped jerk signal, and both MSA and jerk often had a much lower magnitude than during lunges (Fig.  1b, c). Methods As a consequence, the entire tag record was inspected by Tagging operations eye to identify times when feeding lunges occurred. The A Whale Lander tag (Wildlife Computers, Redmond, depth at which each lunge occurred was also determined. WA, USA) was deployed onto a pygmy blue whale from Due to the impact that wave drag can have on the whale a 5.5-m rigid-hulled inflatable using a handheld 6.5-m and on the accelerometer signals expected during lunge carbon fiber pole. The Whale Lander tag (8.9  cm diam - feeding at the surface [39], any potential lunges shallower eter, 6.5  cm tall) was anchored to the dermis with three than 10 m were excluded from the analysis. titanium darts, similar to the darts on the compact dou- ble dart satellite tag [36] that we now call the Low Impact Diving behavior Minimally Percutaneous External-electronics Trans- A zero offset correction of the pressure sensor data was mitter (LIMPET) system, except that these darts were completed using the ‘filter’ method within the diveMove Owen et al. Anim Biotelemetry (2016) 4:17 Page 4 of 11 a 0 -200 -400 0 2 4 6 8 10 12 14 16 18 20 1.5 0.5 0 2 4 6 8 10 12 14 16 18 20 -1 -2 0 2 4 6 8 10 12 14 16 18 20 Time (minutes) b 0 -10 -20 -30 0 2 4 6 8 10 12 14 16 1.5 0.5 0 2 4 6 8 10 12 14 16 -1 -2 0 2 4 6 8 10 12 14 16 Time (minutes) -20 -40 -60 0 2 4 6 8 10 12 14 16 18 1.5 0.5 0 2 4 6 8 10 12 14 16 18 -1 -2 0 2 4 6 8 10 12 14 16 18 Time (minutes) Fig. 1 Dive profiles, minimum specific acceleration (MSA) and jerk of the three dive types (a feeding dive, b migratory dive and c exploratory dive) completed by a pygmy blue whale off the southwest coast of Australia. Lunges (red dots) were identified by looking for periods of increasing MSA and jerk, followed by a rapid drop in both parameters Owen et al. Anim Biotelemetry (2016) 4:17 Page 5 of 11 package [40] in R [41] with a minimum depth of 5 m set Environmental factors influencing exploratory dives as the definition of a dive. This package was also used and residence time to produce a series of statistics for each dive includ- The FastLoc GPS positions were annotated with envi - ing the maximum dive depth and mean bottom depth, ronmental data using a bilinear interpolation method which is the mean dive depth of the whale from the end through the EnvDATA system of Movebank [48]. of the descent to the beginning of the ascent. Each dive Bathymetry data used were the ETOPO1 topography and was categorized into one of three dive types (1) feed- bathymetry grid [49], and the chlorophyll a and sea sur- ing dives  =  dives containing a lunge, (2) migratory face temperature (SST) data (both 8 day temporal resolu- dives  =  dives containing no lunge and varying in depth tion and 4  km spatial resolution) were sourced from the less than 15 m from the mean of the maximum depth of MODIS Ocean dataset from NASA’s OceanColor Web the 10 surrounding dives (five dives before and five after), [50]. These data were used to visually determine whether (3) exploratory dives  =  dives containing no lunge that any patterns existed between the environmental condi- were more than three times deeper than the mean of the tions and when the whale was determined to be in an maximum depth of the 10 surrounding dives. ARS state using the horizontal movement data (FastLoc GPS). In addition, patterns in the timing of exploratory Residence time dives in relation to environmental parameters were also The FastLoc GPS data were filtered to remove any posi - examined visually. Due to the small sample size, no statis- tions with a residual value >30 [42]. A speed filter set tical analyses of habitat associations were completed. All to 25  km/h was then applied to the data to remove any data are presented as mean ± standard deviation. additional positions that were likely to be erroneous. The retained positions were interpolated to provide a posi- Results tion of the whale once every hour. The residence time for A pygmy blue whale was tagged with a Lander tag at each position was calculated using the method described 14:23 (local time) on April 30, 2014, at 31.48°S 114.84°E, by Barraquand and Benhamou [43] in the adehabitatLT approximately 35 nautical miles north of the Perth package [44] in R [41]. This method builds on the first Canyon, Western Australia (Fig.  2a). The tag was posi - passage time method [45] by measuring the time spent tioned approximately 1  m down from the centerline in the vicinity of each location. The mean distance moved of the dorsal ridge, in line with midline of the pecto- each hour was determined for the whole trajectory and ral fins. After tagging, the whale moved north, with the three times the mean distance per hour was then used as tag remaining attached for 7.6  days, falling off at 04:53 the radius for the circle within which residence time was (local time) on May 8, 2014, off the coast of Geraldton, calculated [43]. The animal was allowed to move outside Western Australia (Fig.  2a). During that time the whale of the circle for a maximum of 1  h before any re-entry covered a total distance of 506.3 km, with a mean speed into the circle was not included in the residence time of 2.8  ±  2.2  km/h. Movement from the Perth Canyon calculation for each position data point. The method region to Geraldton represents approximately 20  % of of Lavielle [46, 47] was used to partition the trajectory the total migration distance of this population of pygmy based on the residence time of the whale [43]. To do this, blue whales toward Indonesia [34], and so for this reason, the function ‘lavielle’ within the adehabitatLT [44] pack- the whale is assumed to have been migrating during the age in R [41] was used to generate a contrast matrix that deployment. allowed for the identification of the optimal number of The whale completed a total of 37 lunges, 35 of which segments that the trajectory should be broken into based were within the first 30 h after deployment. Depth of the on the residence time. This was completed by identifying lunges ranged from 13 to 414  m (mean 167  m). For the the number of segments where a sharp break in the value first 3–4  h after deployment the whale fed deep within of the contrast function occurs, identifying the number the water column (mean 328 m, minimum 213 m, maxi- of segments required before any further increase in the mum 414  m) during the afternoon, before switching to number of segments does not significantly reduce the shallow surface feeding early the following morning and value of the contrast function. The best location for the afternoon (mean 16 m, minimum 13 m, maximum 25 m) segmentation to occur based on the residence time was (Fig.  3). Inspection of the accelerometer record revealed then calculated and given as a time of the switch from that exclusion of the upper 10 m of the water column due one segment to the next. Segments with high residence to the potential influence of surface effects on accurate times were considered periods of ‘area-restricted search lunge detection is only likely to have influenced the num - (ARS),’ and segments with low residence time were con- ber of lunges determined to be completed during these sidered periods of ‘transit.’ two surface feeding bouts and not the overall location of Owen et al. Anim Biotelemetry (2016) 4:17 Page 6 of 11 Fig. 2 The study site off the southwest coast of Australia displaying (a) the interpolated track of a pygmy blue whale tagged with a Whale Lander tag for 7.6 days. From the deployment location (yellow star) the whale gradually moved north before turning south again just prior to the tag falling off the whale. Black dots indicate where the whale was considered to be in a transit state, and the red dots indicate ARS behavior. In addition, the location of feeding lunges completed by the whale (green squares) as determined by the accelerometer data is shown. Shipping activity in the area is also shown (b) where lunge feeding occurred. After completing these 26.7 min). The consistency in the depth of migratory dives feeding bouts, the whale migrated north, only lunging throughout the duration of the tag deployment was unre- two more times throughout the remaining 6  days of the lated to sea floor depth, with the mean bottom depth of deployment (151.5 h) (Fig. 3). migratory dives being 14 ± 4 m. When taking into consid- eration that the tag was positioned ~1 m below the dorsal Diving behavior midline of the whale, the mean depth of the back of the The whale completed a total of 1677 dives over the duration whale was ~13  m, just below the predicted 12-m depth of the tag attachment, with 86.1 % classified as either feed - threshold that blue whales should swim under to avoid ing, migratory or exploratory dives. The maximum dive the impact of wave drag (Fig. 3). The whale spent 94 % of depth was 506  m, which, to the best of our knowledge, is its time within the upper 24 m of the water column, and deeper than any previously published dive by a blue whale. 99 % of migratory dives had a mean depth shallower than This dive had a duration of 14.8  min and was completed 24 m. during the bout of deep feeding; however, from inspection A total of 51 exploratory dives were completed, with of the accelerometer data, no lunge was completed during a mean time between exploratory dives of 3.3  ±  3.4  h. this dive. A total of 21 feeding dives were identified, with a Although these dives were deeper than surrounding mean maximum depth of 129 ± 183 m (range 13–505 m) dives, review of the accelerometer record confirmed that and mean dive duration of 7.6  min (maximum 17.5  min). no feeding lunges were performed (Fig.  1c). The mean During the bout of deep feeding, mean dive duration was maximum depth of exploratory dives (107 ± 81 m, range 11.4  min, with the duration of feeding dives shortening 23–320  m) was similar to the mean maximum depth of during the bout of shallow feeding to a mean of 5.2 min. feeding dives (129 m) and did not appear to be related to A total of 81.8 % of the dives were classified as migratory sea floor depth. Mean dive duration of exploratory dives dives, with a mean dive duration of 5.2  min (maximum was 8.6 min (maximum 22.05 min). Owen et al. Anim Biotelemetry (2016) 4:17 Page 7 of 11 Fig. 3 The dive profile of the tagged pygmy blue whale for (top) the entire week of the deployment, (middle) the first 18 h after deployment and (bottom) 6 h after the whale commenced its northward movement. Letters above the middle panel mark when the whale initially performed feed‑ ing behavior at depths below 300 m (a), when it switched to a series of exploratory dives (b) and then shallow feeding behavior (c). After this time, the whale began to migrate (d) and continued this behavior for the majority of the remaining 6 days (151.5 h) of the deployment, as well as some additional exploratory dives. The mean bottom depth of migratory dives was often just below the predicted depth threshold (12 m, red line in bot‑ tom panel) that the whale should swim below in order to avoid wave drag Residence time revealed that switching into an ARS state by the whale A total of 481 FastLoc GPS positions were retained did not appear to be correlated with either SST or chloro- after filtering, with a mean time between positions of phyll a concentration. 