Editor: Nigel Bennett
Optimization of transit strategies while diving in foraging king penguins
Article first published online: 19 MAR 2013
© 2013 The Zoological Society of London
Journal of Zoology
Volume 290, Issue 3, pages 181–191, July 2013
How to Cite
Hanuise, N., Bost, C.-A. and Handrich, Y. (2013), Optimization of transit strategies while diving in foraging king penguins. Journal of Zoology, 290: 181–191. doi: 10.1111/jzo.12026
- Issue published online: 19 JUN 2013
- Article first published online: 19 MAR 2013
- Manuscript Accepted: 24 JAN 2013
- Manuscript Revised: 15 JAN 2013
- Manuscript Received: 26 JUN 2012
- French Ministry of Research
- Institut Polaire Français
- behavioural adjustments;
- depth anticipation;
- foraging success;
- diving angle;
- stroke frequency;
- swimming speed;
- transit time;
- vertical speed
- Top of page
- Supporting Information
Optimal foraging theories predict that air-breathing, diving foragers should maximize time spent at feeding depths, and minimize time spent travelling between surface and depth (transits). The second part of this hypothesis was tested in free-ranging king penguins Aptenodytes patagonicus using measurements of vertical speed, swimming speed, body angle and flipper stroke frequency during transits in relation to an index of foraging success (number of wiggles), during the bottom and the ascent phases of the dive. We found that, except for flipper stroke frequency, all measured variables increased with diving depth and foraging activity. The change in vertical speed was driven mainly by a change in body angle and a slight change in swimming speed. These results suggest a shortening of transit duration in response to increased foraging activity. Whereas the time spent commuting between the surface and foraging depths was reduced when foraging activity was high, vertical speed was only at its maximum over a small part of both ascent and descent phases of the dive. Within the first 10 m of descent, vertical speed increased with maximum dive depth and an index of foraging activity, suggesting that penguins anticipated their diving depth and foraging activity. Our results show that foraging king penguins adjust their diving behaviour in response to both diving depth and foraging activity. Further studies should consider ecological, physiological or mechanical constraints as factors that may limit foraging optimization.
The survival, growth and reproduction of animals depend on their foraging success. Thus, evolution should favour behaviours, including movements where foraging is optimized (Stephens & Krebs, 1986). Air-breathing, diving aquatic predators such as pinnipeds and seabirds are central place foragers (Orians & Pearson, 1979), which forage at sea, but need to come back onto land for breeding duties or to moult. Moreover, when foraging at depth, they have to commute periodically from the surface, where gas exchange takes place, to the depths at which prey are found (Kooyman, 1989). Therefore, they have to continuously make decisions about where, when and what to feed on in conditions that places constraints on air-breathing, endotherm foragers. Such predators thus provide excellent models for studying foraging decisions.
Theoretical models generally assume that diving predators should maximize the proportion of time spent at favourable foraging depths (Houston & Carbone, 1992; Thompson & Fedak, 2001; Mori et al., 2002), which often corresponds to the bottom period of the dive (around the maximum dive depth). Thus, they should reduce both the duration of transit phases, during descent to the bottom and ascent to the surface, and the time spent at the surface recovering from their apnoea. Reduction in surface and transit times may be particularly beneficial for predators that feed on ephemeral, elusive patches of prey. However, they should also minimize oxygen use during the transits, in order to maximize the amount of oxygen available at the foraging depth, thus implying constraints on diving behaviour.
Behavioural adjustments during transit phases of a dive can occur through changes in swimming speed or body angle. Swimming speed is the result of propulsive force, sustained by foot/flipper stroke frequency/intensity (Sato et al., 2003; Lovvorn et al., 2004; Watanuki et al., 2006), and limited by water drag effect. During transit phases, marine mammals and seabirds usually swim at a speed close to the values that minimize the cost of transport, and thus cruising speed is a relatively fixed variable for a given body size (Schmidt-Nielsen, 1972; Culik, Bannasch & Wilson, 1994; Boyd, McCafferty & Walker, 1997; Ropert-Coudert et al., 2002). Therefore, vertical speed of descent or ascent can be modulated in diving birds, such as penguins, by modifying the angle of the transit (Ropert-Coudert et al., 2001).
