1For air-breathing animals in aquatic environments, foraging behaviours are often constrained by physiological capability. The development of oxygen stores and the rate at which these stores are used determine juvenile diving and foraging potential.
2We examined the ontogeny of dive physiology in the threatened Australian sea lion Neophoca cinerea. Australian sea lions exploit benthic habitats; adult females demonstrate high field metabolic rates (FMR), maximize time spent near the benthos, and regularly exceed their calculated aerobic dive limit (cADL). Given larger animals have disproportionately greater diving capabilities; we wanted to determine the extent physiological development constrained diving and foraging in young sea lions.
3Ten different mother/pup pairs were measured at three developmental stages (6, 15 and 23 months) at Seal Bay Conservation Park, Kangaroo Island, South Australia. Hematocrit (Hct), haemoglobin (Hb) and plasma volume were analyzed to calculate blood O2 stores and myoglobin was measured to determine muscle O2. Additionally, FMR's for nine of the juveniles were derived from doubly-labelled water measurements.
4Australian sea lions have the slowest documented O2 store development among diving mammals. Although weaning typically occurs by 17·6 months, 23-month juveniles had only developed 68% of adult blood O2. Muscle O2 was the slowest to develop and was 60% of adult values at 23 months.
5We divided available O2 stores (37·11 ± 1·49 mL O2 kg−1) by at-sea FMR (15·78 ± 1·29 mL O2 min−1 kg−1) to determine a cADL of 2·33 ± 0·24 min for juvenile Australian sea lions. Like adults, young sea lions regularly exceeded cADL's with 67·8 ± 2·8% of dives over theoretical limits and a mean dive duration to cADL ratio of 1·23 ± 0·10.
6Both dive depth and duration appear impacted by the slow development of oxygen stores. For species that operate close to, or indeed above their estimated physiological maximum, the capacity to increase dive depth, duration or foraging effort would be limited. Due to reduced access to benthic habitat and restricted behavioural options, young benthic foragers, such as Australian sea lions, would be particularly vulnerable to resource limitation.
Australian sea lions Neophoca cinerea (Péron) are excellent subjects to study otariid diving ontogeny as they are non-migratory and demonstrate extended dependency, during which pups begin diving. Pups are suckled for 17·6 ± 0·1 months, one of the longest lactation periods in pinnipeds thought to have evolved as an adaptation to a marine environment where resources are limited and show little seasonal fluctuation (Higgins 1993; Gales, Shaughnessy & Dennis 1994). As a result of small population size, small breeding colony size, exposure to human activities and evidence of population declines, Australian sea lions have recently been listed as threatened (EPBC Act 2000).
The Australian sea lion provides an intriguing system as one of few species in which adults regularly exceed cADL's, with almost 80% of dives over predicted limits (Costa et al. 2001). Adults spend 58% of time at-sea underwater and exhibit high field metabolic rates (FMR) (Costa & Gales 2003). Given the extreme foraging behaviour of adults and the potentially limited capabilities of younger animals, we wanted to examine diving ability in Australian sea lion pups and juveniles.
We investigated the ontogeny of O2 stores (hematocrit (Hct), haemoglobin (Hb), plasma volume, Mb), FMR and cADL in Australian sea lions. Although many studies have looked at one or more aspects of cADL development in marine mammals, this is the first to simultaneously measure blood and muscle O2 stores with FMR. By examining the extent physiology limits dive behaviour in young Australian sea lions, we can answer a central question in physiological ecology for this species, provide insight into its threatened status, and contribute to the emerging field of conservation physiology (Wikelski & Cooke 2006).
Materials and methods
Fieldwork was conducted between June 2001 and August 2003 at Seal Bay Conservation Park, Kangaroo Island, South Australia (35°41′S, 136°53′E). A known-aged cohort of 55 pups (28 males and 27 females) was flipper-tagged in 2001 (Fowler et al. 2006). To ensure individuals were only measured once, mothers and pups received a subcutaneous passive microtransponder chip (Destron Fearing Corporation, South St Paul, MN, USA).
Mother/pup pairs were captured simultaneously, sedated with Isoflurane gas anaesthesia, and weighed with a digital scale (± 0·1 kg) (Gales & Mattlin 1998). Pairs were captured during three field seasons: (i) 6-month pups (March 2002); (ii) 15-month pups (November 2002); and (iii) 23-month juveniles (July 2003). Ten different mother/pup pairs were captured each season, with the exception of two 15-month pups and five 23-month juveniles, which were never observed suckling and were captured alone. Adult females suckling young pups were captured in place of their mothers. The remaining 23-month juveniles were observed suckling at least once during the field season, despite the fact weaning usually occurs by 17·6 months (Higgins 1993). In July 2003, only six 23-month juveniles (not sampled in March or November 2002) could be located. Therefore, age for the remaining juveniles was estimated using pelage condition and growth curves constructed from data on mass and standard length (Fowler 2005). One independent juvenile from the previous cohort (aged c. 3 years) was also captured and sampled in July 2003.