23 ± 22 min. When interpolated to estimate the position once every hour, the track had a mean distance between Discussion positions of 2.8  km. Based on this step length, a radius The mean bottom depth of migratory dives completed by of 8.4  km was assigned to calculate the residence time the pygmy blue whale was just below the depth thresh- around each position. Inspection of the contrast matrix old where wave drag is likely to have an influence on the revealed that four segments were the optimal number whale. This depth remained consistent over the entire of breaks in the residence time that allowed for a lower week of the deployment, regardless of bottom bathym- contrast function value. Two distinct patches of ARS etry. By swimming just below the depth where surface behavior were located using the FastLoc GPS data, with drag is predicted to be removed, the whale likely placed the whale spending up to 37  h in these ~220  km circu- itself in an area where it was able to minimize the drag lar areas (Fig.  4). Locations of ARS behavior identified it had to overcome and still remain close to the air sup- with the FastLoc GPS data did not correspond to areas of ply at the surface, potentially saving energy while migrat- lunge feeding behavior as identified by the accelerometer ing. The amount of vertical travel completed by the data (Fig. 2a). whale was minimized, allowing for an increase in the horizontal distance covered by the animal in each dive. Environmental factors influencing exploratory dives This potential energy-saving strategy used by the whale and residence time is different to that of some other air-breathing divers that The whale appeared to complete ARS behavior in deep have been shown to complete gliding descents to depths (~1000  m) areas and transit in more shallow areas much greater than three times their body diameter while (~200  m) (Fig.  4). In addition, while the number of traveling [6, 11, 51, 52]. It is possible that factors such as exploratory dives completed in a 10-h period was high- predation risk or foraging along route may play a role est upon concluding a lunge feeding bout, the number of in driving differences in the diving behavior of migrat - exploratory dives also appeared to increase within these ing animals [12]. Future research should investigate this deep areas. Comparisons with environmental parameters pattern further, in order to determine whether other Owen et al. Anim Biotelemetry (2016) 4:17 Page 8 of 11 050 100 150 Time since deployment (hours) 050 100 150 Time since deployment (hours) Fig. 4 The residence time of a pygmy blue whale (a) and the bathymetry (sea floor depth) of the area (b) off the southwest coast of Australia. The behavioral state assigned based on the residence time is shown by the black (ARS) and red (transit) dots individuals show a similar pattern in their diving behav- In addition, given that the migratory path of the whales ior, and whether variation between species is a function and shipping traffic lanes off Western Australia are par - of body diameter or migratory strategy. allel [34], and not perpendicular (Fig.  2b), it is also pos- While reducing drag is likely to reduce energy expendi- sible that shadowing of ship noise by the bow of a ship ture and increase travel speed, the use of this depth range may impact on a whale’s ability to detect a vessel pres- also placed the pygmy blue whale at a greater risk of ship ence [56]. Mortality due to ship strike is thought to be a strike. During a recorded week of travel and feeding, the factor limiting the recovery of some whale populations pygmy blue whale spent 94  % of its time and completed from past exploitation [57]; however, documenting how 99  % of its migratory dives within the possible depth often whales are hit by ships is difficult [58]. It has been range of large container ship drafts (<24 m). While most estimated that the number of whale deaths related to ship container ships have drafts of ~8  m (e.g., Panamax) (the strike is likely to be higher than documented numbers average in the Perth Canyon region is 9  m, maximum due to unrecorded or unrecognized events, and difficul - 15  m), the draft of some container ships is as deep as ties with assigning the cause of death of stranded animals 24  m (e.g., Chinamax). Modeling of the hydrodynamic to ship strike postmortem [58]. The Conservation Man - zone of influence of ships on whales suggests that even agement Plan for blue whales in Australia [59] highlights when a whale is two times the depth of the ships draft, that the risk of ship strike is greatest for calves or feeding propeller suction will still have a significant impact on animals that are thought to spend the greatest amount of the whale, drawing it toward the hull [53]. The lethal time at the surface. However, if other whales exhibit simi- zone for whales was shown to extend to over three times lar behavior to the whale in this study and optimize their the depth of the ships draft when the whale was directly migratory depth to save energy, then many whales may below the ship [53], so even when only considering con- spend a much larger proportion of their time within the tainer ships with moderate drafts (8 m), use of the upper danger zone for ship strikes than currently thought. 24 m of the water column for extended periods is likely to The consistency of the shallow migratory depths also have increased the risk of ship strike for this whale. The has implications for programming of satellite-linked migratory paths of pygmy blue whales along the Western dive depth-transmitting tags placed onto whales. These Australian coast largely overlap with the main shipping devices require that a dive be defined as submergence routes through the area [34] (Fig.  2b). When respond- below some minimum depth to limit the amount of ing to oncoming ships, blue whales have been shown to dive information necessary to compress for transmis- complete a response dive with a slow decent and a lack sion through the very limited Argos bandwidth. In many of movement away from the ship [54]. While in shallow cases, the definition of a dive is set to be potentially sections of the water column, the Lloyd mirror effect deeper than the predicted depth for efficient travel. For can dramatically influence the ability of a whale to deter - example, if predictions hold for each species, Antarctic mine both the direction and distance of a ship [55, 56]. minke whales (Balaenoptera bonaerensis) need to swim Sea Floor Depth (m) Residence time (hours) −1000−400 0 0 10 20 30 Owen et al. Anim Biotelemetry (2016) 4:17 Page 9 of 11 just below 4  m to avoid wave drag (Table  1). Indeed as observed both occurred over deep water areas, with an predicted, inspection of dives referred to as ‘shallow increase in exploratory diving behavior suggesting poten- non-feeding dives’ by Antarctic minke whales in another tial prey searching. Therefore, deep canyons around the study appeared to occur at approximately 5 m depth [16], southwest coast of Australia may provide suitable feeding shallower than the depth threshold set for defining dives habitat for blue whales. However, depending on the timing in many studies [37, 60–62]. If other whales also optimize of local conditions with the arrival of a whale, some areas their dive depths and this is not taken into account when may have varying success as a feeding area year to year. setting dive definition thresholds, then it is possible that Lunge feeding is an energetically expensive strategy with migratory dives may not be recorded by these devices prey density needing to be above a certain level to induce and a biased picture of the depth ranges used by species feeding behavior by a whale [69, 70]. If an animal enters an could be produced. area where they have reliably fed previously, then they may Typically, studies obtaining horizontal movement data initiate ARS behavior to assist with searching for prey, and lack information on the actual behavioral state of the ani- spend longer in the area, yet not lunge feed if dense prey mal and infer it from track metrics [27, 63–65]. Although is not located successfully [71, 72]. Therefore, while hori - the one whale in this study did not lunge feed to a great zontal movement data were not able to successfully locate extent, it provided an opportunity to address the likeli- areas where lunge feeding occurred, they may still provide hood that the location of lunge feeding behavior by the information on potential foraging sites of whales. whale matched areas identified as ARS, and potential foraging locations inferred from horizontal movement Conclusions data. For this whale, the behavioral states inferred from The migratory dive depth of a pygmy blue whale was FastLoc GPS data did not accurately match the locations found to be highly consistent over the week of obser- where lunge feeding occurred. The method used to infer vation, with the whale appearing to optimize its travel behavior was based on an assumption that areas of low depth to reduce the impact of wave drag near the surface, speed and high turning angles result in higher residence yet remain close to the surface where it must return regu- time which is likely to be indicative of foraging behav- larly to breathe. While this strategy may reduce energy ior; a common assumption in many behavioral models expenditure during migration, it also placed the whale at applied to both FastLoc GPS and Argos data [26–29]. greater risk of ship strike for a much longer period than However, on a fine-scale whales have been shown to currently thought. In addition, periods of ARS identi- feed while still maintaining relatively straight trajectories fied using horizontal movement data did not accurately with low turning angles [33]. Other studies have looked reflect where lunge feeding behavior occurred for this at how well behavior inferred from horizontal movement animal, but may still highlight some foraging areas. This data overlaps with presumed feeding in other taxa with suggests that the core assumptions of movement models mixed conclusions. In southern elephant seals (Mirounga should be further tested in relation to fine-scale acceler - leonina), increases in body condition, and likely foraging ometer data, as the behavioral states obtained may not success, have been used to show that the foraging behav- accurately reflect the feeding behavior of whales. Further ior detected in horizontal movement data is a good proxy development of methods for recognizing actual lunge for foraging success [66]. However, in migratory tuna feeding events by the tag software and subsequent trans- it was found that periods of high residence can also be mission of that information over longer time periods will associated with periods of fasting [67], as was observed in also assist with the identification of whale feeding areas. this study. It is possible that for species that spend their entire lives in the marine environment, periods of resi- Abbreviations dency are also likely to be associated with periods of rest- ARS: area‑restricted search; LIMPET: Low Impact Minimally Percutaneous ing, behavior that is often completed on land for species External‑ electronics Transmitter; MSA: minimum specific acceleration; SST: sea surface temperature. that can haul out. More research is needed in this area to determine whether the behavioral states identified from Authors’ contributions horizontal movement data are reflective of the location of KO developed ideas, completed data analyses, wrote the manuscript, RDA developed ideas, collected data, assisted with manuscript preparation, MNJ feeding by baleen whales. collected data, assisted with manuscript preparation, and CSJ collected data, However, when examining the periods of ARS behavior assisted with manuscript preparation. All authors read and approved the final by the tagged pygmy blue whale it is possible to specu- manuscript. late that the animal was indeed searching for prey during Author details these times when no lunges were observed. Off Western 1 2 Alaska SeaLife Center, Seward, AK 99664, USA. Centre for Whale Research Australia, pygmy blue whales are known to feed along the ( WA Inc.), Fremantle, WA 6959, Australia. School of Fisheries and Ocean Sci‑ ences, University of Alaska Fairbanks, Fairbanks, AK 99775, USA. edges of deep water canyons [68]. The two periods of ARS Owen et al. Anim Biotelemetry (2016) 4:17 Page 10 of 11 Acknowledgements 16. Friedlaender AS, Goldbogen JA, Nowacek DP, Read AJ, Johnston D, Gales The authors would like to thank the crew of the Whale Song and especially N. Feeding rates and under ice foraging strategies of smallest lunge filter Simon Kenion for retrieving the tag. This research was completed under Com‑ feeder, the Antarctic minke whale (Balaenoptera bonaerensis). J Exp Biol. monwealth permit number 2013‑00012. 