Activity budget and diving behavior of gray whales (Eschrichtius robustus) in feeding grounds off coastal British Columbia


  • Lei Lani Stelle,

    1. Department of Biology, University of Redlands, Redlands, CA 92373, U.S.A. and Coastal Ecosystems Research Foundation, The White House, Dawson's Landing, British Columbia, V0N 1M0, Canada E-mail:
    Search for more papers by this author
  • William M. Megill,

    1. Coastal Ecosystems Research Foundation, The White House, Dawson's Landing, British Columbia, V0N 1M0, Canada and Centre for Biomimetic and Natural Technologies, Mechanical Engineering Department, University of Bath, Claverton Down, Bath BA2 7AY, United Kingdom
    Search for more papers by this author
  • Michelle R. Kinzel

    1. Coastal Ecosystems Research Foundation, The White House, Dawson's Landing, British Columbia, V0N 1M0, Canada
    Search for more papers by this author


Behavior and diving patterns of summer resident gray whales (Eschrichtius robustus) foraging on mysids were studied in coastal bays along the north shore of Queen Charlotte Strait, British Columbia. In this region, gray whales were found to feed primarily on planktonic prey rather than on the benthos as in their primary feeding areas further north. During the summers of 1999 and 2000, whales spent most of their time actively feeding or searching for prey (77%), whereas only 15% of their time was spent traveling and 8% socializing. The majority of the dives were short; the mean dive duration was 2.24 min with approximately three respirations per surfacing and 15 s between blows. Whales dove frequently (26.7 h−1), spending only 17% of their time at the surface with an overall blow rate of 1.14 respirations per minute. Activity states were characterized by significantly different diving and respiratory parameters; feeding whales dove more frequently, with shorter intervals between respirations, thus spending less time at the surface compared to when traveling or searching. This diving pattern differs from benthic-feeding whales and likely optimizes capture of the mobile mysid swarms in shallow waters.

Eastern Pacific gray whales (Eschrichtius robustus) migrate annually between summering grounds in the Bering and Chukchi Seas and overwintering areas around Baja California (Rice and Wolman 1971). However, some gray whales do not complete this path, and in recent decades there has been an increase in the number of records of animals summering south of the Bering Sea (Nerini 1984, Darling et al. 1998, Calambokidis et al. 2002); it is unknown whether this reflects a shift in distribution or just increased research effort in these areas. Gray whales migrate along the outer coast of Vancouver Island, British Columbia, and as early as 1958 it was noted that some animals remained in the area as summer residents (Pike 1962). Vancouver Island is located approximately midway between the breeding lagoons of Baja California and the most northern summering grounds. Terminating the migration early reduces the travel time by 2 mo (Darling 1984) and extends the feeding time, which could provide an energetic advantage (Sumich 1984).

Gray whales are considered opportunistic foragers (Nerini 1984). The majority of the population bottom-feeds on an extensive infaunal amphipod community in the northern waters (Rice and Wolman 1971). Recently however, there has been a documented decrease in amphipod productivity and decline in gray whale sightings (Moore et al. 2003) associated with a shift from a benthic to a pelagic-dominated ecosystem in the Northern Bering Sea (Grebmeier et al. 2006). Thus, the Eastern Pacific gray whale population may be nearing the carrying capacity of the Bering and Chukchi Seas (Highsmith and Coyle 1992), which could result in increased use of southern feeding grounds (Le Boeuf et al. 2000). Much of what is known regarding gray whale foraging is based on studies from the benthic-feeding animals (Nerini 1984, Oliver et al. 1984, Würsig et al. 1986) but their behavior and dive patterns are influenced by prey and associated foraging techniques. Swarms of mysids in the water column appear to be a common prey in southern feeding areas (Murison et al. 1984, Kim and Oliver 1989, Dunham and Duffus 2002), and when feeding on planktonic prey, gray whales use surface skimming or engulf prey from the water column (Nerini 1984).