Penguins provide an excellent model to study fine-scale adjustments in diving behaviour as they are efficient avian divers (Wilson, 1995; Green et al., 2003; Ropert-Coudert et al., 2006; Halsey et al., 2007b). Over the past 20 years, continuous technological advances in biologging (Rutz & Hays, 2009) have resulted in improved accuracy when recording data on dive profiles, body angle, flipper stroke frequency and swimming speed. Some techniques can also provide an index of prey capture in penguin species (Bost et al., 2007), thus allowing to link foraging success to diving behaviour. Calculation of mean diving angle has shown that Adélie Pygoscelis adeliae and little penguins Eudyptula minor (Ropert-Coudert et al., 2001, 2006) can adjust their diving behaviour in response to foraging success; these two species exhibit steeper mean descent angles when they have encountered prey in the previous dive.
Anticipation of the characteristics of a dive in terms of duration, depth and foraging success has also been investigated, via for example the number of breaths taken during the preceding surface episode (Wilson, 2003). Studies on diving animals have described a positive relationship between mean descent vertical speed and maximum depth of a dive (Boyd, Reid & Bevan, 1995; Wilson et al., 1996; Charrassin, Le Maho & Bost, 2002). This result could support the idea that divers predict their upcoming dive performance (Fig. 1a), but only if vertical speed is constant along the descent, which does not seem to be the main case (Ropert-Coudert et al., 2001; Wilson et al., 2010, but see Sato et al., 2004), or if its rate of change depends on ultimate dive depth. The same pattern could, however, also result from an increase in vertical speed during the descent phase, rather than anticipation (Fig. 1b).
In the present study, we investigated how a deep, air-breathing diver, the king penguin Aptenodytes patagonicus, maximizes the time spent at profitable foraging depths. King penguins mainly feed on patchy, small, mesopelagic fish distributed at great depths during the day (100–300 m, Bost et al., 2002). Our goal was to test some optimal foraging predictions in a deep-diving predator; maximization of efficient foraging time and minimization of time spent commuting and recovering. We compared in free-ranging penguins instrumented with data loggers the change in vertical and swimming speeds, body angle and flipper stroke frequency in relation to maximum depth. We also investigated whether such deep divers can adjust their diving behaviour in response to the foraging success of the current/previous dive. Firstly, we examined the mean values of these four variables during transit. We predicted penguins would reduce transit time (i.e. duration of dive descent phase from the surface to the bottom and duration of dive ascent phase from the bottom to the surface) by increasing their vertical speed and body angle as a response to higher foraging success during the preceding and/or the current diving event (Wilson & Liebsch, 2003). Secondly, we examined variations of these four variables throughout the depth range experienced during transits. This enabled us to investigate how penguins may anticipate the nature of the dive they are going to undertake in terms of transit rates.
- Top of page
- Supporting Information
Study birds and data loggers
The study was carried out on Possession Island, Crozet Archipelago (46.4°S, 51.8°E) from December 2003 to March 2004. Birds used in the study were king penguins breeding at La Baie du Marin, a colony of approximately 16 000 pairs (Delord, Barbraud & Weimerskirch, 2004). The procedures received the approval of the ethics committee of the French Polar Institute (IPEV) and of the French Ministry of the Environment. Detailed description of the general surgical and handling procedure are given in Froget et al. (2004).
Six breeding male king penguins were captured while brooding an egg and immediately subjected to isoflurane-anaesthesia, during which they were fitted with data loggers. SMAD data loggers (DEPE-IPHC, Strasbourg, France; 80 × 25 × 10 mm, 54 g) were externally attached to the lower-back feathers of each animal to diminish hydrodynamic drag (Bannasch, Wilson & Culik, 1994) and recorded depth every 2 s. SMAD were also programmed to measure tail-to-head (surge) and ventral-to-dorsal (heave) accelerations during two 1-h high-frequency sessions per day when penguins performed deep dives, and stored these measurements 32 times per second. Cross-sectional area (CSA) of the external logger (2.5 cm2) represented less than 1% of the smallest bird's CSA. Modified Mk7 data loggers (Wildlife Computers, Redmond, WA, USA) were also implanted subcutaneously for a study of peripheral temperatures; these results have been described previously in Schmidt (2006). Together, the mass of both loggers (87 g) represented less than 0.8% of the smallest bird's mass. The penguins undertook a foraging trip at sea 15–18 days later, after being relieved by their partners. After their return to the colony, the birds were recaptured and anaesthetized using the same procedure, and the loggers were removed. All the loggers were recovered, of which five had recorded usable data. Data from these loggers were extracted, prepared and analysed using purpose-written computer programs in Matlab 6.0 (The MathsWorks, Natick, MA, USA).