2014;217:2851–4. 17. Fish FE. Aquatic locmotion. In: Tomasi TE, Horton TH, editors. Mammalian Competing interests energetics: interdisciplinary views of metabolism and reproduction. The authors declare that they have no competing interests. Ithaca: Cornell University Press; 1992. p. 34–63. 18. Williams TM. Swimming by sea otters: adaptations for low energetic cost Availability of data and material locomotion. J Comp Physiol A. 1989;164(81):5–24. The dataset supporting the conclusions of this article is available in the 19. Goldbogen JA, Calambokidis J, Croll DA, Harvey JT, Newton KM, Oleson Movebank repository [A week in the life of a pygmy blue whale, https://www. EM, Schorr G, Shadwick RE. Foraging behavior of humpback whales: movebank.org/panel_embedded_movebank_webapp]. kinematic and respiratory patterns suggest a high cost for a lunge. J Exp Biol. 2008;211:3712–9. Ethics approval 20. Tershy B, Acevedo‑ Gutiérrez A, Breese D, Strong C. Diet and feeding Ethical approval for this study was provided by the Western Australian Depart‑ behavior of fin and Bryde’s whales in the Central Gulf of California, ment of Parks and Wildlife (number 2013/45). México. Rev Invest Cient. 1993;1(1):31–7. 21. Doniol‑ Valcroze T, Lesage V, Giard J, Michaud R. Optimal foraging theory Funding predicts diving and feeding strategies of the largest marine predator. Funding for this research was provided by the Centre for Whale Research Behav Ecol. 2011;22:880–8. ( Western Australia). The funding body assisted with data collection and 22. Heide‑ Jørgensen MP, Laidre KL, Jensen MV, Dueck L, Postma LD. Dissolv‑ manuscript preparation. The Whale Lander tag was developed by Wildlife ing stock discreteness with satellite tracking: bowhead whales in Baffin Computers and RDA with funding from the Office of Naval Research and the Bay. Mar Mam Sci. 2006;22(1):34–45. DOD Small Business Technology Transfer (STTR) program. 23. Zerbini AN, Andriolo A, Heide‑ Jørgensen MP, Pizzorno JL, Maia YG, Van‑ Blaricom GR, DeMaster DP, Simões‑Lopes PC, Moreira S, Bethlam C. Satel‑ Received: 2 April 2016 Accepted: 1 August 2016 lite‑monitored movements of humpback whales Megaptera novaeangliae in the Southwest Atlantic Ocean. Mar Ecol Prog Ser. 2006;313:295–304. 24. Garrigue C, Zerbini AN, Geyer Y, Heide‑ Jørgensen MP, Hanaoka W, Clapham P. Movements of satellite‑monitored humpback whales from New Caledonia. J Mamm. 2010;91(1):109–15. 25. Mate BR, Best PB, Lagerquist BA, Winsor MH. Coastal, offshore, and References migratory movements of South African right whales revealed by satellite 1. Alerstam T, Hedenstrom A, Akesson S. Long‑ distance migration: evolution telemetry. Mar Mam Sci. 2011;27(3):455–76. and determinants. Oikos. 2003;103:247–60. 26. Jonsen ID, Mills Flemming J, Myers RA. Robust state space modeling of 2. Rassmussen K, Palacios DM, Calambokidis J, Saborio MT, Dalla Rosa animal movement data. Ecology. 2005;86:2874–80. L, Secchi ER, Steiger GH, Allen JM, Stone GS. Southern Hemisphere 27. Bailey H, Mate BR, Palacios DM, Irvine L, Bograd SJ, Costa DP. Behavioural humpback whales wintering off Central America: insights from estimation of blue whale movements in the Northeast Pacific from water temperature into the longest mammalian migration. Biol Lett. state‑space model analysis of satellite tracks. Endanger Species Res. 2007;3:205–302. 2009;10:93–106. 3. Robbins J, Rosa LD, Allen J, Mattila D, Secchi ER, Friedlaender AS, Stevick 28. Patterson TA, Thomas L, Wilcox C, Ovaskainen O, Matthiopoulos J. Statespace ‑ P, Nowacek D, Steel D. Return movement of a humpback whale between models of individual animal movement. Trends Ecol Evol. 2008;23:87–94. the Antarctic Peninsula and American Samoa: a seasonal migration 29. Patterson TA, Basson M, Bravington MV, Gunn JS. Classifying movement record. Endang Species Res. 2011;13:117–21. behaviour in relation to environmental conditions using hidden Markov 4. Mate BR, Ilyashenko VY, Bradford AL, Vertyankin VV, Tsidulko GA, Rozhnov models. J Anim Ecol. 2009;78:1113–23. VV, Irvine LM. Critically endangered western gray whales migrate to the 30. Best PB, Sekiguchi K, Findlay KP. A suspended migration of humpback eastern North Pacific. Biol Lett. 2015;11:20150071. whales Megaptera novaeangliae on the west coast of South Africa. Mar 5. Fish FE. Transitions from drag‑based to lift ‑based propulsion in mamma‑ Ecol Prog Ser. 1995;118:1–12. lian swimming. Am Zool. 1996;36:628–41. 31. Stockin KA, Burgess EA. Opportunistic feeding of an adult male hump‑ 6. Williams TM, Davis RW, Fuiman LA, Francis J, Le Boeuf BJ, Horning M, back whale (Megaptera novaeangliae) migrating along the coast of south Calambokidis J, Croll DA. Sink or swim: strategies for cost‑ efficient diving eastern Queensland, Australia. Aquat Mamm. 2005;31(1):120–3. by marine mammals. Science. 2000;288:133–6. 32. Visser F, Hartman KL, Pierce GJ, Valavanis VD, Huisman J. Timing of migra‑ 7. Goldman JA. Eec ff ts of the free water surface on animals that jump out of tory baleen whales at the Azores in relation to the North Atlantic spring the water. Ph.D. thesis, Duke University, Durham, NC; 2001 p. 256 bloom. Mar Ecol Prog Ser. 2011;440:267–79. 8. Vennell R, Pease D, Wilson B. Wave drag on human swimmers. J Biomech. 33. Owen K, Warren JD, Noad MJ, Donnelly D, Goldizen AW, Dunlop RA. 2006;39:664–71. Eec ff t of prey type on the fine ‑scale feeding behaviour of migrating east 9. Blake RW. Biological implications of the hydrodynamics of swimming at Australian humpback whales. Mar Ecol Prog Ser. 2015;541:231–44. or near the surface and in shallow water. Bioinspir Biomim. 2009;4:1–9. 34. Double MC, Andrews‑ Goff V, Jenner KCS, Jenner MN, Laverick SM, Branch 10. Hertel H. Structure, form, movement. New York: Reinhold Publishing Co; TA, Gales NJ. Migratory movements of pygmy blue whales (Balaenoptera musculus brevicauda) between Australia and Indonesia as revealed by 11. Gleiss AC, Norman B, Wilson RP. Moved by that sinking feeling: variable satellite telemetry. PLoS One. 2014;9(4):e93578. diving geometry underlies movement strategies in whale sharks. Func 35. Johnson MP, Tyack PL. A digital acoustic recording tag for measuring Ecol. 2011;25(3):595–607. the response of wild marine mammals to sound. IEEE J Ocean Eng. 12. Davis RW, Weihs D. Locomotion in diving elephant seals: physi‑ 2003;28(1):3–12. cal and physiological constraints. Philos Trans R Soc Lond B Biol Sci. 36. Andrews RD, Pitman RL, Ballance LT. Satellite tracking reveals distinct 2007;362(1487):2141–50. movement patterns for Type B and Type C killer whales in the southern 13. Nousek‑McGregor AE. The cost of locomotion in North Atlantic right Ross Sea, Antarctica. Polar Biol. 2008;31:1461–8. whales, Eubalaena glacialis. PhD thesis. Duke University; 2010. 37. Simon M, Johnson M, Madsen PT. Keeping momentum with a mouthful 14. Tyack PL, Johnson M, Aguilar Soto N, Sturlese A, Madsen PT. Extreme div‑ of water: behavior and kinematics of humpback whale lunge feeding. J ing of beaked whales. J Exp Biol. 2006;209:4238–53. Exp Biol. 2012;215:3786–98. 15. Watwood S, Miller PJO, Johnson M, Madsen PT, Tyack P. Deep diving for‑ 38. Goldbogen JA, Calambokidis J, Shadwick R, Oleson EM, McDonald MA, aging behaviour of sperm whales (Physeter macrocephalus). J Anim Ecol. Hildebrand JA. Kinematics of foraging dives and lunge‑feeding in fin 2006;75:814–25. whales. J Exp Biol. 2006;209:1231–44. Owen et al. Anim Biotelemetry (2016) 4:17 Page 11 of 11 39. Owen K, Dunlop RA, Monty JP, Chung D, Noad MJ, Donnelly D, Goldizen 58. Laist DW, Knowlton AR, Mead JG, Collet AS, Podesta M. Collisions AW, Mackenzie T. Detecting surface‑feeding behavior by rorqual whales between ships and whales. Mar Mamm Sci. 2001;17(1):35–75. in accelerometer data. Mar Mamm Sci. 2016;32(1):327–48. 59. Conservation Management Plan for the Blue Whale—A Recovery Plan 40. Luque SP. Dive analysis and calibration. R Package version 1.3.9. CRAN. under the Environment Protection and Biodiversity Conservation Act 2015. R‑project.org/package = diveMove; 2015. 1999, Commonwealth of Australia 2015. 41. R Development Core Team. R: A language and environment for statistical 60. Croll DA, Acevedo‑ Gutiérrez A, Tershy BR, Urban‑Ramırez J. The diving computing, reference index version 3.1.3. R Foundation for Statistical behaviour of blue and fin whales: Is dive duration shorter than expected Computing, Vienna, Austria. 2015. URL http://www.Rproject.org. based on oxygen stores? Compar Biochem Physiol A. 2001;129:797–809. 42. Witt MJ, Åkesson S, Broderick AC, Coyne MS, Ellick J, Formia A, Hays 61. Heide‑ Jørgensen MP, Laidre KL, Wiig Ø, Jensen MV, Dueck L, Maiers GC, Luschi P, Stroud S, Godley BJ. Assessing accuracy and utility of LD, Schmidt HC, Hobbs RC. From Greenland to Canada in ten days: satellite‑tracking data using Argos‑linked Fastloc‑ GPS. Anim Behav. tracks of bowhead whales, Balaena mysticetus, across Baffin Bay. Arctic. 2010;80:571–81. 2003;56(1):21–31. 43. Barraquand F, Benhamou S. Animal movements in heterogeneous 62. Nielsen NH, Laidre K, Larsen RS, Heide‑ Jørgensen MP. identification of landscapes: identifying profitable places and homogeneous movement potential foraging areas for bowhead whales in Baffin Bay and adjacent bouts. Ecology. 2008;89(12):3336–48. waters. Arctic. 2015;68(2):169–79. 44. Calenge C. Analysis of Animal Movements in R: the adehabitatLT Package. 63. Forester JD, Ives AR, Turner MG, Andersen DP, Fortin D, Beyer HL, Smith 2015. R Package version n 0.3.20. CRAN. R‑project.org/package = ade ‑ DW, Boyce MS. State‑space models link elk movement patterns to habitatLT. 2015. landscape characteristics in Yellowstone National Park. Ecol Monograph. 