Activity budgets provide useful information on the relative importance of various behaviors, and these can also be combined with estimates of activity specific metabolic rates to investigate energetics (Costa and Williams 1999). It is difficult to make many generalizations about the behavior of gray whales because few studies report activity budgets and those that do often rely on relatively small sample sizes or incomplete behavioral categories. Studies utilizing aerial observations provide large sample sizes (e.g., Moore et al. 1986, Clarke et al. 1989) but their values, which are based on activities of groups, are not equivalent to traditional activity budgets based on the behavior of individuals. Aerial observations present a snapshot of what each individual is doing at any one time rather than demonstrating how a single individual's activities vary over a longer time span. In addition to detailed activity budgets, focal animal studies can provide information on respiratory and diving patterns. These data can be used to correct sighting records (Okamura et al. 2006), assess physiological influences, and serve as an indirect measure of metabolic rates (Sumich 1983). Relationships between respiratory characteristics and activity state can also help to identify a whale's activity when this is not clear from surface behaviors.

Our study focused on a seasonal feeding aggregation of gray whales just north of Vancouver Island along the mainland coast of British Columbia (Calambokidis et al. 2002). In this area the whales forage in coastal bays, feeding in the abundant kelp beds on mysids (family Mysidae) that form dense, large swarms which often extend from the bottom to near the surface (Stelle 2001). Whales were easily observed from both shore stations and small boats, and since they were in shallow waters, their activity was obvious from surface behaviors. This remote study area is an excellent location to provide baseline information on activity budgets and diving patterns of gray whales as it receives relatively little human impact because there are no whale watch vessels and few recreational boaters.

The ecosystem shift occurring in the Bering Sea (Grebmeier et al. 2006) has impacted gray whale distribution and foraging range (Moore et al. 2003), thus it is essential that we have a solid understanding of how their behavior varies with habitat and prey type. The objectives of this study were to (1) compile an activity budget for gray whales in this southern feeding area, (2) describe their diving behavior, and (3) determine if there were any significant relationships between the whales' diving and respiratory patterns with their associated activity.


During the summer months of 1999 and 2000, gray whales were observed throughout a 16-km stretch along the mainland coast of the southern central coast of British Columbia, Canada (51°102′N, 127°134′W to 51°10′N, 126°146′W). The majority of the data was collected in three main study areas: North, Burnett, and Silvester bays, with occasional records from the connecting areas (Fig. 1). The bays and surrounding coastal waters ranged in depth from 2 to 68 m, but were <20 m deep on average. Observation platforms included shore stations in each bay and several motor boats (6–13 m in length). Binoculars were used when necessary, but most observations could be done without visual aids. Observations were made from 1 July to 8 September 1999 and from 26 June to 22 August 2000. Each bay was monitored from 1 to 7 d per week, for 1–11 h per day during the daylight hours (0700–2200) by a minimum of two observers, for a total of over 580 h. All whale sightings were recorded and focal and group data were collected for as many individuals as possible.

Figure 1.

Study area map; the inset shows location relative to Vancouver Island and mainland British Columbia.

Focal observations were conducted when only one whale was in view. Individuals were recognized using natural markings (Darling 1984, Calambokidis et al. 2002), which made it possible to confirm that the same animal was followed throughout an observation period. Photographs were not taken during behavioral observations so as to avoid disturbing their natural behavior. Focal observations were terminated when the whale left the field of view or another whale entered the same bay. When multiple whales were present, group data were collected, which included recording the behaviors of each of the whales, but respiration rates were not recorded due to possible confusion in determining the identity of individuals. Movement of whales and time spent in each location were also tracked on charts of each bay during both focal and group observations.