Hematocrit (Hct) declines as the spleen expands under general anaesthesia, so we took initial blood samples using manual restraint (Zapol et al. 1989; Ponganis et al. 1992; Costa, Gales & Crocker 1998). We measured Hct in quadruplicate the same day of collection in capillary tubes following centrifugation for 5 min at 11 500 r.p.m. For individuals that could not be sampled using manual restraint (one 6-month pup, one 15-month pup, two 23-month juveniles), we used a minimum of three sampling points and linear regression to hindcast Hct before administration of anaesthesia.
To determine whole-blood haemoglobin concentration (Hb), 10 µL whole blood were added to 2·5 mL Drabkins solution (Kit 525A, Sigma Diagnostics, St Louis, MO, USA) and later assayed in duplicate using the cyan-methhaemoglobin photometric method (ICSH 1967). Samples were read at 540 nm (Spectronic 1001, Bausch & Lomb, Rochester, NY, USA) and Hb was determined by comparison with standard dilution curves. Following methods for Hct, linear regression was used to hindcast Hb when necessary. We determined mean corpuscular haemoglobin content (MCHC) using the equation:
MCHC = (Hb × 100) × Hct − 1.
Plasma volume was determined using Evans Blue dilution (ICSH 1967). A 10-mL blood sample was drawn from the caudal gluteal vein, followed by an intravenous injection of pre-weighed Evans Blue dye (Sigma Diagnostics) approximating a dosage of 0·6 mg kg−1 (Costa et al. 1998). The syringe was flushed with blood to ensure injections were intravenous and all dye was administered. Two to three serial samples followed at 10 min intervals.
Blood samples were kept on ice until centrifuged the same day for 10 min at 3400 revs min−1. Plasma was kept frozen for a maximum of 3 months, when samples were thawed and centrifuged for 10 min. Plasma optical densities were determined at 624 and 740 nm following El-Sayed, Goodall & Hainsworth (1995), with modifications by Foldager & Blomqvist (1991). Adjusted absorbances were logarithmically transformed and linear regression used to determine dye concentration at time of injection. If the regression was not significant or the line's slope was positive (two 6-month pups and three 23-month juveniles), adjusted absorbance values were averaged (Jørgensen et al. 2001; Arnould et al. 2003). Plasma volume was calculated as distributional volume of injected dye (El-Sayed et al. 1995).
Biopsies were collected from the dorsal triceps and pectoralis complex locomotor muscles to analyze Mb. Additionally, we obtained muscle opportunistically from 10 fresh carcasses, ranging in age from 1 week to adult. Samples were frozen at −80 °C until analyses. We determined Mb following methodology detailed in Reynafarje (1963), as modified by Castellini & Somero (1981). Buffer blanks and elephant seal muscle of known Mb were used as assay controls.
where Vb is blood volume, 0·33 is the percentage arterial blood, 0·66 is the percentage venous blood, (capacitance coefficient of O2 in blood) = Hb × 1·34 mL O2, and is O2 saturation of venous blood. An O2 carrying capacity of 1·34 mL O2 (g−1) was assumed (Kooyman 1989). We also assumed 75% of arterial blood O2 was available during a dive (with 15% used to maintain vital body and brain functions; 95% O2 saturation to 20%: Ponganis et al. 1997c; Costa et al. 2001) and mixed venous blood had an O2 content 5% by volume less than initial (Ponganis et al. 1993), so = [( – 50) ()−1] × 100 (Davis & Kanatous 1999).
Muscle O2 stores were calculated using the equation:
where 0·3 is the fraction of muscle mass in the body (Kooyman et al. 1983). This is also identical to the fraction of muscle mass found from complete dissection of 1-month Steller sea lion Eumetopias jubatus pups (Richmond et al. 2006). For two 23-month-old juveniles for which muscle biopsies were not available, mean Mb determined for the age class was used to calculate muscle O2.
Following Costa et al. (2001), lung O2 stores were derived from allometric estimates of lung volume for otariids:
Concurrent measurements of CO2 production and diving behaviour were carried out on eight of the 23-month Australian sea lions and the one 3-year-old to determine at-sea metabolism using oxygen-18 doubly-labelled water (Lifson & McClintock 1966; Nagy & Costa 1980; Speakman 1997). For comparison with published adult FMR, methodologies were identical to Costa & Gales (2003).