45. Fauchald P, Tveraa T. Using first ‑passage time in the analysis of area‑ 2007;77(2):285–99. restricted search and habitat selection. Ecology. 2003;84(2):282–8. 64. Heithaus MR, Frid A, Wirsing AJ, Dill LM, Fourqurean JW, Burkholder D, 46. Lavielle M. Detection of multiple changes in a sequence of dependent Thomson J, Bejder L. State‑ dependent risk‑taking by green sea turtles variables. Stoch Proc Appl. 1999;83:79–102. mediates top‑ down effects of tiger shark intimidation in a marine ecosys‑ 47. Lavielle M. Using penalized contrasts for the change‑point problem. tem. J Anim Ecol. 2007;76(5):837–44. Signal Process. 2005;85:1501–10. 65. Jonsen ID, Myers RA, James MC. Identifying leatherback turtle foraging 48. Dodge S, Bohrer G, Weinzierl R, Davidson SC, Kays R, Douglas D, Cruz S, behaviour from satellite telemetry using a switching state‑space model. Han J, Brandes D, Wilkelski M. The environmental‑ data automated track Mar Ecol Prog Ser. 2007;337:255–64. annotation (Env‑DATA) system: linking animal tracks with environmental 66. Dragon A, Bar‑Hen A, Monestiez P, Guinet C. Comparative analysis of data. Mov Ecol. 2013;1:3. methods for inferring successful foraging areas from Argos and GPS 49. Amante C, Eakins BW. ETOPO1 1 Arc‑minute global relief model: proce ‑ tracking data. Mar Ecol Prog Ser. 2012;452:253–67. dures, data sources and analysis. NOAA Technical Memorandum NESDIS 67. Bestley S, Patterson TA, Hindell MA, Gunn JS. Feeding ecology of wild NGDC‑24. National Geophysical Data Center, NOAA. 2009. doi:10.7289/ migratory tunas revealed by archival tag records of visceral warming. J V5C8276M. Accessed 1 Oct 2015. Anim Ecol. 2008;77:1223–33. 50. NASA Ocean Biology (OB.DAAC). Sea‑ viewing wide field‑ of‑ view sensor 68. Rennie S, Hanson CE, McCauley RD, Pattiaratchi C, Burton C, Bannister J, (SeaWiFS) ocean color data, 2014.0 Reprocessing. NASA OB.DAAC, Green‑ Jenner C, Jenner MN. Physical properties and processes in the Perth Can‑ belt, MD, USA. doi:10.5067/ORBVIEW‑2/SEAWIFS_OC.2014.0. Accessed 01 yon, Western Australia: Links to water column production and seasonal Dec 2015. pygmy blue whale abundance. J Mar Syst. 2009;77:21–44. 51. Davis RW, Fuiman LA, Williams TM, Le Boeuf BJ. Three‑ dimensional 69. Piatt JF, Methven DA. Threshold foraging behavior of baleen whales. Mar movements and swimming activity of a northern elephant seal. Comp Ecol Prog Ser. 1992;84:205–10. Biochem Physiol A. 2001;129:759–70. 70. Hazen EL, Friedlaender AS, Goldbogen JA. Blue whales (Balaenoptera 52. Sato K, Mitani Y, Cameron MF, Siniff DB, Naito Y. Factors affecting stroking musculus) optimize foraging efficiency by balancing oxygen use and patterns and body angle in diving Weddell seals under natural condi‑ energy gain as a function of prey density. Sci Adv. 2015;1(9):e1500469. tions. J Exp Biol. 2003;206:1461–70. 71. Weimerskirch H. Are seabirds foraging for unpredictable resources? Deep 53. Silber GK, Slutsky J, Bettridge S. Hydrodynamics of a whale/ship collision. Sea Res 2. 2007;54:211–23. J Exp Mar Biol Ecol. 2010;391:10–9. 72. Thums T, Bradshaw CJA, Hindell MA. In situ measures of foraging success 54. McKenna MF, Calambokidis J, Oleson EM, Laist DW, Goldbogen JA. and prey encounter reveal marine habitat‑ dependent search strategies. Simultaneous tracking of blue whales and large ships demonstrates Ecology. 2011;92(6):1258–70. limited behavioral responses for avoiding collision. Endanger Species Res. 73. Konishi K, Tamura T, Zenitani R, Bando T, Kato H, Walløe L. Decline in 2015;27:219–32. energy storage in the Antarctic minke whale (Balaenoptera bonaerensis) 55. Urick RJ. Principles of underwater sound. New York: McGraw‑Hill; 1983. in the Southern Ocean. Polar Biol. 2008;31:1509–20. 56. Terhune JM, Verboom WC. Right whales and ship noise. Mar Mamm Sci. 74. Woodward BL, Winn JP, Fish FE. Morphological specializations of baleen 1999;15:256–8. whales associated with hydrodynamic performance and ecological niche. 57. Irvine LM, Mate BR, Winsor MH, Palacios DM, Bograd SJ, Costa DP, Bailey J Morphol. 2006;267:1284–94. H. Spatial and temporal occurrence of blue whales off the US West Coast, with implications for management. PLoS One. 2014;9(7):e102959. Submit your next manuscript to BioMed Central and we will help you at every step: • We accept pre-submission inquiries • Our selector tool helps you to find the most relevant journal • We provide round the clock customer support • Convenient online submission • Thorough peer review • Inclusion in PubMed and all major indexing services • Maximum visibility for your research Submit your manuscript at www.biomedcentral.com/submit http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Animal Biotelemetry Springer Journals

A week in the life of a pygmy blue whale: migratory dive depth overlaps with large vessel drafts

Loading next page...
 
/lp/springer-journals/a-week-in-the-life-of-a-pygmy-blue-whale-migratory-dive-depth-overlaps-X20BmyPywr
Publisher
Springer Journals
Copyright
Copyright © 2016 by The Author(s)
Subject
Life Sciences; Animal Systematics/Taxonomy/Biogeography; Conservation Biology/Ecology; Terrestial Ecology; Bioinformatics; Freshwater & Marine Ecology
eISSN
2050-3385
DOI
10.1186/s40317-016-0109-4
Publisher site
See Article on Publisher Site

Abstract

Background: The use of multi‑ sensor tags is increasingly providing insights into the behavior of whales. However, due to limitations in tag attachment duration and the transmission bandwidth of the Argos system, little is known about fine ‑ scale diving behavior over time or the reliability of assigning behavioral states based on horizontal move‑ ment data for whale species. How whales use the water column while migrating has not been closely examined, yet the strategy used is likely to influence the vulnerability of whales to ship strike. Here we present information from a rare week long multi‑ sensor tag deployment on a pygmy blue whale (Balaenoptera musculus brevicauda) that pro‑ vided a great opportunity to examine the fine ‑ scale diving behavior of a migrating whale and to compare the occur‑ rence of feeding lunges with assigned behavioral states. Results: The depth of migratory dives was highly consistent over time and unrelated to local bathymetry. The mean depth of migratory dives (~13 m when corrected for the tag position on the whale) was just below the threshold depth (12 m) that blue whales are predicted to travel below to remove the influence of wave drag at the surface. The whale spent 94 % of observed time and completed 99 % of observed migratory dives within the range of large container ship drafts (<24 m). Areas of high residence identified using the horizontal movement data (FastLoc GPS) did not reflect where lunge feeding occurred. Conclusions: The lack of correspondence between areas of high residence inferred from horizontal movement data and the locations where the whale performed feeding lunges highlights the need for further research to determine whether movement models can accurately detect whale feeding areas or only areas of prey searching. While migrat‑ ing, the whale made dives to a depth that is likely to allow it to avoid wave drag and maximize horizontal movement. Although this strategy may reduce energy expenditure during migration, it also placed the whale at greater risk of ship strike for a much longer period than currently thought. If other whales have similar diving behavior to this animal during migration, many whale species may spend much longer periods than currently estimated within the parts of the water column where the risk of ship strike is high. Keywords: Area‑ restricted search, Accelerometer, Balaenoptera musculus brevicauda, Biotelemetry, Diving, Feeding, Migration, Residence time, Ship strike, Wave drag Background by baleen whales that make annual migrations between Annual migrations involve a large input of energy from high-latitude feeding grounds and low-latitude breeding an animal in order to travel successfully between alter- grounds [2–4]. The large majority of research effort on nately favorable environments [1]. Some of the long- whale behavior has focused on their feeding and breed- est migrations of any mammal species are completed ing grounds, with comparatively less known about their behavior while migrating. In order to make migration successful, whales should have evolved ways to reduce *Correspondence: kylieo@alaskasealife.org energy expenditure while migrating between feeding Alaska SeaLife Center, Seward, AK 99664, USA Full list of author information is available at the end of the article © 2016 The Author(s). This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/ publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Owen et al. Anim Biotelemetry (2016) 4:17 Page 2 of 11 and breeding sites. Morphological adaptations such as bouts reached a depth just deeper than that predicted a streamlined body shape to reduce drag, and the use of to be required to remove wave drag based on their body oscillatory propulsion allowing for thrust generation on diameter [13]. Studies have shown that cetaceans regu- both the upward and downward stroke reduce energy larly intersperse shallow dives between their foraging expenditure and increase propulsion efficiency [5]. In dives [14–16], and while it has been recognized that addition to these morphological adaptations, behavioral continuous swimming along the surface of the ocean is adaptations have also been shown in many marine mam- rarely completed by aquatic mammals [17, 18], there have mals, such as the use of energy-efficient stroke and glide been limited data available to determine whether the dive behavior while diving [6]. depths of migrating whales are optimized, i.e., consist- When swimming close to the surface, the production ently just below the depth threshold where wave drag has of wave drag can increase the drag force felt by an animal minimal influence or whether whales utilize other parts by 2.5–5 times that of an animal moving at depth [7–9]. of the water column while migrating. If whales do opti- The impact of wave drag decreases with depth until it mize their travel depths while migrating, then based on becomes negligible at depths of greater than three times measures of body diameter, it is likely that many species the body diameter of the animal [10]. When swimming of whales may be spending a significant amount of time deeper than this threshold at a given speed, the drag within the depth range where the risk of ship strike is forces acting on a whale remain constant with increasing high. depth [10]. Given that whales are an air-breathing mam- The use of biotelemetry and biologging devices to mal, the most efficient and energy-conserving place for a record the behavior of whales has been rapidly expand- whale to swim might be just below this depth threshold, ing in recent years. The majority of data describing the removing the influence of wave drag, while remaining in diving depths of whales come from feeding grounds, close proximity to the surface to reduce the travel dis- with whales shown to target the shallowest section of the tance to their air supply. As a result, the depth that dif- densest patch of prey [19], and increase the amount of ferent species of whales are predicted to swim at while feeding lunges completed per dive when prey is deeper to migrating should vary between species based on varia- account