Recorded behaviors were used to determine activity states during the observation period. Common behaviors included: fluke—tail raised completely at start of dive, arch—peduncle raised but tail not shown, flipjack—whale rolling to one side as it dives, and tailswish—part or all of the tail breaks the water's surface without concurrent surfacing of the entire animal (i.e., no respiration possible). Behaviors and movement patterns indicated the overall activity of the animal: traveling, resting, socializing, searching, or feeding. These activity states were defined as follows: traveling—directed, relatively straight-line movement at a steady speed without stopping, resting—stationary animal quiescent near the surface, socializing—two or more individuals within approximately two body lengths of one another and apparently interacting, searching—examination of potential foraging areas with no clear feeding events, and feeding—apparent feeding in the water column. Searching was characterized by whales' moving between kelp beds and stopping for short investigations before continuing to another potential site; these behaviors often preceded and followed feeding events. Active feeding was inferred from surface behaviors (e.g., Murison et al. 1984): repeated dives in the same kelp bed with frequent changes of direction at the surface or underwater to remain in the same location; twisting maneuvers resulting in only half of the fluke being visible during the dive (flipjack); tail breaking the surface of the water while the animal was submerged, usually in a circular pattern (tailswish) indicating that the whale was rotating underwater with its head near the substrate; pectoral flippers visible above the water as the whale rolled on its side; and bubbles (function unknown). Feeding was identified based on behaviors as described above, along with the amount of time spent in one location; for example, if a whale made only a single dive in a kelp bed it was classified as searching. Measures of respiratory rates or dive duration (DT) were not used to categorize activity. Although feeding could not be directly observed, extensive scuba observations revealed that mysids were the only abundant prey in the study area (Stelle 2001). That gray whales were actively feeding on the mysids was confirmed by the presence of exoskeletons in fecal samples collected from whales characterized as “feeding” based on behavior.

To measure dive parameters, the exact time of every exhalation (i.e., blow) by focal whales was recorded to the second. Whales often respired multiple times during a single surfacing event with brief submergences between blows. These submergences were not classified as a “dive” because there was no potential for activity (e.g., feeding or searching) since the animals remained near the surface; instead they are termed blow intervals. Sumich (1983) recognized these short breath-holds as a method to achieve efficient rates of oxygen utilization. Malcolm and Duffus (2000), using data collected with time-depth recorders, also ascribed the function of oxygen recharge to the short shallow dives they recorded and termed “interventilation.” Submergences were classified as dives when the tail was displayed (e.g., fluke or flipjack), appendages (e.g., tail or pectoral flipper) emerged while the whale was underwater or directly observing that the whale was not near the surface. We did not use a predetermined time of submergence to identify dives because some true dives were observed to be shorter in duration than some blow intervals.

Focal records were analyzed quantitatively to provide data on respiration, diving patterns, and activity budgets. DT, number of blows per surfacing period (NB), duration of time at the surface between dives (ST), and the interval between blows (BI) were measured from all recorded dive cycles during focal observations, and the mean values calculated for each observation period. We also calculated the mean blow rate (BR = NB/[DT + ST]) and percent of the ventilatory cycle at the surface (PCST = ST/[ST + DT]× 100) following Dolphin (1987a) along with the dive rate (DR = number of dives/sum[ST + DT], expressed per hour). Activity budgets were compiled by calculating the proportion (expressed as a percentage) of the observation time a whale spent in each activity (including time at depth and at the surface) during each record, based on both focal and group data.

For statistical purposes each observation period was considered a sample rather than each respiration/dive cycle because of the lack of independence due to the physiological effect of previous dives on subsequent dives and respirations. Ideally, each individual animal should be treated as a sampling unit but we were unable to apply that stringent standard in this analysis because we could not confirm whale identities between observations. A concurrent photo-ID study revealed that at least 20 individual whales in 1999 and 24 whales in 2000 were resident in our study area, with 10 of the same animals found in both summers (unpublished data).

All statistical analyses were performed with Minitab 14 (Minitab Inc., State College, PA). Although the variables were not normally distributed, variances were equal and the sample size (n= 342) is so large that parametric analyses are appropriate according to the central limit theorem. To examine activity budgets and diving patterns, we ran general linear models with year, bay, and activity as factors. For the dive analysis, only feeding, searching, and traveling were examined; socializing and resting were not included because of the small sample size associated with these less common activities. Post hoc comparisons of significant effects were examined using the Bonferroni simultaneous tests. For all statistical tests, null hypotheses were rejected at P≤ 0.05. All values are reported as the grand mean of the observation means (±SD), but when range is included it refers to the absolute values observed rather than the range of the means.