Pre-injection blood samples were taken to determine isotope background values, followed by intraperitoneal injections of 60–80 mL 15% oxygen-18 water (H218O) and 18·5 MBq/mL tritiated water (HTO) in 3 mL sterile saline. Syringes were weighed ( ± 0·001 g) before and after injections to determine masses injected. After 3 h equilibration, body mass was recorded and a 10 mL blood sample collected to determine isotope concentrations at the start of the experimental period. Juveniles were equipped with Wildlife Computers (Redmond, WA, USA) time/depth recorders (TDR's) and VHF radio transmitters (Sirtrack Ltd, Havelock, New Zealand) (Fowler et al. 2006). Juveniles were recaptured after 5–8 days to record body mass, collect final blood samples, and recover TDR's.
Tritium specific activity was determined by scintillation spectrometry (Tri-Carb 2100TR, Packard, Canberra, ACT, Australia) of duplicate aliquots 0·2 mL pure water (distilled from plasma samples) in 10 mL scintillation fluid (Ultima Gold scintillation fluid, Packard Bio Science, Meriden, CT, USA). Specific activity of H218O was determined by mass ratio spectro-metry (Metabolic Solutions, Nashua, NH, USA).
Initial dilutions of H218O were used to determine total body water (TBW) (Nagy & Costa 1980). Final TBW was calculated by the equation from initial TBW corrected for change in mass. We calculated lean body mass from TBW, assuming a hydration constant of 74·2% reported for California sea lions Zalophus californianus (Oftedal, Iverson & Boness 1987), and calculated CO2 production using Speakman's (1997) two-pool model to correct for errors associated with isotope fractionation. Water influx was calculated using equations (5) and (6) in Nagy & Costa (1980), assuming an exponentially changing body water pool.
As Australian sea lions’ diet is not well-known and six juveniles were observed suckling at least once during the field season, we followed calculations for Steller sea lion juveniles (Richmond et al. 2006) and used a respiratory quotient (RQ) of 0·76 (19·3 kJ L−1 O2). This assumes a 50 : 50 lipid : protein fuel source intermediate between nursing pups’ lipid-rich diet and foraging adult's protein-rich diet (Schmidt-Nielsen 1997; Iverson, Frost & Lang 2002). Assuming a diet of 100% lipid or 100% protein alters the RQ by less than 5%. We divided CO2 production by RQ to determine O2 consumption.
Data collected over measurement intervals included variable amounts of onshore FMR and were normalized to estimate metabolism at-sea. Percentages of time spent at-sea were calculated from TDR data and following methods from Costa & Gales (2003) for adult Australian sea lions, we plotted FMR data (containing both at-sea and onshore components) as a function of percentage time spent at-sea. Least squares linear regression was used to predict FMR for each animal at their respective percentage time spent at-sea (Costa & Gales 2003). The difference (residual) between predicted and actual FMR was added to extrapolated FMR where the animal spent 0% time at-sea to determine onshore metabolism. At-sea FMR was calculated from the equation:
FMR = at-sea FMR (% time at-sea) + onshore FMR (% time ashore).
Dive behaviour data recorded from these juveniles and reported in Fowler et al. (2006) were used to calculate percentages of dives over cADL and ratios of mean dive duration to cADL.
For comparison with at-sea FMR, we substituted estimates of otariid DMR from the literature and recalculated cADL's for juvenile Australian sea lions. Hastie, Rosen & Trites (2006) used metabolic rates of Steller sea lion females trained to dive to depth in the open ocean to construct a model predicting O2 consumption. Although some studies have shown during prolonged breath-holds adult pinnipeds may lower metabolism to resting (Hurley & Costa 2001; Sparling & Fedak 2004), young pinnipeds have less metabolic control and juvenile Weddell seals Leptonychotes weddellii appear unable to do so; cADL's based on resting metabolism (RMR) overestimated ADL's indicated by changes in LA by 60% (Rea & Costa 1992; Ponganis et al. 1993; Burns & Castellini 1996). We therefore chose 2 × RMR (determined by multiplying BMR by age-specific scaling factors estimated for Steller sea lions: Winship, Trites & Rosen 2002) to represent minimum cost of transport (Feldkamp 1987; Costa 1991; Arnould & Boyd 1996). Finally, based on the only direct measurements of ADL in an otariid, we used a value of 17·8 mL O2 min−1 kg−1 determined by measuring post-submersion blood LA in similarly-sized (41·4 kg) juvenile California sea lions (Ponganis et al. 1997c).
Using the different estimates of DMR, surface metabolic rates (MR) were then calculated from:
When statistical differences were determined by one-way analysis of variance, post-hoc comparisons were made using Tukey tests. If transforming data did not achieve normality and equal variances, differences were determined by Kruskal–Wallis one-way analysis of variance on ranks and Dunn's post-hoc test was used.