Dives >50 m, hereafter called ‘deep dives’, were used for analysis as they represent the majority of the foraging dives of king penguins (Charrassin et al., 1998). For each dive analysed, the following parameters were calculated: maximum dive depth (m), dive duration (s), subsequent surface interval duration until the next dive of any depth (s), subsequent time interval until the next deep dive (s), rank of the dive in a bout (i.e. sequence of successive dives), number of wiggles during the bottom phase or the entire dive. Wiggles are a particular, undulation-like pattern in the dive profile over time. In king penguins, the number of ingestions recorded per dive is highly related to the number of wiggles (Bost et al., 2007), so we used the number of wiggles as a proxy of foraging success during that dive. Wiggles, dive phases (descent, bottom, ascent) and bouts were characterized as described by Halsey, Bost & Handrich (2007a). Vertical speeds were calculated as depth differential divided by the sampling interval of depth data. Acceleration data were filtered in high- (>0.5 Hz) and low-frequency components to calculate flipper stroke frequency and body angle of the penguins, respectively, by using procedures adapted from Sato et al. (2003). As the loggers were not exactly parallel to the longitudinal axes of the birds, angle between the horizontal and the direction of diving was corrected assuming that it was on average 0° while the birds were at the sea surface between dives. Descending angles and vertical speeds are represented as negative values. Swimming speed Ss was calculated using vertical speed Vs and dive angle θ with the following equation. Ss = Vs/sin(θ), assuming that the birds moved in the direction of the head-to-tail axis (Watanuki et al., 2003). In order to reduce imprecision when both Vs and sin (θ) reached zero, swimming speed was only calculated when vertical speed was greater than 0.9 m s −1 and when dive angle was greater than 25°. Thus, swimming speed was calculated for descent and ascent phases, but not for the bottom phase. Body angle, swimming speed and flipper stroke frequency were only available when acceleration data were recorded, that is during the two daily 1-h high-frequency recording sessions.
Generalized linear mixed models (GLMMs)
Data were analysed using GLMMs in R 2.9 software (R Foundation for Statistical Computing, Vienna, Austria) to identify the factors influencing the parameters of the transit phases between the surface and the dive bottom. Dive rank and individual identity were included in the models as random effects. Temporal autocorrelation was accounted for by incorporating a lag one autoregressive term (Beck et al., 2003). These models examined how five diving variables (maximum dive depth, dive duration, surface interval duration, rank of the dive in a bout and number of wiggles) affect vertical speed, swimming speed, body angle and flipper stroke frequency. The number of wiggles was counted during the entire previous dive or during the bottom phase of the current dive, in order to assess the effects of the feeding success on the descent and ascent phases, respectively.
A first step was to evaluate the effects of the diving variables on the mean characteristics of the transit. A total of eight GLMMs were built, two for each of the four parameters of the transit phases (vertical speed, swimming speed, body angle and flipper stroke frequency), for descent and ascent phases, respectively. Mean values of parameters do not describe the variation of these parameters in the water column, but rather give only two average values between the surface and the bottom, for each dive. Hence, a second step was to investigate this further by considering the variation during transits of the different parameters. GLMMs analyses were thus conducted to determine the effects of depth on the parameters of transit phases during the first 100 m of the descent and during the last 100 m of the ascent. Data were then divided in 20 5-m bins, from 0–5 m to 95–100 m, and 20 GLMMs were built for each transit phase variable and transit phase (one model for each bin, see Appendix S1). Maximum dive depth, dive duration, surface interval duration, rank of the dive in a bout, number of wiggles (continuous variables) and all second-order interaction were used in the GLMMs. Non-significant terms were then removed, one iteration at a time, by backwards elimination. Non-significant main effects were kept in the model if the variable in question was part of a statistically significant interaction (Halsey et al., 2007b). Although the variables were continuous, we split the two main independent in three bins (number of wiggles: 0–2, 3–4, 5–12, maximum dive depth: 50–95, 95–120, 120–260 m) for illustration purposes.