Data were recorded for 86 whale sightings on 32 d in 1999 and 315 whale sightings on 49 d in 2000. Each whale was observed for an average of 44 min (range of 5 min to 3.68 h) in 1999 and 60 min (range of 6 min to 5.58 h) in 2000, for a total of over 388 whale hours. Since whale identities were unconfirmed, these whale numbers do not represent individual animals but instead reflect distinct sightings or data records.

Activity Budgets

The whales spent most of their time engaged in foraging activities. There was no significant effect of bay or year so with all values combined, whales spent an average of 39% of their time actively feeding, 38% searching, 15% traveling, 8% socializing, and only 0.2% resting. There was no difference between the time spent feeding and searching but they spent significantly more time in those foraging activities than traveling (T=−11.96, P= 0.0000; T=−11.65, P= 0.000; vs. feeding and searching, respectively), socializing (T= 15.47, P= 0.0000; T= 15.16, P= 0.0000), and resting (T=−19.11, P= 0.0000; T= 18.80, P= 0.0000). Whales also spent significantly more time traveling than socializing (T= 3.51, P= 0.0046) and resting (T= 7.15, P= 0.0000) and more time socializing than resting (T=−3.64, P= 0.0028).

Respiration and Diving Characteristics

Whales exhibited the traditionally described dive pattern of one dive followed by a number of blows at the surface separated by shallow submergences (Sumich 1983), or blow intervals. The majority of dives were short in duration; mean DT was 2.24 ± 0.86 min (Fig. 2A) with an overall range of 8 s to 11.0 min. Surfacings were also brief with a mean duration of 24.1 ± 15.7 s in 1999 and 30.9 ± 19.7 s in 2000 (Fig. 2B, F= 5.68, P= 0.018). During a surfacing event, whales would respire 1 to 14 times with a mean of 2.47 ± 1.16 in 1999 and 2.95 ± 1.23 in 2000 (Fig. 2C, F= 8.47, P= 0.004) and the mean interval between blows was 15.4 ± 4.73 s (Fig. 2D). Blow rate was significantly higher in 2000 (1.17 ± 0.317 min−1) than 1999 (0.998 ± 0.269 min−1; F= 12.81, P= 0.000) and was the only variable that showed a significant effect of bay (F= 4.65, P= 0.01), being lower in Burnett Bay (1.01 ± 0.281 min−1) than North Bay (1.18 ± 0.314 min−1; T= 3.33, P= 0.0028) but no differences with Silvester Bay (1.12 ± 0.324 min−1). DR was not influenced by year or bay and averaged 26.7 ± 11.39 dives per hour. The overall percentage of time spent at the surface was 14.2%± 7.34% in 1999 and 17.5%± 7.63% in 2000 (F= 6.92, P= 0.009).

Figure 2.

Histograms of mean diving and respiratory parameters recorded during focal observations: (A) dive duration, DT; (B) surface time, ST; (C) number of blows, NB; and (D) blow interval, BI.

Relationships to Activity

Feeding whales displayed significantly different diving and respiratory patterns compared to their other activities and the same trends were found in both years (Fig. 3). Feeding whales made briefer surfacings (1999: 19.0 ± 12.9 s, 2000: 27.2 ± 16.1 s) than did traveling (1999: 30.6 ± 17.4 s, 2000: 35.9 ± 29.0 s) or searching whales (1999: 28.1 ± 16.9 s, 2000: 32.0 ± 15.6 s). The interval between successive blows was shortest during feeding (13.8 ± 2.65 s), intermediate when searching (15.7 ± 4.93 s), and longest when traveling (18.3 ± 6.29 s). Feeding whales spent a lower percent of their time at the surface (1999: 12.3 ± 4.86%, 2000: 16.1 ± 6.25%) than did traveling (1999: 18.2 ± 8.62%, 2000: 19.7 ± 10.26%) or searching whales (1999: 15.17 ± 8.80%, 2000: 17.8 ± 6.91%). Although there was no difference in mean DT with activity, feeding whales made more dives per hour (30.0 ± 12.81 h−1), than when traveling (23.9 ± 9.02 h−1) or searching (23.0± 8.27 h−1). Neither the number of blows per surfacing nor the mean blow rates were influenced by activity.