for the increase in travel time required to reach tion in body diameter (Table 1). the prey [19–21]. However, when considering migratory However, an alternate or combined strategy is that behavior, the vast majority of studies focus on the hori- whales also use gliding descents during migration, con- zontal movement of whales in an attempt to delineate verting potential energy from their negative buoyancy population connectivity and the location of unknown to kinetic energy as a result of lift generated from their breeding or feeding areas, e.g., [22–25]. Very little infor- pectoral fins [11]. However, in a theoretical model for mation is available on the diving behavior of whales while elephant seals (Mirounga angustirostris) it has been sug- migrating. In addition, models of horizontal movement gested that there is little to no energetic advantage of are also used to infer potential behavioral states, based on gliding descents over horizontal swimming at a depth of the assumption that periods of slow travel speed and high three times their body diameter when transiting to for- turning angles represent area-restricted search (ARS) aging areas [12]. In addition, one study of North Atlan- behavior and likely foraging areas [26–29]. However, it tic right whales (Eubalaena glacialis) on their feeding is unknown whether these areas accurately reflect where grounds suggested that traveling dives between foraging lunge feeding behavior occurs for whale species. Given Table 1 The depth threshold that various baleen whale species should swim below in order to avoid the impact of wave drag close to the surface Species Length N Girth max Diameter Wave drag depth Wave drag (m) (m) (m) threshold (m) depth-rounded threshold (m) Blue whale (Balaenoptera musculus) 24.72 3 12.19 3.88 11.64 12 Humpback whale (Megaptera novaeangliae) 13.5 29 10.08 3.21 9.63 10 Gray whale (Eschrichtius robustus) 12.29 5 6.86 2.18 6.54 7 North Atlantic right whale (Eubalaena glacialis) 15.02 10 10.3 3.28 9.84 10 Antarctic minke whale 8.36 2872 4.28 1.36 4.08 4 (Balaenoptera bonaerensis) Data on the girth of whale species were taken from [73] (Antarctic minke whales) and [74] (all other species). Diameter is calculated from girth by assuming a circular shape. The wave drag depth-rounded threshold is calculated to represent the level of accuracy of the pressure sensor in the biotelemetry device used in this study Owen et al. Anim Biotelemetry (2016) 4:17 Page 3 of 11 the increase in evidence of whales feeding while migrat- designed to only penetrate 4.5  cm into the tissue, and ing [30–33], and suggestions that some species such as there was only one row of three backward facing barbs, blue whales (Balaenoptera musculus) may feed year- each 10  mm long by 4  mm wide. The tag contained a round [27, 34], understanding whether ARS behavior is three-axis accelerometer and a three-axis magnetom- likely to indicate where lunge feeding occurs is necessary. eter, each sampled at 16  Hz, a pressure sensor (accu- The ability to describe migratory behavior in detail has racy  ±  1  m) and a wet–dry sensor sampled at 4  Hz and been largely constrained due to limitations in tag attach- a light level sensor and fast-response thermistor sam- ment methods and the transmission bandwidth of the pled at 2  Hz. The tag included a FastLoc GPS receiver available telemetry systems. In order to provide a long (Wildtrack Telemetry Systems Ltd, Leeds, UK) and was temporal scale, the amount of data that are collected satellite-linked, programmed to transmit diving and posi- by a tag needs to be reduced. In addition, the high-res- tion information via the Argos system. However, the olution data required to determine when lunge feeding high-resolution data were obtained upon recovery of the behavior occurs and to examine fine-scale diving behav - floating tag. ior are too large to be transmitted, requiring retrieval of the tag, which usually limits the sampling duration. For Identifying lunge feeding behavior cetaceans, the sampling duration is typically 1 day or less In order to determine when feeding lunges occurred because high-resolution, multi-sensor tags are normally throughout the deployment, the minimum specific attached with suction cups (e.g., [35]). Here we present acceleration (MSA) was calculated using the norm of data from a rare week long deployment of a multi-sen- the acceleration vector in all three axes (x, y and z) and sor tag attached with small barbed darts to a pygmy blue subtracting the influence of gravity [37]. The rate of whale (Balaenoptera musculus brevicauda) that provided change in the MSA signal, ‘jerk,’ was also calculated [37]. highly accurate information on the horizontal movement Although these two parameters are calculated from the of the whale (using FastLoc GPS), as well as the dive pro- same sensor, jerk is likely to provide a stronger signal of file and fine-scale movement data collected by a three- rapid changes in the forces acting on a whale at the tag axis accelerometer. Although our sample size is only one location, such as those that occur when the mouth opens animal, it is rare for comparisons of the behavioral states during a feeding lunge [37]. Peaks in MSA have been inferred from horizontal movement data and the actual shown to be a reliable indicator of lunge feeding behav- location of lunge feeding behavior to be able to be made ior by humpback whales (Megaptera novaeangliae), for whale species, due to a lack of accelerometer data at with MSA found to be two to seven times greater dur- the broad temporal scales required to designate behav- ing lunges than during descents or ascents [37]. However, ioral states. Our aim was to describe the behavior of the inspection of the data recorded in this study revealed that whale over the week, in order to (1) determine whether the MSA during descents was often higher than that of the depth of migratory dive behavior matches the predic- lunges (Fig.  1a), making automatic detection difficult. tions made of optimal swimming depth, (2) determine Instances of lunges could be easily identified by eye using whether areas identified as ARS using a behavioral clas - the MSA and jerk signal, with lunges having a distinct sification method from horizontal movement data match rise in MSA and jerk as a result of the gradual increase in areas where lunge feeding behavior occurred and (3) acceleration, followed by rapid deceleration and drop in investigate the environmental factors that may result in both MSA and jerk when the mouth opens (Fig. 1a) [38]. a whale completing exploratory dives and increasing ARS In contrast, periods of fluking had a diamond- or round- behavior while migrating. shaped jerk signal, and both MSA and jerk often had a much lower magnitude than during lunges (Fig.  1b, c). Methods As a consequence, the entire tag record was inspected by Tagging operations eye to identify times when feeding lunges occurred. The A Whale Lander tag (Wildlife Computers, Redmond, depth at which each lunge occurred was also determined. WA, USA) was deployed onto a pygmy blue whale from Due to the impact that wave drag can have on the whale a 5.5-m rigid-hulled inflatable using a handheld 6.5-m and on the accelerometer signals expected during lunge carbon fiber pole. The Whale Lander tag (8.9  cm diam - feeding at the surface [39], any potential lunges shallower eter, 6.5  cm tall) was anchored to the dermis with three than 10 m were excluded from the analysis. titanium darts, similar to the darts on the compact dou- ble dart satellite tag [36] that we now call the Low Impact Diving behavior Minimally Percutaneous External-electronics Trans- A zero offset correction of the pressure sensor data was mitter (LIMPET) system, except that these darts were completed using the ‘filter’ method within the diveMove Owen et al. Anim Biotelemetry (2016) 4:17 Page 4 of 11 a 0 -200 -400 0 2 4 6 8 10 12 14 16 18 20 1.5 0.5 0 2 4 6 8 10 12 14 16 18 20 -1 -2 0 2 4 6 8 10 12 14 16 18 20 Time (minutes) b 0 -10 -20 -30 0 2 4 6 8 10 12 14 16 1.5 0.5 0 2 4 6 8 10 12 14 16 -1 -2 0 2 4 6 8 10 12 14 16 Time (minutes) -20 -40 -60 0 2 4 6 8 10 12 14 16 18 1.5 0.5 0 2 4 6 8 10 12 14 16 18 -1 -2 0 2 4 6 8 10 12 14 16 18 Time (minutes) Fig. 1 Dive profiles, minimum specific acceleration (MSA) and jerk of the three dive types (a feeding dive, b migratory dive and c exploratory dive) completed by a pygmy blue whale off the southwest coast of Australia. Lunges (red dots) were identified by looking for periods of increasing MSA and jerk, followed by a rapid drop in both parameters Owen et al. Anim Biotelemetry (2016) 4:17 Page 5 of 11 package [40] in R [41] with a minimum depth of 5 m set Environmental factors influencing exploratory dives as the definition of a dive. This package was also used and residence time to produce a series of statistics for each dive includ- The FastLoc GPS positions were annotated with envi - ing the maximum dive depth and mean bottom depth, ronmental data using a bilinear interpolation method which is the mean dive depth of the whale from the end through the EnvDATA system of Movebank [48]. of the descent to the beginning of the ascent. Each dive Bathymetry data used were the ETOPO1 topography and was categorized into one of three dive types (1) feed- bathymetry grid [49], and the chlorophyll a and sea sur- ing dives  =  dives containing a lunge, (2) migratory face temperature (SST) data (both 8 day temporal resolu- dives  =  dives containing no lunge and varying in depth tion and 4  km spatial resolution) were sourced from the less than 15 m from the mean of the maximum depth of MODIS Ocean dataset from NASA’s OceanColor Web the 10 surrounding dives (five dives before and five after), [50]. These data were used to visually determine whether (3) exploratory dives  =  dives containing no lunge that any patterns existed between the environmental condi- were more than three times deeper than the mean of the tions and when the whale was determined to be in an maximum depth of the 10 surrounding dives. ARS state using the horizontal movement data (FastLoc GPS). In addition, patterns in the timing of exploratory Residence time dives in relation to environmental parameters were also The FastLoc GPS data were filtered to remove any posi - examined visually. Due to the small sample size, no statis- tions with a residual value >30 [42]. A speed filter set tical analyses of habitat associations were completed. All to 25  km/h was then applied to the data to remove any data are presented as mean ± standard deviation. additional positions that were likely to be erroneous. The retained positions were interpolated to provide a posi- Results tion of the whale once every hour. The residence time for A pygmy blue whale was tagged with a Lander tag at each position was calculated using the method described 14:23 (local time) on April 30, 2014, at 31.48°S 114.84°E, by Barraquand and Benhamou [43] in the adehabitatLT approximately 35 nautical miles north of the Perth package [44] in R [41]. This method builds on the first Canyon, Western Australia (Fig.  2a). The tag was posi - passage time method [45] by measuring the time spent tioned approximately 1  m down from the centerline in the vicinity of each location. The mean distance moved of the dorsal ridge, in line with midline of the