Australian sea lions had fully developed adult Hct and Hb by 15 months (Table 1). Six-month pups were the only age class with significantly lower Hct and Hb (Hct: H3 = 28·85, P < 0·001; Hb: H3 = 24·21, P < 0·001). Values for adult females agree closely with published values for this species (Costa et al. 2001). As there were no significant differences between sexes within age classes, data were combined (t-test: t7 = 0·11, P = 0·92). There were no significant differences between age classes for MCHC (F3,46 = 2·19, P = 0·10), which remained relatively constant throughout development and was 34·6 ± 0·5 g/dL at 6 months, 35·4 ± 0·9 g/dL at 15 months, 38·5 ± 0·6 g/dL at 23 months and 35·8 ± 1·1 g/dL for adult females.
Table 1. Summary of O2 storage parameters for different age classes of Australian sea lions (mean ± SE); * = values significantly different from adult. Ranges of ages are reported in parentheses following mean ages and n is given in parentheses below other values. Although hematocrit and haemoglobin were fully developed by 15 months, plasma volume, blood volume, and muscle myoglobin were slower to develop
Haemoglobin (g dL−1)
Plasma volume (L)
Plasma volume (mL kg−1)
Blood volume (L)
Blood volume (mL kg−1)
6·1 ± 0·2 (5·4–7·1)
30·0 ± 1·7* (10)
39·3 ± 1·0* (10)
13·6 ± 0·5* (10)
1·6 ± 0·1* (7)
52·4 ± 5·4* (7)
2·6 ± 0·2* (7)
83·5 ± 7·5* (7)
0·8 ± 0·2* (4)
14·5 ± 0·2 (13·4–15·7)
44·5 ± 2·0* (10)
51·4 ± 0·7 (10)
18·2 ± 0·4 (10)
2·0 ± 0·2* (10)
45·6 ± 3·4* (10)
4·2 ± 0·4* (10)
93·8 ± 7·1* (10)
1·3 ± 0·1* (5)
22·6 ± 0·2 (22·1–22·9)
48·3 ± 2·6* (9)
49·5 ± 1·0 (9)
19·0 ± 0·4 (9)
2·8 ± 0·3* (8)
60·1 ± 4·1 (8)
5·7 ± 0·6* (8)
120·9 ± 8·6* (8)
1·6 ± 0·2* (6)
88·2 ± 2·1 (21)
51·7 ± 0·5 (21)
18·6 ± 0·6 (21)
6·7 ± 1·3 (2)
83·7 ± 12·5 (2)
14·2 ± 3·0 (2)
178·3 ± 30·9 (2)
2·7 ± 0·1 (3)
Plasma and blood volumes were slower to develop (Table 1). Well beyond the age of average weaning, 23-month juveniles demonstrated mass-specific blood volumes only 68% of adult blood volumes. Pups had significantly lower mass-specific plasma and blood volumes than adult females (plasma volume: F3,20 = 6·89, P = 0·002; blood volume: F3,20 = 10·03, P < 0·001) and although plasma volumes for 23-month juveniles were lower than adult values, this was not significant (P = 0·08). Plasma volume was 10% of lean body mass in both juvenile and adult Australian sea lions.
Muscle Mb increased linearly with age (r2 = 0·82, P < 0·001) and mass (r2 = 0·80, P < 0·001). However, development was comparatively slow: 23-month juveniles only developed Mb to 60% of adult capacities (Table 1). Muscle Mb was significantly different across age classes (F4,16 = 49·63, P < 0·001), with the exception of values between 15 and 23 months (P = 0·07). The measured Mb for adult females (2·7 ± 0·1 g/100 g wet tissue) agrees closely with the published value (2·8 g%) and the difference is within the assay's resolution (Costa et al. 2001).
total oxygen stores
Mass-specific blood, muscle and total O2 stores were significantly higher in older animals (blood: F3,20 = 19·28, P < 0·001; muscle: F3,14 = 58·46, P < 0·001; total: F3,27 = 56·73, P < 0·001) and had not reached adult capacities by 23 months (Fig. 1). Even by 3 years, mass-specific blood O2 was only 29·0 mL O2 kg−1, muscle O2 was 8·7 mL O2 kg−1 and total stores were 43·8 mL O2 kg−1 (78% of adult total stores). Total available mass-specific O2 stores increased significantly with age (r2 = 0·66, P < 0·001) and mass (r2 = 0·70, P < 0·001).