- Top of page
- Supporting Information
The five instrumented king penguins performed 7631 deep dives out of a total of 29 299 dives (Table 1). Swimming speed, body angle and flipper stroke frequency were calculated during 572 deep dives (Table 2). Mean vertical speed during descent and ascent were comparable. Mean descent dive angle was steeper than mean ascent angle. Mean flipper stroke frequency was higher during descent than during ascent, and had intermediate values during the bottom phase. Swimming speed was higher during ascent than during descent.
|Bird A1||Bird A2||Bird A3||Bird A4||Bird A5|
|Trip duration (d)||14.4||14.9||20.2||19.4||29.3|
|Total number of dives||5144||4066||5591||4411||10087|
|Mean maximum depth (m)||95 ± 23||120 ± 29||108 ± 30||109 ± 37||135 ± 34|
|Mean duration (s)||211 ± 26||222 ± 36||226 ± 37||224 ± 47||248 ± 44|
|Descent vertical speed (m s−1)||−1.27 ± 0.25||−1.50 ± 0.25||−1.27 ± 0.22||−1.24 ± 0.20||−1.32 ± 0.22|
|Ascent vertical speed (m s−1)||1.25 ± 0.28||1.29 ± 0.26||1.28 ± 0.28||1.21 ± 0.26||1.43 ± 0.30|
|Bird A1||Bird A2||Bird A3||Bird A4||Bird A5|
|Total number of dives||270||233||211||150||190|
|Mean maximum depth (m)||90 ± 19||124 ± 27||100 ± 31||100 ± 24||120 ± 31|
|Mean duration (s)||207 ± 24||229 ± 27||220 ± 29||214 ± 38||237 ± 35|
|Mean vertical speed (m s−1)||−1.23 ± 0.24||−1.54 ± 0.27||−1.23 ± 0.24||−1.28 ± 0.20||−1.25 ± 0.26|
|Mean speed (m s−1)||1.80 ± 0.11||1.86 ± 0.09||1.76 ± 0.12||1.63 ± 0.08||1.82 ± 0.09|
|Mean body angle (°)||−48.3 ± 9.2||−61.4 ± 10.6||−48.1 ± 11.3||−55.9 ± 10.0||−47.5 ± 10.4|
|Mean stroke frequency (Hz)||1.51 ± 0.12||1.52 ± 0.08||1.33 ± 0.11||1.31 ± 0.09||1.49 ± 0.13|
|Mean stroke frequency (Hz)||1.05 ± 0.33||1.30 ± 0.18||1.00 ± 0.24||1.08 ± 0.18||1.08 ± 0.21|
|Mean vertical speed (m s−1)||1.28 ± 0.28||1.25 ± 0.22||1.22 ± 0.32||1.33 ± 0.26||1.37 ± 0.37|
|Mean speed (m s−1)||2.25 ± 0.16||2.30 ± 0.11||2.09 ± 0.15||2.16 ± 0.18||2.18 ± 0.11|
|Mean body angle (°)||33.3 ± 8.4||30.7 ± 7.1||36.3 ± 9.7||37.6 ± 8.0||40.5 ± 11.6|
|Mean stroke frequency (Hz)||0.32 ± 0.25||0.59 ± 0.20||0.39 ± 0.19||0.38 ± 0.18||0.64 ± 0.17|
Mean transit rates analysis
Both maximum dive depth and number of wiggles impacted on mean descent and ascent vertical and swimming speeds, body angle and flipper stroke frequency.
Mean vertical and swimming speeds during descent increased significantly as maximum dive depth increased and as number of wiggles during the previous dive increased (Table 3, Fig. 2a,c). Mean vertical and swimming speeds during ascent increased significantly as maximum dive depth increased and as number of wiggles during the bottom phase of the current dive increased (Table 3, Fig. 2b,d).
|Factor||Vertical speed||Swimming speed||Body angle||Stroke frequency|
|Maximum dive depth||0.011***||0.014***||0.001***||0.005***||0.593***||0.390***||NS||0.001*|
|Number of wigglesa||0.041***||0.118***||0.007***||0.024***||4.700***||4.916***||0.023**||−0.029***|
|Rank in a bout||0.001*||−0.001**||0.001**||NS||0.109***||0.028*||0.001*||NS|
|Max. depth × No wigglesa||−0.000*||NS||NS||NS||NS||NS||−0.000*||NS|
|Max. depth × Dive duration||−0.000***||−0.000***||NS||NS||−0.002***||−0.001***||NS||NS|
|Max. depth × Surface duration||−0.000***||−0.000***||NS||−0.000*||NS||NS||NS||NS|
|Max. depth × Rank||−0.000***||0.000***||NS||−0.000**||NS||NS||NS||NS|
|No wigglesa × Dive duration||NS||−0.000***||NS||NS||−0.017**||−0.016***||NS||NS|
|No wigglesa × Surface duration||NS||NS||NS||NS||NS||0.018***||NS||NS|
|No wigglesa × Rank||−0.000***||0.000**||NS||NS||−0.020**||NS||−0.000**||NS|
|Dive duration × Surface duration||0.000***||0.000***||NS||0.000**||NS||NS||NS||NS|
|Dive duration × Rank||0.000***||0.000***||NS||0.000*||NS||NS||NS||NS|
|Surface duration × Rank||0.000**||0.000***||NS||NS||NS||NS||NS||NS|
Mean descent angle increased significantly as maximum dive depth increased and as number of wiggles during the previous dive increased (Table 3, Fig. 2e). Similarly, mean ascent angle increased significantly as maximum dive depth increased and as number of wiggles during the bottom phase of the current dive increased (Table 3, Fig. 2f).