Figure 3.

Relationship between activity (F-feeding, S-searching, T-traveling) and diving/respiratory parameters; variables that were significantly different between years are grouped by year and activity. Boxplot includes the 25th to 75th percentile, whiskers bound the values 1.5× the interquartile range, and outliers are indicated by solid circles; the median (solid line) and mean (solid square) values are shown within the boxes. Asterisks (***) indicate significant effects of activity at P < 0.05. (A) Blow intervals (BI) were shortest when feeding (T= 3.49, P= 0.0016), intermediate when searching (T= 3.98, P= 0.0003) and longest when traveling (T= 6.82, P= 0.0000). (B) Surface time (ST) was shortest when feeding compared to searching (T= 2.50, P= 0. 0382) or traveling (T= 3.28, P= 0. 0034) with no difference between searching and traveling. (C) Dive time (DT) did not differ significantly with activity (F= 1.46, P= 0.233). (D) Percent of ventilatory cycle at the surface (PCST) was significantly shorter when feeding compared to traveling (F= 3.60, P= 0.0011) but no differences with searching. (E) Dive rate (DR) was highest when feeding compared to searching (T=−5.25, P= 0. 0000) or traveling (T=−3.15, P= 0. 0053) with no difference between searching and traveling. (F) Number of blows (NB) did not differ significantly with activity (F= 1.77, P= 0.172). (G) Blow rate (BR) did not differ significantly with activity (F= 0.43, P= 0.649).


Activity Budgets

Feeding and searching for suitable prey patches were the dominant activities we observed, which is not surprising since gray whales need to store substantial body fat during the summer months to meet their energetic requirements while migrating and breeding (Rice and Wolman 1971) and are not thought to derive much nourishment from opportunistic feeding (Oliver et al. 1983) observed during migration (Sund 1975, Wellington and Anderson 1978) and in their winter breeding grounds (Sanchez-Pacheco et al. 2001). Other studies of gray whales in their summer feeding grounds have found a similar emphasis on feeding-related activities, followed by traveling, socializing, and resting (Table 1). Our results (77% for feeding and searching combined) fall within the range of reported values (41% to 87%) for time spent in feeding-related activities. Direct comparisons to other studies remain difficult, however, because of differences in how behaviors are categorized. For example, Mallonee (1991) was the only other study to include a searching category. Searching activities may have been grouped as feeding or traveling by other researchers, causing an overestimate of both those values. Although these defined activities represent a continuum of behaviors rather than discrete states, it is important to try and make the distinction because time spent actively feeding is necessary to make accurate estimates of energetic intake and the potential impact of gray whales on their prey populations. We found significant differences in the respiratory and diving parameters associated with these activities, supporting the assertion that the behaviors classified as searching were truly different from both active feeding and traveling.

Table 1.  Activity budgets for gray whales in summer feeding grounds reported in the literature.
LocationFeeding (%)Searching (%)Traveling (%)Socializing (%)Resting (%)Other (%)Source
  1. Note: NR indicates variables not reported; sample size refers to the number of observation periods, or the number of whales as indicated.

  2. aAerial observations, values reported as percent of whales engaged in each activity.

  3. bWe grouped numerous categories based on probable functions listed by author, values reported as percent of all surface-dive sequences.

  4. cAverage of values reported for 1990 and 1991 separately.

  5. dPrey differed in each bay with tube-dwelling amphipods in Ahous Bay and mysids in Pachena Bay.