pecto- each hour was determined for the whole trajectory and ral fins. After tagging, the whale moved north, with the three times the mean distance per hour was then used as tag remaining attached for 7.6  days, falling off at 04:53 the radius for the circle within which residence time was (local time) on May 8, 2014, off the coast of Geraldton, calculated [43]. The animal was allowed to move outside Western Australia (Fig.  2a). During that time the whale of the circle for a maximum of 1  h before any re-entry covered a total distance of 506.3 km, with a mean speed into the circle was not included in the residence time of 2.8  ±  2.2  km/h. Movement from the Perth Canyon calculation for each position data point. The method region to Geraldton represents approximately 20  % of of Lavielle [46, 47] was used to partition the trajectory the total migration distance of this population of pygmy based on the residence time of the whale [43]. To do this, blue whales toward Indonesia [34], and so for this reason, the function ‘lavielle’ within the adehabitatLT [44] pack- the whale is assumed to have been migrating during the age in R [41] was used to generate a contrast matrix that deployment. allowed for the identification of the optimal number of The whale completed a total of 37 lunges, 35 of which segments that the trajectory should be broken into based were within the first 30 h after deployment. Depth of the on the residence time. This was completed by identifying lunges ranged from 13 to 414  m (mean 167  m). For the the number of segments where a sharp break in the value first 3–4  h after deployment the whale fed deep within of the contrast function occurs, identifying the number the water column (mean 328 m, minimum 213 m, maxi- of segments required before any further increase in the mum 414  m) during the afternoon, before switching to number of segments does not significantly reduce the shallow surface feeding early the following morning and value of the contrast function. The best location for the afternoon (mean 16 m, minimum 13 m, maximum 25 m) segmentation to occur based on the residence time was (Fig.  3). Inspection of the accelerometer record revealed then calculated and given as a time of the switch from that exclusion of the upper 10 m of the water column due one segment to the next. Segments with high residence to the potential influence of surface effects on accurate times were considered periods of ‘area-restricted search lunge detection is only likely to have influenced the num - (ARS),’ and segments with low residence time were con- ber of lunges determined to be completed during these sidered periods of ‘transit.’ two surface feeding bouts and not the overall location of Owen et al. Anim Biotelemetry (2016) 4:17 Page 6 of 11 Fig. 2 The study site off the southwest coast of Australia displaying (a) the interpolated track of a pygmy blue whale tagged with a Whale Lander tag for 7.6 days. From the deployment location (yellow star) the whale gradually moved north before turning south again just prior to the tag falling off the whale. Black dots indicate where the whale was considered to be in a transit state, and the red dots indicate ARS behavior. In addition, the location of feeding lunges completed by the whale (green squares) as determined by the accelerometer data is shown. Shipping activity in the area is also shown (b) where lunge feeding occurred. After completing these 26.7 min). The consistency in the depth of migratory dives feeding bouts, the whale migrated north, only lunging throughout the duration of the tag deployment was unre- two more times throughout the remaining 6  days of the lated to sea floor depth, with the mean bottom depth of deployment (151.5 h) (Fig. 3). migratory dives being 14 ± 4 m. When taking into consid- eration that the tag was positioned ~1 m below the dorsal Diving behavior midline of the whale, the mean depth of the back of the The whale completed a total of 1677 dives over the duration whale was ~13  m, just below the predicted 12-m depth of the tag attachment, with 86.1 % classified as either feed - threshold that blue whales should swim under to avoid ing, migratory or exploratory dives. The maximum dive the impact of wave drag (Fig. 3). The whale spent 94 % of depth was 506  m, which, to the best of our knowledge, is its time within the upper 24 m of the water column, and deeper than any previously published dive by a blue whale. 99 % of migratory dives had a mean depth shallower than This dive had a duration of 14.8  min and was completed 24 m. during the bout of deep feeding; however, from inspection A total of 51 exploratory dives were completed, with of the accelerometer data, no lunge was completed during a mean time between exploratory dives of 3.3  ±  3.4  h. this dive. A total of 21 feeding dives were identified, with a Although these dives were deeper than surrounding mean maximum depth of 129 ± 183 m (range 13–505 m) dives, review of the accelerometer record confirmed that and mean dive duration of 7.6  min (maximum 17.5  min). no feeding lunges were performed (Fig.  1c). The mean During the bout of deep feeding, mean dive duration was maximum depth of exploratory dives (107 ± 81 m, range 11.4  min, with the duration of feeding dives shortening 23–320  m) was similar to the mean maximum depth of during the bout of shallow feeding to a mean of 5.2 min. feeding dives (129 m) and did not appear to be related to A total of 81.8 % of the dives were classified as migratory sea floor depth. Mean dive duration of exploratory dives dives, with a mean dive duration of 5.2  min (maximum was 8.6 min (maximum 22.05 min). Owen et al. Anim Biotelemetry (2016) 4:17 Page 7 of 11 Fig. 3 The dive profile of the tagged pygmy blue whale for (top) the entire week of the deployment, (middle) the first 18 h after deployment and (bottom) 6 h after the whale commenced its northward movement. Letters above the middle panel mark when the whale initially performed feed‑ ing behavior at depths below 300 m (a), when it switched to a series of exploratory dives (b) and then shallow feeding behavior (c). After this time, the whale began to migrate (d) and continued this behavior for the majority of the remaining 6 days (151.5 h) of the deployment, as well as some additional exploratory dives. The mean bottom depth of migratory dives was often just below the predicted depth threshold (12 m, red line in bot‑ tom panel) that the whale should swim below in order to avoid wave drag Residence time revealed that switching into an ARS state by the whale A total of 481 FastLoc GPS positions were retained did not appear to be correlated with either SST or chloro- after filtering, with a mean time between positions of phyll a concentration. 23 ± 22 min. When interpolated to estimate the position once every hour, the track had a mean distance between Discussion positions of 2.8  km. Based on this step length, a radius The mean bottom depth of migratory dives completed by of 8.4  km was assigned to calculate the residence time the pygmy blue whale was just below the depth thresh- around each position. Inspection of the contrast matrix old where wave drag is likely to have an influence on the revealed that four segments were the optimal number whale. This depth remained consistent over the entire of breaks in the residence time that allowed for a lower week of the deployment, regardless of bottom bathym- contrast function value. Two distinct patches of ARS etry. By swimming just below the depth where surface behavior were located using the FastLoc GPS data, with drag is predicted to be removed, the whale likely placed the whale spending up to 37  h in these ~220  km circu- itself in an area where it was able to minimize the drag lar areas (Fig.  4). Locations of ARS behavior identified it had to overcome and still remain close to the air sup- with the FastLoc GPS data did not correspond to areas of ply at the surface, potentially saving energy while migrat- lunge feeding behavior as identified by the accelerometer ing. The amount of vertical travel completed by the data (Fig. 2a). whale was minimized, allowing for an increase in the horizontal distance covered by the animal in each dive. Environmental factors influencing exploratory dives This potential energy-saving strategy used by the whale and residence time is different to that of some other air-breathing divers that The whale appeared to complete ARS behavior in deep have been shown to complete gliding descents to depths (~1000  m) areas and transit in more shallow areas much greater than three times their body diameter while (~200  m) (Fig.  4). In addition, while the number of traveling [6, 11, 51, 52]. It is possible that factors such as exploratory dives completed in a 10-h period was high- predation risk or foraging along route may play a role est upon concluding a lunge feeding bout, the number of in driving differences in the diving behavior of migrat - exploratory dives also appeared to increase within these ing animals [12]. Future research should investigate this deep areas. Comparisons with environmental parameters pattern further, in order to determine whether other Owen et al. Anim Biotelemetry (2016) 4:17 Page 8 of 11 050 100 150 Time since deployment (hours) 050 100 150 Time since deployment (hours) Fig. 4 The residence time of a pygmy blue whale (a) and the bathymetry (sea floor depth) of the area (b) off the southwest coast of Australia. The behavioral state assigned based on the residence time is shown by the black (ARS) and red (transit) dots individuals show a similar pattern in their diving behav- In addition, given that the migratory path of the whales ior, and whether variation between species is a function and shipping traffic lanes off Western Australia are par - of body diameter or migratory strategy. allel [34], and not perpendicular (Fig.  2b), it is also pos- While reducing drag is likely to reduce energy expendi- sible that shadowing of ship noise by the bow of a ship ture and increase travel speed, the use of this depth range may impact on a whale’s ability to detect a vessel pres- also placed the pygmy blue whale at a greater risk of ship ence [56]. Mortality due to ship strike is thought to be a strike. During a recorded week of travel and feeding, the factor limiting the recovery of some whale populations pygmy blue whale spent 94  % of its time and completed from past exploitation [57]; however, documenting how 99  % of its migratory dives within the possible depth often whales are hit by ships is difficult [58]. It has been range of large container ship drafts (<24 m). While most estimated that the number of whale deaths related to ship container ships have drafts of ~8  m (e.g., Panamax) (the strike is likely to be higher than documented numbers average in the Perth Canyon region is 9  m, maximum due to unrecorded or unrecognized events, and difficul - 15  m), the draft of some container ships is as deep as ties with assigning the cause of death of stranded animals 24  m (e.g., Chinamax). Modeling of the hydrodynamic to ship strike postmortem [58]. The Conservation Man - zone of influence of ships on whales suggests that even agement Plan for blue whales in Australia [59] highlights when a whale is two times the depth of the ships draft, that the risk of ship strike is greatest for calves or feeding propeller suction will still have a significant impact on animals that are thought to spend the greatest amount of the whale, drawing it toward the hull [53]. The lethal time at the surface. However, if other whales exhibit simi- zone for whales was shown to extend to over three times lar behavior to the whale in this study and optimize their the depth of the ships draft when the whale was directly migratory depth to save energy, then many whales may below the ship [53], so even when only considering con- spend a much larger proportion of their time within the tainer ships with moderate drafts (8 m), use of the upper danger zone for ship strikes than currently thought. 