Mean descent flipper stroke frequency increased significantly as number of wiggles during the previous dive increased (Table 3, Fig. 2g). Furthermore, mean ascent flipper stroke frequency increased significantly as maximum dive depth increased and as number of wiggles during the bottom phase of the current dive decreased (Table 3, Fig. 2h).
For both descent and ascent, the range of changes was large in vertical speed (33 and 60%) and in body angle (33 and 44%), and greatly lower in swimming speed (7 and 10%).
Transit rates during the first/last 100 m analysis
Here, we investigated the variation during transit of different parameters: vertical speed, swimming speed, dive angle and flipper stroke frequency in relation to maximum dive depth (Fig. 3) and the number of wiggles (Fig. 4).
Vertical speed and body angle followed the same pattern of variations during descent phases. Firstly, they increased from 1.0 m s −1 and 45° to a maximum (up to 1.8 m s −1 and 80°) at about half of the maximum dive depth, then started to decrease as the bird approached the bottom. When considered below a 5-m depth step in the water column, the birds' vertical speed and body angle were positively affected by both maximum dive depth (Fig. 3a,e) and number of wiggles during the previous dive (Fig. 4a,e). Swimming speed during descent sharply increased in the first 5 m, then slightly increased before reaching a maximum value at about 2.0 m s −1 (Fig. 3c). Flipper stroke frequency decreased during descent from around 2.0 Hz in the first 5 m to around 1.0 Hz at the beginning of the bottom phase, and was positively affected by maximum dive depth (Fig. 3g).
Vertical speed during ascent increased except during the last 30 m where it slightly decreased, and was positively affected by both maximum dive depth (Fig. 3b) and number of wiggles during the bottom of the current dive (Fig. 4b). Swimming speed during ascent remained constant at about 2.0 m s −1 until a depth of 30–40 m where it started to increase, reaching a maximum value of 2.5–3.0 m s −1 at a depth of 15–20 m, and was positively affected by maximum dive depth during the last 40 m (Fig. 3d). Body angle during ascent increased except during the last 40 m where it quickly decreased, and was positively affected by both maximum dive depth (Fig. 3f) and number of wiggles during the bottom of the current dive (Fig. 4f). Flipper stroke frequency during ascent continuously decreased from around 0.9 Hz at the end of the bottom period to 0 Hz at the surface. The suppression of stroke movements appeared at a depth equal to approximately 35% of maximum dive depth. Ascent flipper stroke frequency was negatively affected by the number of wiggles during the bottom phase of the current dive (Fig. 4h).
- Top of page
- Supporting Information
Theoretical studies of diving behaviour have proposed strategies that maximize the proportion of time spent submerged mostly based on the use/recovery of oxygen reserves (Carbone & Houston, 1994). Thus, divers should maximize the time spent in a favourable patch at depth by maximizing the oxygen store available at the foraging depth. If diving predators increase time spent foraging in a patch, that is at the bottom of a dive, they should in turn reduce the time spent commuting or recovering at the surface. The present study shows that deep divers such as king penguins can adjust their transit time from the surface to the bottom of a dive in response to the success of the previous dive, and from the bottom to the surface in response to the success of the current one. Thus, this study is one of the first to show both anticipation and optimization of diving. In king penguins, these adjustments result from changes in vertical speed, which are driven mainly by large changes in diving angle and slight changes in swimming speed.
Transit rates by 5-m steps throughout descent and ascent
Previous studies have described mean swimming and vertical speeds, body angle or flipper stroke frequency in penguins and other diving seabirds in relation to depth. However, few of them have focussed on variations in these rates within descent or ascent phases (Watanuki et al., 2005, 2006; Cook et al., 2010). Here, for the first time, we report on the progressive changes occurring with current depth in four parameters influencing the transit duration between the surface and the dive bottom in a deep diver.