Mean and maximum dive depth (Fig. 2a,b), and mean (r2 = 0·51, P < 0·001), and maximum dive duration (r2 = 0·56, P < 0·001) increased with increasing O2 stores (Fowler et al. 2006).
field metabolic rates (fmr)
Mean FMR for 23-month juveniles was 15·44 ± 1·30 mL O2 min−1 kg−1 (see Appendix for FMR data). Field metabolism for the 3-year-old was 18·25 mL O2 min−1 kg−1. Mass-specific FMR's were slightly higher for juveniles than values calculated for adult Australian sea lions using Speakman's (1997) two-pool model (Costa & Gales 2003), although this was not significant (t27 = 0·26, P = 0·80). At-sea FMR was calculated to be 15·78 ± 1·29 mL O2 min−1 kg−1 at 23 months and 18·41 mL O2 min−1 kg−1 at 3 years.
calculated aerobic dive limit (cadl)
By dividing total available O2 by at-sea FMR, we found juvenile Australian sea lions developed a cADL of 2·33 ± 0·24 min by 23 months (Table 2). Like adults, juveniles exceeded cADL's on the majority of dives (Table 2). Furthermore, there were no significant differences between ratios of mean dive duration to cADL for adults and juveniles (t17 = –1·29, P = 0·90), indicating both groups exceeded cADL's to similar extents.
Table 2. Calculated aerobic dive limits (cADL) for Australian sea lions (mean ± SE). Values for adult females from Costa et al. (2001); * = values significantly different from adult. Masses are presented as means from initial captures and recaptures. At-sea field metabolic rates (FMR) were calculated using equations from Speakman (1997) and corrected for percentage of time spent onshore. Juvenile total available O2 stores were significantly lower than adult (t19=−7·72, P < 0·001), but juveniles exceeded cADL's as often (t16 = 1·81, P = 0·09)
Mean mass (kg)
Total available O2 (mL)
At-sea FMR (mL O2 min−1)
Mean dive duration (min)
Dives > cADL (%)
Mean dive duration (cADL)−1
22·6 ± 0·1
47·7 ± 2·7*
1668·8 ± 126·2*
748·81 ± 66·67*
2·33 ± 0·24
2·75 ± 0·16*
67·8 ± 2·8
1·23 ± 0·10
77·8 ± 3·8
3656·0 ± 179·0
1589·91 ± 101·53
2·34 ± 0·11
3·14 ± 0·16
79·4 ± 4·8
1·38 ± 0·11
There were no significant differences between DMR's determined from FMR, at-sea FMR, predicted O2 consumption, 2 × RMR, and post-submersion LA (F4,38 = 1·41, P = 0·25: Table 3). The mass-specific value predicted for Steller sea lions diving to 50 m was remarkably similar to FMR and at-sea FMR measurements for juvenile Australian sea lions diving to depths of 44 ± 4 m (Fowler et al. 2006; Hastie et al. 2006; Table 3). There were also no significant differences between percentages of dives over cADL (F4,33 = 0·44, P = 0·78), ratios of mean dive duration to cADL (F4,38 = 1·16, P = 0·35), or surface MR (F4,30 = 0·10, P = 0·98; Table 3).
Table 3. Summary of different diving metabolic rates (DMR) used to calculate aerobic dive limits (cADL) in juvenile Australian sea lions (n= 8, mean ± SE). Field metabolism (FMR) was determined using doubly-labelled water (Lifson & McClintock 1966); at-sea FMR (bold) was corrected for percentage of time spent onshore. For comparison, predicted O2 consumption for Steller sea lions (SSL) diving to 50 m (Hastie et al. 2006), 2 × resting metabolism (RMR) representing minimum cost of transport, and DMR determined by measuring post-submersion blood lactate (LA) in similarly-sized juvenile California sea lions (CSL) (Ponganis et al. 1997c) are presented. There were no significant differences between alternate cADL's (H4 = 3·22, P = 0·52)
Diving MR (mL O2 min−1 kg−1)
Dives > cADL (%)
Mean dive duration (cADL)−1
Surface MR (mL O2 min−1 kg−1)
15·44 ± 1·30
2·39 ± 0·25
66·7 ± 2·8
1·21 ± 0·10
16·54 ± 1·27
15·78 ± 1·29
2·33 ± 0·24
67·8 ± 2·8
1·23 ± 0·10
15·78 ± 1·29
2·30 ± 0·11
67·7 ± 4·4
1·25 ± 0·09
16·25 ± 4·29
2 × RMR
17·53 ± 0·21
2·02 ± 0·11
71·4 ± 3·8
1·42 ± 0·11
11·45 ± 4·39
CSL post-submersion LA
1·99 ± 0·11
71·5 ± 3·8
1·44 ± 0·11
10·97 ± 4·42
There were increases in minimum post-dive surface intervals (PDSI) as Australian sea lion juveniles approached cADL's (Fig. 3a,b), with minimum PDSI's 2·2 times longer for dives over cADL's than dives under (Fowler et al. 2006).