Arctica (n= 1,543 whales)41NR49NR6.43.5Moore et al. (1986)
Chukchi Seaa (n= 821 whales)63NR249.12.12.7Clarke et al. (1989)
Northern CAb (n= 11 obs)264616NR7.63.7Mallonee (1991)
Agate Bay, CAc (n= 21 whales)70NR18NRNR12Hawkinson (1992)
Bamfield, BCd (Ahous Bay, n= 5 obs/Pachena Bay, n= 14 obs)87/62NR0/20NR0/613/12Guerrero (1989)
Central Coast, BC (n= 394 obs)39381580.2NRThis study

Time spent in non-feeding related activities was generally similar between studies (Table 1). Most researchers have found that travel accounts for between 13% and 24% of a whale's time. Whales in our study only spent 15% of their time traveling, but this value is likely an underestimate because most of our observations were made within foraging bays and whales traveled more when moving between the bays. Moore et al. (1986) observed a predominance of travel (48%), which suggests they may have included searching behaviors in the travel category. Also, they used aerial surveys and comparisons between activity budgets based on individuals and estimates of group activity from aerial observations must be made with caution. Resting behavior was rarely observed in any of the studies, but anecdotal evidence suggests that the animals spend a larger proportion of their time resting at night (Guerrero 1989; W. Megill, personal observation), when few, if any, observations were made. Clarke et al. (1989) was the only other study to consider socializing as a separate category, and our values are very similar to theirs. The occurrence of social interactions appears to increase in the fall, as the breeding season approaches since, in years following this study, mating behavior was observed in our study area in the late summer and early autumn (W. Megill, personal observation).

Diving Behavior

Gray whales generally make short dives compared to other mysticetes: Bowheads average 6.3 min (Würsig et al. 1984), blue whales 6.6 min (Croll et al. 2001), and humpbacks 3.0 min (Dolphin 1987a). Sumich (1983) calculated a maximum aerobic breath hold near 6 min for migrating gray whales swimming at speeds less than 2 m·s−1. Although we did observe a few dives beyond this limit, 98.6% of the recorded dives were <6 min and the overall mean dive time was 2.24 min, indicating that the majority of dives displayed in this study were well within the aerobic dive limit. Blow rate can serve as an indicator of physiological demand (Dolphin 1987b). The mean blow rate of 1.14 breaths·min−1 that we observed was similar to values for other summer resident whales (Table 2) but higher than the mean reported by Sumich (1983) of 0.72 breaths·min−1, and is more comparable to his maximum value of 1.12 breaths·min−1. This suggests that foraging activities are more energetically expensive than migrating in gray whales, likely because during migration they are swimming at speeds near their minimum cost of transport (Sumich 1983), and they may have a reduced metabolic rate while fasting. In addition, the body maneuvers (Woodward 2006) and diving patterns associated with feeding and searching may require more energy than steady swimming. Acevedo-Gutierrez et al. (2002) demonstrated that lunge-feeding by blue and fin whales was significantly more energetically expensive than non-foraging dives and limited the duration of their dives.

Table 2.  Respiratory and diving parameters for gray whales reported in the literature. Study location, major activity, and primary prey (when applicable) are listed. Sample size (n) refers to the number of each occurrence observed for each variable, except in this study where we considered each observation period as a sample.
Location (activity)Dive time (min)Number of blows per surfacingBlow interval (s)Blow rate (min−1)Surface time (min, %)Source
  1. Note: NR indicates variables not reported. Parentheses around the surface time values indicate we converted the reported value from minutes to a percentage by dividing the mean surface time by the sum of the mean surface time + mean dive time.

  2. aValues from tagged animals, therefore values included in each parameter differ from the other studies: dive rate includes submergences between blows so blow intervals could not be calculated and surface time does not include the blow intervals.

  3. bValues from feeding animals only, does not include other activities.