During descent, instantaneous vertical speed changed with current depth. The pattern of changes was mainly due to variations in body angle: penguins first increased their descent angle from the surface to the middle of the descent, up to a value of 50–60° and then decreased it. Swimming speed quickly reached values around 1.8 m s −1 and gradually, but very slightly, increased during descent. During these dive phases, the range of speeds recorded correspond to minimal cost of transport in horizontally swimming king penguins (Culik et al., 1996), suggesting that energetic constraints strongly reduce the span of changes in swimming speed. Flipper stroke frequency was at a maximum at the beginning of the descent, in the first metres of the water column where positive buoyancy is high, and then decreased. This initial vigorous flipper stroking suggests hard work undertaken against positive buoyancy at shallow depths (Sato et al., 2002).
During ascent, instantaneous vertical speed changed with current depth, in relation to both changes in body angle and swimming speed. Body angle increased during the first part of the ascent and sharply decreased during the second part. Swimming speed remained approximately constant at around 2.0 m s −1 during the first part of ascent, and increased up to 2.5 m s −1 prior to surfacing. As a result of changes in these two parameters, vertical speed slowly increased, then stabilized and gradually decreased in the last 20–30 m of ascent. Flipper stroke frequency was low at the beginning of ascent, and stroking decreased until ceasing just prior to surfacing. An increase in swimming speed despite a decrease in flipper beat frequency confirms that penguins use positive buoyancy to ascend passively over the last 40 m (Sato et al., 2002). Despite large increases in swimming speed before surfacing, reduction of body angle leads to a limited increase in vertical speed. Two main hypotheses could explain such behaviour, which results in delayed surfacing, horizontal travelling and avoidance of decompression consequences (Sato et al., 2002, 2004).
It is still unclear how seabirds avoid decompression sickness; ascending slowly to the surface has been one proposed hypothesis (Sato et al., 2002; Fahlman et al., 2007). However, one of our studied penguins (A5) did not exhibit a reduction in body angle and regularly surfaced at vertical speeds exceeding 2.2 m s −1. The second main hypothesis explaining delaying ascent is the use of buoyancy to travel horizontally (Sato et al., 2002). This would predict that penguins increase the horizontal component of the ascent phase after not encountering a prey patch, in order to prospect a bigger volume of the water column. Conversely, we could predict that penguins would minimize horizontal travelling after encountering a prey patch, in order to maximize the probability of relocating the same patch. Indeed, we observed that ascent angles were higher and ascent flipper stroke frequency was lower after encountering prey, thus reducing horizontal travelling. However, as no data were available on the 3-D structure of the dives or on the surface locations between successive dives, we cannot confirm this hypothesis of horizontal travelling in the search of a new foraging patch.
Effects of higher feeding success on transit rates
The present study indicates that king penguins exhibited higher vertical speed during transit times, linked with a steeper body angle and a small increase in swimming speed following productive foraging during the preceding dive or during the current one. Similar results have been reported in two smaller penguin species performing shallower dives. In Adélie penguins, mean angles of ascent and subsequent descent are steeper after bottom phases where prey ingestions occurred (Ropert-Coudert et al., 2001); and in little penguins, mean descent angles were steeper after dives where prey pursuit occurred (Ropert-Coudert et al., 2006).
Together, these results show that penguins are able to optimize their diving behaviour by adjusting their transit. King penguins feed on myctophid fishes patchily distributed in dense monospecific shoals during the day (Perissinotto & McQuaid, 1992). When the penguin has fed successfully on a favourable patch, we can assume the preferred foraging option is to attempt to relocate the same patch before its dispersion after returning to the surface. By shortening their post-dive interval and descending faster, the penguins increase their probability of encountering the same patch in the following dive.
The fact that penguins ascended with a lower flipper stroke frequency after finding more prey during the bottom of the current dive was unexpected. Furthermore, this lower flipper stroke frequency seems not to handicap a faster ascent to the surface, which could be explained by higher buoyancy. We might hypothesize that penguins anticipated encountering prey and consequently increased respiratory air volume before submergence, which increased buoyancy up-thrust when ascending. Increased descent flipper stroke frequency after highly foraging dives, presumably to overcome this additional buoyancy, strongly supports this hypothesis. Such a relationship was already found when considering the maximum depth of a dive: the deeper the dives, the higher their initial buoyancy and earlier in the ascent phase penguins cease using their flippers (Sato et al., 2002). Moreover, Magellanic penguins Spheniscus magellanicus have also previously been shown to modulate their air volume in relation to both dive depth and foraging success (Wilson, 2003). Otherwise, the hypothesis of the described behaviour being primarily energy-saving may also be proposed. Diminishing flipper stroke frequency could be a means of reducing locomotory costs after a successful dive where pursuing and catching prey during wiggles must have been energetically costly.