Juvenile Australian sea lions well beyond the age of average weaning had not developed adult O2 stores. Juveniles were also not reaching adult dive depths or durations, suggesting they are limited by physiological capability. Like adults, juveniles appeared to be operating close to, or above their estimated physiological maximum, with the majority of dives over cADL's. These factors suggest juvenile Australian sea lions have limited ability to compensate for environmental fluctuations, restricted potential foraging habitat and restricted behavioural options.
blood oxygen stores
Hematocrit (Hct) and Hb increased while MCHC remained constant, suggesting although red blood cells mature early, young pups have fewer cells. However, Hct and Hb were fully developed by 15 months, so lower blood O2 stores in older pups resulted from differences in mass-specific blood volumes (Table 1). Age-specific differences in plasma and blood volumes may be largely due to changes in body composition and hydration. We found, like harbour seals Phoca vitulina and Steller sea lions, although mass-specific plasma and blood volumes fluctuated throughout development, ratios of plasma volume to lean body mass remained relatively constant (Burns et al. 2005; Richmond et al. 2006).
muscle oxygen stores
Increases in Mb and muscle O2 stores lagged behind development of blood O2 stores in Australian sea lions. In previous studies on diving vertebrates, Mb had also not reached adult levels by onset of independent foraging (Weber, Hemmingsen & Johansen 1974; Thorson & LeBoeuf 1994; Noren et al. 2001). As Mb is the slowest O2 storage parameter to mature in species studied to date, its development appears to be a major limiting factor on diving ability in juveniles.
Because Australian sea lions have to meet physiological demands of benthic diving at an early age, they may deal with prolonged development in muscle by storing proportionately more O2 in blood and less in muscle than other pinnipeds (Fowler et al. 2006). On average, adult otariids store 51% of total O2 in blood and 33% in muscle (Kooyman 1989; Richmond et al. 2006) and phocids store 65% in blood and 25% in muscle (Kooyman 1989; Burns et al. 2005; Noren, Iverson & Boness 2005). Australian sea lion adults store 70% in blood and 20% in muscle (Costa et al. 2001), which more closely resembles stores of deeper diving phocids than otariids. In Australian sea lion juveniles, blood O2 stores accounted for 1·53 ± 0·20 min (62·2%) of cADL's and muscle O2 stores accounted for 0·47 ± 0·06 min (19·7%).
total oxygen stores
Protracted blood volume and Mb maturation, as well as low body mass contributed to low total available O2 stores in pups and juveniles. Fowler et al. (2006) reported young Australian sea lions dived benthically like adults, but did not achieve adult dive depths or durations. By 23 months, dive behaviour and dive physiology developed to similar percentages of adult capacity: maximum depth was 76% adult depth, maximum duration was 77% adult duration, and total mass-specific O2 stores were 70% adult stores (Fowler et al. 2006). The positive relationship between ontogeny of O2 stores and ontogeny of dive performance suggests slow physiological development may constrain foraging capacities (Fig. 2).
Australian sea lions demonstrated the slowest development of total O2 stores documented to date in mammalian divers and only developed 55% of adult female mass-specific total O2 before average weaning. In comparison, 9-month Steller sea lions develop 81% (Richmond et al. 2006), 12-week northern elephant seals Mirounga angustirostris develop 74% (Thorson & LeBoeuf 1994), 41-day grey seals Halichoerus grypus develop 67% (Noren et al. 2005) and Weddell seals aged 41–50 days develop 64% (Burns & Castellini 1996) prior to onset of independent foraging. Although exact ages were not available, juvenile New Zealand sea lions Phocarctos hookeri developed an estimated 87% (Costa et al. 1998) and juvenile California sea lions developed 59% (Kuhn et al. 2006).
Galapagos fur seals Arctocephalus galapagoensis are the only pinniped with a typical lactation interval longer than Australian sea lions (Trillmich 1986). Horning & Trillmich (1997) found similar development of Hct and Hb in Galapagos fur seals, but did not collect data on blood volume or Mb. However, in the divers studied, species developing total O2 stores earlier tend to wean at younger ages (Arnould et al. 2003).
This is the second study to calculate at-sea FMR or DMR for an immature pinniped. While we estimated DMR for juvenile Australian sea lions (n = 8) to be 4·03 times BMR, Lydersen & Hammill (1993) used a value of 5·88 times BMR for ringed seal P. hispida pups (n = 3). This difference cannot be accounted for by differences in body mass (18·6 ± 0·3 vs 47·7 ± 2·7 kg), as basal metabolism was scaled to mass0·75 in both studies. However, ringed seals were studied at a younger age and their higher DMR may reflect immature metabolic control, faster growth rates and/or increased thermoregulatory demands in their arctic environment (Kleiber 1975; Rea & Costa 1992; Lydersen & Hammill 1993).
calculated aerobic dive limit (cadl)
Given juvenile Australian sea lions were found to exceed theoretical aerobic limits on 67·8% of dives, there are three possibilities: (i) cADL's were miscalculated; (ii) Australian sea lions are highly adapted to deal with LA; or (iii) this species is operating at or near its physiological limit in a marginal foraging environment (Costa & Gales 2003; Chilvers et al. 2006).