24 m of the water column for extended periods is likely to The consistency of the shallow migratory depths also have increased the risk of ship strike for this whale. The has implications for programming of satellite-linked migratory paths of pygmy blue whales along the Western dive depth-transmitting tags placed onto whales. These Australian coast largely overlap with the main shipping devices require that a dive be defined as submergence routes through the area [34] (Fig.  2b). When respond- below some minimum depth to limit the amount of ing to oncoming ships, blue whales have been shown to dive information necessary to compress for transmis- complete a response dive with a slow decent and a lack sion through the very limited Argos bandwidth. In many of movement away from the ship [54]. While in shallow cases, the definition of a dive is set to be potentially sections of the water column, the Lloyd mirror effect deeper than the predicted depth for efficient travel. For can dramatically influence the ability of a whale to deter - example, if predictions hold for each species, Antarctic mine both the direction and distance of a ship [55, 56]. minke whales (Balaenoptera bonaerensis) need to swim Sea Floor Depth (m) Residence time (hours) −1000−400 0 0 10 20 30 Owen et al. Anim Biotelemetry (2016) 4:17 Page 9 of 11 just below 4  m to avoid wave drag (Table  1). Indeed as observed both occurred over deep water areas, with an predicted, inspection of dives referred to as ‘shallow increase in exploratory diving behavior suggesting poten- non-feeding dives’ by Antarctic minke whales in another tial prey searching. Therefore, deep canyons around the study appeared to occur at approximately 5 m depth [16], southwest coast of Australia may provide suitable feeding shallower than the depth threshold set for defining dives habitat for blue whales. However, depending on the timing in many studies [37, 60–62]. If other whales also optimize of local conditions with the arrival of a whale, some areas their dive depths and this is not taken into account when may have varying success as a feeding area year to year. setting dive definition thresholds, then it is possible that Lunge feeding is an energetically expensive strategy with migratory dives may not be recorded by these devices prey density needing to be above a certain level to induce and a biased picture of the depth ranges used by species feeding behavior by a whale [69, 70]. If an animal enters an could be produced. area where they have reliably fed previously, then they may Typically, studies obtaining horizontal movement data initiate ARS behavior to assist with searching for prey, and lack information on the actual behavioral state of the ani- spend longer in the area, yet not lunge feed if dense prey mal and infer it from track metrics [27, 63–65]. Although is not located successfully [71, 72]. Therefore, while hori - the one whale in this study did not lunge feed to a great zontal movement data were not able to successfully locate extent, it provided an opportunity to address the likeli- areas where lunge feeding occurred, they may still provide hood that the location of lunge feeding behavior by the information on potential foraging sites of whales. whale matched areas identified as ARS, and potential foraging locations inferred from horizontal movement Conclusions data. For this whale, the behavioral states inferred from The migratory dive depth of a pygmy blue whale was FastLoc GPS data did not accurately match the locations found to be highly consistent over the week of obser- where lunge feeding occurred. The method used to infer vation, with the whale appearing to optimize its travel behavior was based on an assumption that areas of low depth to reduce the impact of wave drag near the surface, speed and high turning angles result in higher residence yet remain close to the surface where it must return regu- time which is likely to be indicative of foraging behav- larly to breathe. While this strategy may reduce energy ior; a common assumption in many behavioral models expenditure during migration, it also placed the whale at applied to both FastLoc GPS and Argos data [26–29]. greater risk of ship strike for a much longer period than However, on a fine-scale whales have been shown to currently thought. In addition, periods of ARS identi- feed while still maintaining relatively straight trajectories fied using horizontal movement data did not accurately with low turning angles [33]. Other studies have looked reflect where lunge feeding behavior occurred for this at how well behavior inferred from horizontal movement animal, but may still highlight some foraging areas. This data overlaps with presumed feeding in other taxa with suggests that the core assumptions of movement models mixed conclusions. In southern elephant seals (Mirounga should be further tested in relation to fine-scale acceler - leonina), increases in body condition, and likely foraging ometer data, as the behavioral states obtained may not success, have been used to show that the foraging behav- accurately reflect the feeding behavior of whales. Further ior detected in horizontal movement data is a good proxy development of methods for recognizing actual lunge for foraging success [66]. However, in migratory tuna feeding events by the tag software and subsequent trans- it was found that periods of high residence can also be mission of that information over longer time periods will associated with periods of fasting [67], as was observed in also assist with the identification of whale feeding areas. this study. It is possible that for species that spend their entire lives in the marine environment, periods of resi- Abbreviations dency are also likely to be associated with periods of rest- ARS: area‑restricted search; LIMPET: Low Impact Minimally Percutaneous ing, behavior that is often completed on land for species External‑ electronics Transmitter; MSA: minimum specific acceleration; SST: sea surface temperature. that can haul out. More research is needed in this area to determine whether the behavioral states identified from Authors’ contributions horizontal movement data are reflective of the location of KO developed ideas, completed data analyses, wrote the manuscript, RDA developed ideas, collected data, assisted with manuscript preparation, MNJ feeding by baleen whales. collected data, assisted with manuscript preparation, and CSJ collected data, However, when examining the periods of ARS behavior assisted with manuscript preparation. All authors read and approved the final by the tagged pygmy blue whale it is possible to specu- manuscript. late that the animal was indeed searching for prey during Author details these times when no lunges were observed. Off Western 1 2 Alaska SeaLife Center, Seward, AK 99664, USA. Centre for Whale Research Australia, pygmy blue whales are known to feed along the ( WA Inc.), Fremantle, WA 6959, Australia. School of Fisheries and Ocean Sci‑ ences, University of Alaska Fairbanks, Fairbanks, AK 99775, USA. edges of deep water canyons [68]. The two periods of ARS Owen et al. Anim Biotelemetry (2016) 4:17 Page 10 of 11 Acknowledgements 16. Friedlaender AS, Goldbogen JA, Nowacek DP, Read AJ, Johnston D, Gales The authors would like to thank the crew of the Whale Song and especially N. Feeding rates and under ice foraging strategies of smallest lunge filter Simon Kenion for retrieving the tag. This research was completed under Com‑ feeder, the Antarctic minke whale (Balaenoptera bonaerensis). J Exp Biol. monwealth permit number 2013‑00012. 2014;217:2851–4. 17. Fish FE. Aquatic locmotion. In: Tomasi TE, Horton TH, editors. Mammalian Competing interests energetics: interdisciplinary views of metabolism and reproduction. The authors declare that they have no competing interests. Ithaca: Cornell University Press; 1992. p. 34–63. 18. Williams TM. Swimming by sea otters: adaptations for low energetic cost Availability of data and material locomotion. J Comp Physiol A. 1989;164(81):5–24. The dataset supporting the conclusions of this article is available in the 19. Goldbogen JA, Calambokidis J, Croll DA, Harvey JT, Newton KM, Oleson Movebank repository [A week in the life of a pygmy blue whale, https://www. EM, Schorr G, Shadwick RE. Foraging behavior of humpback whales: movebank.org/panel_embedded_movebank_webapp]. kinematic and respiratory patterns suggest a high cost for a lunge. J Exp Biol. 2008;211:3712–9. Ethics approval 20. Tershy B, Acevedo‑ Gutiérrez A, Breese D, Strong C. Diet and feeding Ethical approval for this study was provided by the Western Australian Depart‑ behavior of fin and Bryde’s whales in the Central Gulf of California, ment of Parks and Wildlife (number 2013/45). México. Rev Invest Cient. 1993;1(1):31–7. 21. Doniol‑ Valcroze T, Lesage V, Giard J, Michaud R. Optimal foraging theory Funding predicts diving and feeding strategies of the largest marine predator. Funding for this research was provided by the Centre for Whale Research Behav Ecol. 2011;22:880–8. ( Western Australia). The funding body assisted with data collection and 22. Heide‑ Jørgensen MP, Laidre KL, Jensen MV, Dueck L, Postma LD. Dissolv‑ manuscript preparation. The Whale Lander tag was developed by Wildlife ing stock discreteness with satellite tracking: bowhead whales in Baffin Computers and RDA with funding from the Office of Naval Research and the Bay. Mar Mam Sci. 2006;22(1):34–45. DOD Small Business Technology Transfer (STTR) program. 23. Zerbini AN, Andriolo A, Heide‑ Jørgensen MP, Pizzorno JL, Maia YG, Van‑ Blaricom GR, DeMaster DP, Simões‑Lopes PC, Moreira S, Bethlam C. Satel‑ Received: 2 April 2016 Accepted: 1 August 2016 lite‑monitored movements of humpback whales Megaptera novaeangliae in the Southwest Atlantic Ocean. Mar Ecol Prog Ser. 2006;313:295–304. 24. Garrigue C, Zerbini AN, Geyer Y, Heide‑ Jørgensen MP, Hanaoka W, Clapham P. Movements of satellite‑monitored humpback whales from New Caledonia. J Mamm. 2010;91(1):109–15. 25. Mate BR, Best PB, Lagerquist BA, Winsor MH. Coastal, offshore, and References migratory movements of South African right whales revealed by satellite 1. Alerstam T, Hedenstrom A, Akesson S. Long‑ distance migration: evolution telemetry. Mar Mam Sci. 2011;27(3):455–76. and determinants. Oikos. 2003;103:247–60. 26. Jonsen ID, Mills Flemming J, Myers RA. Robust state space modeling of 2. Rassmussen K, Palacios DM, Calambokidis J, Saborio MT, Dalla Rosa animal movement data. Ecology. 