Anticipation of dive depth
As shown in Fig. 1, a positive relationship between mean descent vertical speed and maximum depth of a dive is not sufficient to support the idea that divers predict their upcoming dive performance. Here, we analysed vertical transit rate by 5-m steps throughout the descent, for different maximum dive depths, and showed that as soon as after the first 5-m depth and at any given depth below, vertical speed increased with maximum dive depth. In our opinion, this observation strongly supports the suggestion that a behavioural anticipation occurred. When having to reach greater depths during a dive, penguins increased their vertical descent rate by increasing both body angle and flipper stroke frequency from the beginning of the dive. Higher flipper stroke frequency during descent at deeper depths could be related to greater work done against positive buoyancy, caused by greater air volume inhaled in anticipation of longer dive durations, as has been previously suggested (Sato et al., 2002).
- Top of page
- Supporting Information
Contrary to well-studied laboratory and terrestrial systems, optimal foraging studies have rarely dealt with quantifying foraging strategies in marine systems (Hindell, 2008). Here, we show that king penguins adjust their transit rate in terms of dive angle and swimming speed, based on the number of feeding opportunities they are likely to have during the next dive. To the best of our knowledge, these results are the first to show both anticipation and optimization of diving in response to depth and foraging success. Such decisions are likely to increase the efficiency of foraging efforts, both in terms of prey capture success when close to a foraging patch and in terms of deciding to attempt to find a new patch. Future work including the measurement of respiration activity and energy expenditure will provide fruitful insights into our understanding of animal optimal diving behaviour.
- Top of page
- Supporting Information
N.H. was funded by the French Ministry of Research. This work was supported by the Institut Polaire Français (IPEV, programme n°394, resp. CAB). We greatly thank A. Schmidt for contributing to fieldwork, A. Kato, P.A. Pistorius and an anonymous reviewer for helpful comments, C. Saraux and D. Babel for data of the control group, and L.G. Halsey and V.A. Viblanc for improvement of the English.
- Top of page
- Supporting Information
- 1994). Hydrodynamic aspects of design and attachment of a back-mounted device in penguins. J. Exp. Biol. 194, 83–96. , & (
- 2003). Sex differences in the diving behaviour of a size-dimorphic capital breeder: the grey seal. Anim. Behav. 66, 777–789. , , & (
- 2002). Feeding of diving predators and diel vertical migration of prey: king penguins' diet versus trawl sampling at Kerguelen Islands. Mar. Ecol. Prog. Ser. 227, 51–61. , , & (
- 2007). Changes in dive profiles as an indicator of feeding success in king and Adélie penguins. Deep Sea Res. Part II Top. Stud. Oceanogr. 54, 248–255. , , , , , & (
- 1995). Swimming speed and allocation of time during the dive cycle in antarctic fur seals. Anim. Behav. 50, 769–784. , & (
- 1997). Variation in foraging effort by lactating Antarctic fur seals: response to simulated increased foraging costs. Behav. Ecol. Sociobiol. 40, 135–144. , & (
- 1994). Patterns in the diving behaviour of the pochard, Aythya ferina : a test of an optimality model. Anim. Behav. 48, 457–465. & (
- 1998). Foraging strategies of incubating and brooding king penguins Aptenodytes patagonicus. Oecologia 114, 194–201. , , , , , & (
- 2002). Seasonal changes in the diving parameters of king penguins (Aptenodytes patagonicus). Mar. Biol. 141, 581–589. , & (
- 2010). Buoyancy under control: underwater locomotor performance in a deep diving seabird suggests respiratory strategies for reducing foraging effort. PLoS ONE 5, e9839. doi: 10.1371/journal.pone.0009839 , , , & (
- 1994). External devices on penguins – how important is shape. Mar. Biol. 118, 353–357. , & (
- 1996). Diving energetics in king penguins (Aptenodytes patagonicus). J. Exp. Biol. 199, 973–983. , , , , , & (
- 2004). Long-term trends in the population size of king penguins at Crozet archipelago: environmental variability and density dependence? Polar Biol. 27, 793–800. , & (