(i) cADL's were miscalculated
To accurately determine cADL's, available O2 stores and O2 use underwater must be known. Although body mass, Hb, Hct, blood volume and Mb were measured, muscle mass and respiratory volumes were estimated from the literature (Kooyman et al. 1971, 1983; Richmond et al. 2006). We assumed 75% of arterial O2 and 100% of muscle O2 was available during a dive. Our calculations should be conservative and over-estimate ADL as Davis & Kanatous (1999) reported only 63% of total O2 stores may be accessible and not all muscle O2 can be utilized aerobically.
Our approach is supported by an earlier study on California sea lions that measured increases in LA in freely-diving trained animals (Ponganis et al. 1997c). Researchers found a close agreement between measured ADL's and cADL's based on estimates of total O2 stores similar to presented here (Ponganis et al. 1997c). Furthermore, FMR's derived from doubly-labelled water measurements on wild adult females (Costa, Antonelis & DeLong 1990), such as presented here, gave more accurate predictions of ADL than calculations using metabolic rates estimated from flumes.
However, we compared at-sea FMR to a number of different estimates for otariids and found no differences between DMR's and cADL's (Table 3). Furthermore, DMR estimates from the literature resulted in slightly lower cADL's, with juveniles consequently exceeding cADL more often and to a greater extent, so calculations using at-sea FMR appear conservative. We therefore feel until direct measurements can be made, at-sea FMR provides a useful DMR estimate for juvenile Australian sea lions. Additionally, at-sea FMR was used to determine cADL's in adult Australian sea lions, as well as other adult pinnipeds, allowing for internally consistent comparisons (Costa et al. 2001, 2004).
(ii) Australian sea lions are highly adapted to deal with LA
Weddell seals can remain submerged for durations three times their ADL, indicating sufficient O2 left to supply the nervous system and, to at least some extent, locomotor muscles (Kooyman et al. 1980; Butler 2004). Likewise, muscle has probably begun proportional energy production by anaerobic glycolysis at the point of net production of LA (Kooyman 2006). These factors complicate assumptions about O2 store management which must be made to calculate cADL. Without measurements of post-dive LA in additional species, or a better understanding of how divers manage and use O2, it is impossible to draw definitive conclusions about LA levels solely from cADL.
There were increases in minimum PDSI's as Australian sea lion juveniles approached maximum dive durations (Fig. 3a, b), which may indicate LA production. However, there was no evidence of extended PDSI's or cumulative effects predicted to clear accumulated LA (G.L. Kooyman & T.G. Kooyman 1995). Lack of extended surface recovery has also been documented in a number of adult penguins and pinnipeds found to dive beyond their cADL, including king penguins Aptenodytes patagonicus (Handrich et al. 1997), northern elephant seals (Le Boeuf et al. 1989), and New Zealand (Crocker et al. 2000; Chilvers et al. 2006) and Australian sea lions (Costa & Gales 2003). Although these data indicate some species use more aerobic capacity than others, none of these studies were able to measure changes in LA. Results highlight the potentially problematic nature of assuming aerobic dives if PDSI's are not protracted and further research into whether or how these animals exceed their ADL is required. Until direct measurements of post-dive LA are taken, other indirect measurements of effort, such as heart rate or acceleration, would be instructive (Butler et al. 1992, 1995; Sato et al. 2002).
If Australian sea lions were relying on anaerobic metabolism for the majority of dives, they must be able to buffer, recycle and/or burn LA. As juveniles only exceeded cADL's by a mean factor of 1·23, this would lead to a minor metabolic imbalance of acidosis. Using the Weddell seal as a model, exceeding its 20 min ADL by a factor of 1·23 is equivalent to a 25 min dive resulting in a blood LA concentration of c. 2 µmol and requiring less than 10 min surface recovery (Kooyman et al. 1980). Likewise, resting LA turnover rates (26·2 µmol min−1 kg−1) measured in the more similarly-sized harbour seal (40·3 kg) would require 9 min recovery (Davis et al. 1991). Yet, only 1·1% of PDSI's reached 9 min for juvenile Australian sea lions. However, relative differences were similar: for 25 min dives, Weddell seals require PDSI's two to three times longer than aerobic dives, while minimum PDSI's for Australian sea lions were 2·2 times longer for dives over cADL's than dives under.