San Ignacioa, Mexico (Breeding)1.6 (SE = 0.02, n= 11,080)NRNRNR0.07, 4.5% (SE = 0.01, n= 11,229)Harvey and Mate (1984)
Bering Sea, AK (Feeding-amphipods)3.2 (SD = 1.27, n= 905)4.2 (SD = 2.23, n= 1,050)13.5 (SD = 7.27, n= 3,503)1.05 (n= 811)0.89 (21.9%) (SD = 0.73, n= 1,062)Würsig et al. (1986)
Northern CA (Feeding-benthic)3.4 (SD = 1.56, n= 194)3.5 (SD = 1.90, n= 187)18.4 (SD = 10.5, n= 4,98)NR0.78 (18.7%) (SD = 0.69, n= 188)Mallonee (1991)
Agate Bay, CAb (Feeding-cumaceans)3.1 (SD = 1.13, n= 2,443)4.5 (SD = 1.94, n= 2,439)17.6 (SD = 13.5, n= 3,816)NR1.2 (27.9%) (SD = 0.89, n= 2,455)Hawkinson (1992)
Bamfield, BC (Feeding-amphipods)3.7 (SD = 0.68, n= 17)3.8 (SD = 0.8, n= 22)NRNR0.72 (16.3%) (SD = 0.20, n= 22)Oliver et al. (1984)
Bamfield, BC (Feeding-mysids)1.3 (SE = 0.70, n= 175)NR11.8 (SE = 3.75, n= 197)NRNRMurison et al. (1984)
Bamfield, BCb (Feeding-amphipods)3.7 (SD = 1.04, n= 248)4.2 (SD = 1.29, n= 250)14.9 (SD = 3.26, n= 826)0.92 (n= 244)0.9 (19.6%) (SD = 0.33, n= 250)Guerrero (1989)
(Feeding-mysids)1.6 (SD = 1.08, n= 1,515)2.4 (SD = 1.78, n= 1,494)12.4 (SD = 8.32, n= 2,197)1.21 (n= 1379)0.4 (20%) (SD = 0.39, n= 1,494) 
Central Coast, BC (Feeding-mysids)2.2 (SD = 0.86, n= 366)2.9 (SD = 1.33, n= 373)15.4 (SD = 4.37, n= 360)1.14 (SD = 0.367, n= 366)0.50 16.9% (SD = 0.341, n= 372)This study

Effects of Activity

Whale activity in this study area can be characterized by distinct diving and respiratory patterns. When feeding on mysids, the whales made more frequent dives with shorter intervals between blows, thus they spent less time at the surface than they did while searching or traveling. Other studies have also examined the effects of activity on diving patterns. Bottom-feeding whales make longer dives, respire more times per surfacing and spend longer at the surface compared to other activities (Würsig et al. 1986, Guerrero 1989, Mallonee 1991, Hawkinson 1992).

A comparison of respiratory/dive parameters with other studies on gray whales emphasizes the effect of prey and feeding methods on diving patterns (Table 2). The duration of dives appears to be longest in areas where whales were feeding on benthic prey, whereas mysid feeding whales exhibited dive times approximately 50% shorter on average. Whales feeding on mysids respire fewer times per surfacing period than those feeding on amphipods, and the reports of mean blow intervals range from 11.8 to 18.4 s with mysid feeding whales on the low end of this range. Guerrero (1989) directly compared prey types and found that whales made shorter dives, respired fewer times with shorter intervals, and displayed briefer surfacing periods when feeding on mysids than when feeding on benthic amphipods.