- 2007). To what extent might N-2 limit dive performance in king penguins? J. Exp. Biol. 210, 3344–3355. , , , & (
- 2004). Heart rate and energetics of free-ranging king penguins (Aptenodytes patagonicus). J. Exp. Biol. 207, 3917–3926. , , , , , & (
- 2003). Energetics of diving in macaroni penguins. J. Exp. Biol. 206, 43–57. , , & (
- 2007a). A thorough and quantified method for classifying seabird diving behaviour. Polar Biol. 30, 991–1004. , & (
- 2007b). Fine-scale analyses of diving energetics in king penguins Aptenodytes patagonicus : how behaviour affects costs of a foraging dive. Mar. Ecol. Prog. Ser. 344, 299–309. , , , , , , & (
- 2008). To breathe or not to breathe: optimal strategies for finding prey in a dark, three-dimensional environment. J. Anim. Ecol. 77, 847–849. (
- 1992). The optimal allocation of time during the diving cycle. Behav. Ecol. 3, 255–265. & (
- 1989). Diverse divers: physiology and behavior. Berlin: Springer-Verlag. (
- 2004). Stroke patterns and regulation of swim speed and energy cost in free-ranging Brunnich's guillemots. J. Exp. Biol. 207, 4679–4695. , , , & (
- 2002). An application of optimal diving models to diving behaviour of Brunnich's guillemots. Anim. Behav. 64, 739–745. , , & (
- 1979). On the theory of central place foraging. In Analysis of ecological systems: 155–177. Horn, D.J. , Mitchell, R.D. & Stairs, G.R. (Eds). Columbus, OH: Ohio State University Press. & (
- 1992). Land-based predator impact on vertically migrating zooplankton and micronekton advected to a southern-ocean archipelago. Mar. Ecol. Prog. Ser. 80, 15–27. & (
- 2001). Time/depth usage of Adelie penguins: an approach based on dive angles. Polar Biol. 24, 467–470. , , , , & (
- 2002). Swim speed of free-ranging Adélie penguins Pygoscelis adeliae and its relation to the maximum depth of dives. J. Avian Biol. 33, 94–99. , , , , , & (
- 2006). Foraging strategies and prey encounter rate of free-ranging Little Penguins. Mar. Biol. 149, 139–148. , , & (
- 2009). New frontiers in biologging science. Biol. Lett. 5, 289–292. & (
- 2002). Buoyancy and maximal diving depth in penguins: do they control inhaling air volume? J. Exp. Biol. 205, 1189–1197. , , , , , , , & (
- 2003). Factors affecting stroking patterns and body angle in diving Weddell seals under natural conditions. J. Exp. Biol. 206, 1461–1470. , , , & (
- 2004). Why do macaroni penguins choose shallow body angles that result in longer descent and ascent durations? J. Exp. Biol. 207, 4057–4065. , , & (
- 2006). Etude de la thermorégulation en mer chez le manchot royal : mécanismes et conséquences énergétiques. PhD thesis, Strasbourg University. (
- 1972). Locomotion – energy cost of swimming, flying, and running. Science 177, 222–228. (
- 1986). Foraging theory. Princeton, NJ: Princeton University Press. & (
- 2001). How long should a dive last? A simple model of foraging decisions by breath-hold divers in a patchy environment. Anim. Behav. 61, 287–296. & (
- 2003). Stroke and glide of wing-propelled divers: deep diving seabirds adjust surge frequency to buoyancy change with depth. Proc. R. Soc. Lond., B, Biol. Sci. 270, 483–488. , , , & (
- 2005). Regulation of stroke and glide in a foot-propelled avian diver. J. Exp. Biol. 208, 2207–2216. , , , , , & (
- 2006). Swim speeds and stroke patterns in wing-propelled divers: a comparison among alcids and a penguin. J. Exp. Biol. 209, 1217–1230. , , , , , & (
- 1995). Foraging ecology. In The penguins: 81–106. Williams, T.D. (Ed.). Oxford: Oxford University Press. (
- 2003). Penguins predict their performance. Mar. Ecol. Prog. Ser. 249, 305–310. (
- 2003). Up-beat motion in swinging limbs: new insights into assessing movement in free-living aquatic vertebrates. Mar. Biol. 142, 537–547. & (
- 1996). Diving behaviour of Gentoo penguins, Pygoscelis papua; factors keeping dive profiles in shape. Mar. Biol. 126, 153–162. , , & (
- 2010). Pedalling downhill and freewheeling up; a penguin perspective on foraging. Aquat. Biol. 8, 193–202. , , , & (
- Top of page
- Supporting Information
Appendix SI. Effects of maximum dive depth and number of wiggles on vertical speed, swimming speed, body angle and flipper stroke frequency during descent and ascent phases: statistical values.
Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.