2005;86:2874–80. L, Secchi ER, Steiger GH, Allen JM, Stone GS. Southern Hemisphere 27. Bailey H, Mate BR, Palacios DM, Irvine L, Bograd SJ, Costa DP. Behavioural humpback whales wintering off Central America: insights from estimation of blue whale movements in the Northeast Pacific from water temperature into the longest mammalian migration. Biol Lett. state‑space model analysis of satellite tracks. Endanger Species Res. 2007;3:205–302. 2009;10:93–106. 3. Robbins J, Rosa LD, Allen J, Mattila D, Secchi ER, Friedlaender AS, Stevick 28. Patterson TA, Thomas L, Wilcox C, Ovaskainen O, Matthiopoulos J. Statespace ‑ P, Nowacek D, Steel D. Return movement of a humpback whale between models of individual animal movement. Trends Ecol Evol. 2008;23:87–94. the Antarctic Peninsula and American Samoa: a seasonal migration 29. Patterson TA, Basson M, Bravington MV, Gunn JS. Classifying movement record. Endang Species Res. 2011;13:117–21. behaviour in relation to environmental conditions using hidden Markov 4. Mate BR, Ilyashenko VY, Bradford AL, Vertyankin VV, Tsidulko GA, Rozhnov models. J Anim Ecol. 2009;78:1113–23. VV, Irvine LM. Critically endangered western gray whales migrate to the 30. Best PB, Sekiguchi K, Findlay KP. A suspended migration of humpback eastern North Pacific. Biol Lett. 2015;11:20150071. whales Megaptera novaeangliae on the west coast of South Africa. Mar 5. Fish FE. Transitions from drag‑based to lift ‑based propulsion in mamma‑ Ecol Prog Ser. 1995;118:1–12. lian swimming. Am Zool. 1996;36:628–41. 31. Stockin KA, Burgess EA. Opportunistic feeding of an adult male hump‑ 6. Williams TM, Davis RW, Fuiman LA, Francis J, Le Boeuf BJ, Horning M, back whale (Megaptera novaeangliae) migrating along the coast of south Calambokidis J, Croll DA. Sink or swim: strategies for cost‑ efficient diving eastern Queensland, Australia. Aquat Mamm. 2005;31(1):120–3. by marine mammals. Science. 2000;288:133–6. 32. Visser F, Hartman KL, Pierce GJ, Valavanis VD, Huisman J. Timing of migra‑ 7. Goldman JA. Eec ff ts of the free water surface on animals that jump out of tory baleen whales at the Azores in relation to the North Atlantic spring the water. Ph.D. thesis, Duke University, Durham, NC; 2001 p. 256 bloom. Mar Ecol Prog Ser. 2011;440:267–79. 8. Vennell R, Pease D, Wilson B. Wave drag on human swimmers. J Biomech. 33. Owen K, Warren JD, Noad MJ, Donnelly D, Goldizen AW, Dunlop RA. 2006;39:664–71. Eec ff t of prey type on the fine ‑scale feeding behaviour of migrating east 9. Blake RW. Biological implications of the hydrodynamics of swimming at Australian humpback whales. Mar Ecol Prog Ser. 2015;541:231–44. or near the surface and in shallow water. Bioinspir Biomim. 2009;4:1–9. 34. Double MC, Andrews‑ Goff V, Jenner KCS, Jenner MN, Laverick SM, Branch 10. Hertel H. Structure, form, movement. New York: Reinhold Publishing Co; TA, Gales NJ. Migratory movements of pygmy blue whales (Balaenoptera musculus brevicauda) between Australia and Indonesia as revealed by 11. Gleiss AC, Norman B, Wilson RP. Moved by that sinking feeling: variable satellite telemetry. PLoS One. 2014;9(4):e93578. diving geometry underlies movement strategies in whale sharks. Func 35. Johnson MP, Tyack PL. A digital acoustic recording tag for measuring Ecol. 2011;25(3):595–607. the response of wild marine mammals to sound. IEEE J Ocean Eng. 12. Davis RW, Weihs D. Locomotion in diving elephant seals: physi‑ 2003;28(1):3–12. cal and physiological constraints. Philos Trans R Soc Lond B Biol Sci. 36. Andrews RD, Pitman RL, Ballance LT. Satellite tracking reveals distinct 2007;362(1487):2141–50. movement patterns for Type B and Type C killer whales in the southern 13. Nousek‑McGregor AE. The cost of locomotion in North Atlantic right Ross Sea, Antarctica. Polar Biol. 2008;31:1461–8. whales, Eubalaena glacialis. PhD thesis. Duke University; 2010. 37. Simon M, Johnson M, Madsen PT. Keeping momentum with a mouthful 14. Tyack PL, Johnson M, Aguilar Soto N, Sturlese A, Madsen PT. Extreme div‑ of water: behavior and kinematics of humpback whale lunge feeding. J ing of beaked whales. J Exp Biol. 2006;209:4238–53. Exp Biol. 2012;215:3786–98. 15. Watwood S, Miller PJO, Johnson M, Madsen PT, Tyack P. Deep diving for‑ 38. Goldbogen JA, Calambokidis J, Shadwick R, Oleson EM, McDonald MA, aging behaviour of sperm whales (Physeter macrocephalus). J Anim Ecol. Hildebrand JA. Kinematics of foraging dives and lunge‑feeding in fin 2006;75:814–25. whales. J Exp Biol. 2006;209:1231–44. Owen et al. Anim Biotelemetry (2016) 4:17 Page 11 of 11 39. Owen K, Dunlop RA, Monty JP, Chung D, Noad MJ, Donnelly D, Goldizen 58. Laist DW, Knowlton AR, Mead JG, Collet AS, Podesta M. Collisions AW, Mackenzie T. Detecting surface‑feeding behavior by rorqual whales between ships and whales. Mar Mamm Sci. 2001;17(1):35–75. in accelerometer data. Mar Mamm Sci. 2016;32(1):327–48. 59. Conservation Management Plan for the Blue Whale—A Recovery Plan 40. Luque SP. Dive analysis and calibration. R Package version 1.3.9. CRAN. under the Environment Protection and Biodiversity Conservation Act 2015. R‑project.org/package = diveMove; 2015. 1999, Commonwealth of Australia 2015. 41. R Development Core Team. R: A language and environment for statistical 60. Croll DA, Acevedo‑ Gutiérrez A, Tershy BR, Urban‑Ramırez J. The diving computing, reference index version 3.1.3. R Foundation for Statistical behaviour of blue and fin whales: Is dive duration shorter than expected Computing, Vienna, Austria. 2015. URL http://www.Rproject.org. based on oxygen stores? Compar Biochem Physiol A. 2001;129:797–809. 42. Witt MJ, Åkesson S, Broderick AC, Coyne MS, Ellick J, Formia A, Hays 61. Heide‑ Jørgensen MP, Laidre KL, Wiig Ø, Jensen MV, Dueck L, Maiers GC, Luschi P, Stroud S, Godley BJ. Assessing accuracy and utility of LD, Schmidt HC, Hobbs RC. From Greenland to Canada in ten days: satellite‑tracking data using Argos‑linked Fastloc‑ GPS. Anim Behav. tracks of bowhead whales, Balaena mysticetus, across Baffin Bay. Arctic. 2010;80:571–81. 2003;56(1):21–31. 43. Barraquand F, Benhamou S. Animal movements in heterogeneous 62. Nielsen NH, Laidre K, Larsen RS, Heide‑ Jørgensen MP. identification of landscapes: identifying profitable places and homogeneous movement potential foraging areas for bowhead whales in Baffin Bay and adjacent bouts. Ecology. 2008;89(12):3336–48. waters. Arctic. 2015;68(2):169–79. 44. Calenge C. Analysis of Animal Movements in R: the adehabitatLT Package. 63. Forester JD, Ives AR, Turner MG, Andersen DP, Fortin D, Beyer HL, Smith 2015. R Package version n 0.3.20. CRAN. R‑project.org/package = ade ‑ DW, Boyce MS. State‑space models link elk movement patterns to habitatLT. 2015. landscape characteristics in Yellowstone National Park. Ecol Monograph. 45. Fauchald P, Tveraa T. Using first ‑passage time in the analysis of area‑ 2007;77(2):285–99. restricted search and habitat selection. Ecology. 2003;84(2):282–8. 64. Heithaus MR, Frid A, Wirsing AJ, Dill LM, Fourqurean JW, Burkholder D, 46. Lavielle M. Detection of multiple changes in a sequence of dependent Thomson J, Bejder L. State‑ dependent risk‑taking by green sea turtles variables. Stoch Proc Appl. 1999;83:79–102. mediates top‑ down effects of tiger shark intimidation in a marine ecosys‑ 47. Lavielle M. Using penalized contrasts for the change‑point problem. tem. J Anim Ecol. 2007;76(5):837–44. Signal Process. 2005;85:1501–10. 65. Jonsen ID, Myers RA, James MC. Identifying leatherback turtle foraging 48. Dodge S, Bohrer G, Weinzierl R, Davidson SC, Kays R, Douglas D, Cruz S, behaviour from satellite telemetry using a switching state‑space model. Han J, Brandes D, Wilkelski M. The environmental‑ data automated track Mar Ecol Prog Ser. 2007;337:255–64. annotation (Env‑DATA) system: linking animal tracks with environmental 66. Dragon A, Bar‑Hen A, Monestiez P, Guinet C. Comparative analysis of data. Mov Ecol. 2013;1:3. methods for inferring successful foraging areas from Argos and GPS 49. Amante C, Eakins BW. ETOPO1 1 Arc‑minute global relief model: proce ‑ tracking data. Mar Ecol Prog Ser. 2012;452:253–67. dures, data sources and analysis. NOAA Technical Memorandum NESDIS 67. Bestley S, Patterson TA, Hindell MA, Gunn JS. Feeding ecology of wild NGDC‑24. National Geophysical Data Center, NOAA. 2009. doi:10.7289/ migratory tunas revealed by archival tag records of visceral warming. J V5C8276M. Accessed 1 Oct 2015. Anim Ecol. 2008;77:1223–33. 50. NASA Ocean Biology (OB.DAAC). Sea‑ viewing wide field‑ of‑ view sensor 68. Rennie S, Hanson CE, McCauley RD, Pattiaratchi C, Burton C, Bannister J, (SeaWiFS) ocean color data, 2014.0 Reprocessing. NASA OB.DAAC, Green‑ Jenner C, Jenner MN. Physical properties and processes in the Perth Can‑ belt, MD, USA. doi:10.5067/ORBVIEW‑2/SEAWIFS_OC.2014.0. Accessed 01 yon, Western Australia: Links to water column production and seasonal Dec 2015. pygmy blue whale abundance. J Mar Syst. 2009;77:21–44. 51. Davis RW, Fuiman LA, Williams TM, Le Boeuf BJ. Three‑ dimensional 69. Piatt JF, Methven DA. Threshold foraging behavior of baleen whales. Mar movements and swimming activity of a northern elephant seal. Comp Ecol Prog Ser. 1992;84:205–10. Biochem Physiol A. 2001;129:759–70. 70. Hazen EL, Friedlaender AS, Goldbogen JA. Blue whales (Balaenoptera 52. Sato K, Mitani Y, Cameron MF, Siniff DB, Naito Y. Factors affecting stroking musculus) optimize foraging efficiency by balancing oxygen use and patterns and body angle in diving Weddell seals under natural condi‑ energy gain as a function of prey density. Sci Adv. 2015;1(9):e1500469. tions. J Exp Biol. 2003;206:1461–70. 71. Weimerskirch H. Are seabirds foraging for unpredictable resources? Deep 53. Silber GK, Slutsky J, Bettridge S. Hydrodynamics of a whale/ship collision. Sea Res 2. 2007;54:211–23. J Exp Mar Biol Ecol. 2010;391:10–9. 72. Thums T, Bradshaw CJA, Hindell MA. In situ measures of foraging success 54. McKenna MF, Calambokidis J, Oleson EM, Laist DW, Goldbogen JA. and prey encounter reveal marine habitat‑ dependent search strategies. Simultaneous tracking of blue whales and large ships demonstrates Ecology. 2011;92(6):1258–70. limited behavioral responses for avoiding collision. Endanger Species Res. 73. Konishi K, Tamura T, Zenitani R, Bando T, Kato H, Walløe L. Decline in 2015;27:219–32. energy storage in the Antarctic minke whale (Balaenoptera bonaerensis) 55. Urick RJ. Principles of underwater sound. New York: McGraw‑Hill; 1983. in the Southern Ocean. Polar Biol. 2008;31:1509–20. 56. Terhune JM, Verboom WC. Right whales and ship noise. Mar Mamm Sci. 74. Woodward BL, Winn JP, Fish FE. Morphological specializations of baleen 1999;15:256–8. whales associated with hydrodynamic performance and ecological niche. 57. Irvine LM, Mate BR, Winsor MH, Palacios DM, Bograd SJ, Costa DP, Bailey J Morphol. 2006;267:1284–94. H. Spatial and temporal occurrence of blue whales off the US West Coast, with implications for management. PLoS One. 2014;9(7):e102959. Submit your next manuscript to BioMed Central and we will help you at every step: • We accept pre-submission inquiries • Our selector tool helps you to find the most relevant journal • We provide round the clock customer support • Convenient online submission • Thorough peer review • Inclusion in PubMed and all major indexing services • Maximum visibility for your research Submit your manuscript at www.biomedcentral.com/submit

Journal

Animal BiotelemetrySpringer Journals

Published: Aug 12, 2016

References