There are a number of potential reasons for the distinctive diving patterns exhibited by mysid feeding whales. The brief surfacings maximize the time spent submerged in shallow waters and thus the time available to consume prey. An explanation for making frequent dives instead of extending the DT comes from in situ observations that mysids temporarily scatter when disturbed (Modlin 1990, Stelle 2001). Thus, the whales may surface between feeding sweeps to allow the swarms to reform for efficient capture. The diving/respiratory parameters along with behavioral observations suggest that once a whale finds a suitable patch of mysids, the feeding animal dives frequently in the same area, exhibiting extreme flexibility and maneuverability to return to the same location (Woodward 2006). The cue to terminate a feeding bout is unknown. Based on optimal foraging theory (MacArthur and Pianka 1966) the whale should feed in a patch until the density or abundance becomes low enough that is energetically more worthwhile to move to a new patch. Yet, scuba divers observed no obvious change in density or abundance of mysid swarms after whales fed in kelp beds sampled earlier in the day (Stelle 2001). It is possible that mysids in previously disturbed swarms exhibit a greater escape response so that after a period of feeding disturbance in one area the swarms become more difficult for the whale to capture, and the whale moves on to another area. When feeding on either planktonic or benthic prey, gray whales display a similar feeding technique of rolling on their sides (Nerini 1984, Woodward 2006) and using suction (Ray and Schevill 1974), thus the main difference is the mobility of the prey. Future studies could employ underwater viewing systems to determine if this observed diving pattern optimizes capture of planktonic prey before they become too dispersed to be captured efficiently.

Relationships between diving and respiratory patterns with activity can help to identify behavioral states in the absence of other cues. Würsig et al. (1986) suggest that the duration of dives may be a useful indicator of feeding activity. Yet the strong effect of prey type on dive time may confound conclusions, especially in areas where the whales shift between available prey choices (e.g., Darling et al. 1998, Guerrero 1989, Dunham and Duffus 2001). The mean interval between blows was found to be significantly shorter in feeding whales in all studies, except in Agate Bay (Hawkinson 1992), regardless of prey. Thus blow interval, which can be easily measured, may serve as a valuable indicator of activity states in gray whales.


It is important to recognize that activities and diving patterns are subject to individual variability and will also differ with season and geographical region thus the data reported here is only representative of summer behavior during daylight hours in the habitat studied. Sampling the same individuals repeatedly over the season could bias our results but we had a large enough sample size (at least 35 individual whales and 388 h of data) that the effect should be negligible. Activity had the same influence on diving patterns in both years, with feeding whales exhibiting significantly distinctive diving behavior. Although we found significant differences in some dive parameters between years, activity budgets were the same in each year thus the between-year effects may be due to the larger sample size in 2000 and not to any biological differences in their diving behaviors. These common trends along with comparisons with other studies suggest that diving behavior is meaningfully affected by the activity of the whale.

Gray whales feed low in the trophic web and are opportunistic feeders (Nerini 1984, Darling et al. 1998) thus they can serve as conspicuous bio-indicators of shifts in biological productivity (Moore et al. 2003). The major ecosystem shift occurring in the Bering Sea will likely have significant community-wide impacts (Grebmeier et al. 2006). Clearly gray whales can switch their diet between benthic and planktonic prey, yet questions remain as to whether feeding on mysids provides equivalent nutrition and energy, and whether there is enough suitable habitat to support a stable population. As the whales expand their range and shift their distribution, this baseline knowledge of their activity budgets and diving behavior in a pelagic-dominated feeding area will help us to analyze any changes in their behavior. Future research should focus on quantifying the physiological requirements of the various activities and measuring the energetic gains from different prey to examine optimal foraging theory in gray whales and help predict how they will respond to changes in ecosystem productivity.


We would like to thank all of the individuals who contributed to this research. W. Hamner of University of California, Los Angeles provided support and guidance throughout the project. Many crew members and volunteers worked with the Coastal Ecosystems Research Foundation in the field, including M. Berger, D. Boggs, N. Cameron, K. Gray, M. Heath, J. Holloway, J. Muise, C. Picco, D. Randall, J. Scott-Ashe, J. Taylor, A. Walkus Jr., J. Winn, and B. Woodward. Statistical advice was provided by R. Ambrose at UCLA, D. Mathiason and J. Halavin at Rochester Institute of Technology, with assistance from J. Blackwood and P. Martino. The manuscript was improved by revisions suggested by F. Pough at RIT and anonymous reviewers. This work was funded by grants from UCLA, University Research Expeditions Program, Earthwatch Institute, and the BC Ministry of Environment, Lands & Parks.