Alternatively, juvenile Australian sea lions could use accumulated LA as substrate on following dives (Castellini et al. 1988). One molecule of ATP is produced for each molecule of LA through glycolysis, yet 17 molecules of ATP are produced by subsequent oxidation of one molecule of LA (Fedak & Thompson 1993). Therefore, in the unlikely event juveniles were producing energy exclusively anaerobically during the 0·42 min (25 s) beyond cADL's, it would take 7·10 min (0·42 × 17) of aerobic diving using LA as substrate until blood LA returned to normal (Butler 2004). However, harbour seals swimming at 35% VO2max increase LA turnover to 39·7 µmol min−1 kg−1, which would decrease recovery time for a blood LA concentration of 2 µmol to 5 min; swimming at 50% VO2max would require a 4 min recovery (Davis et al. 1991). With mean dive durations of 2·75 min, juveniles would have to perform 1–3 dives within the cADL following every dive over (Fowler et al. 2006). As 67·8% of dives were over, juvenile Australian sea lions may use additional or alternative methods to deal with LA.
(iii) This species is operating at or near its physiological limit in a marginal foraging environment
Regardless of potential problems, cADL can be a useful conceptual tool for determining foraging effort and efficiency in relation to environmental quality, as well as providing comparisons between different ages, sexes and species (Costa et al. 2004; Kooyman 2006). Animals operating well within their physiological capacity are able to draw on reserves to pursue prey deeper, dive longer or extend foraging bouts if necessary (Boyd et al. 1994). Both Australian sea lion adults and juveniles appear to operate at or near their physiological maximum in their present habitat, so the capacity to increase dive depth, duration, or foraging effort would be limited (Costa et al. 2001). Additionally, young sea lions do not reach adult dive depths or durations (Fowler et al. 2006). Restricted behavioural options would make juveniles even more susceptible to shifts in prey availability.
The ability to exploit different dive depths is important in determining potential foraging habitat, especially for benthic divers such as Australian sea lions (Costa & Gales 2003; Fowler et al. 2006). One limitation of benthic foraging is restricted available habitat due to the finite extent of continental shelves. This limitation is even more severe for pups that cannot or do not reach adult depths. Although juvenile Australian sea lions exceeded cADL's on the majority of dives and to similar extents as adult Australian sea lions, New Zealand sea lions and Australian fur seals A. pusillus doriferus, they did not achieve depths typical of other benthic foragers (Costa et al. 2004; Fig. 4).
Relative to many other species, which exceed theoretical ADL's on 4%–10% of dives, Australian sea lions use more metabolic scope to forage (Gentry, Kooyman & Goebel 1986; Feldkamp, Long & Antonelis 1989; Boyd & Croxall 1996). Additionally, Australian sea lions demonstrate the slowest development of O2 stores during one of the longest lactation periods, strongly supporting the idea that rates of O2 development are related to investment duration in pinnipeds (Arnould et al. 2003). With restricted access to benthic habitat and limited behavioural options, juvenile Australian sea lions appear particularly susceptible to resource limitation. Shallower dive depths would also restrict juveniles to near-shore waters, which would increase interaction risks with fisheries. These results have potential implications for the conservation and management of other species, especially those that forage benthically, such as New Zealand, Steller, and southern sea lions Otaria flavescens and Australian fur seals (Werner & Campagna 1995; Merrick & Loughlin 1997; Costa et al. 2004), as well as for benthically foraging seabirds such as penguins and cormorants (Croxall et al. 1991; Bost, Puetz & Lage 1994; Gremillet et al. 1998; Tremblay & Cherel 2000). Although development of oxygen stores have not been well studied in these species, ontogeny may constrain diving behaviour and limit potential foraging habitat, making young divers more susceptible to resource limitation.
Successful juvenile recruitment is critical for a long-lived species with a low reproductive rate, small total population and small breeding colony size (Environmental Protection and Biodiversity Conservation Act 1999). The recent listing of Australian sea lions as threatened was in response to poor population performance across their range, in stark contrast to the rapidly recovering sympatric New Zealand fur seal A. forsteri (Gales, Haberley & Collins 2000). Lack of physiological reserves for juvenile Australian sea lions may be a major limitation on population potential. Our results suggest slow development of oxygen stores limits diving ability in juvenile Australian sea lions, and similar to adults, juveniles are working at or near their physiological capacity.
This work was supported by UC MEXUS, NSF AAAS WISC, JEB, ONR, Wildlife Computers, National Geographic, Myers Oceanographic and Marine Biology Trust, AMNH Lerner Grey, ACS, Friends of LML, Project AWARE, Sigma Xi, Sealink, Clairol, South Australia National Parks and Wildlife. The following people provided invaluable field assistance: Seal Bay Conservation Park staff, D. Higgins, N. Rourke, D. Needham, Melbourne Zoo (S. Blanchard, G. McDonald), Z. Boland, C. Farber, J. Gibbens, H. Mostman, S. Sataar, S. Simmons, Y. Tremblay, M. Weise. S. Shaffer, S. Villegas, M. Weise provided assistance with laboratory analyses. J. Estes, H. Fowler, K. J. Fowler, G. Kooyman, T. Williams, Costa lab group provided helpful comments. Research was carried out under South Australian Department for Environment and Heritage permit #G24475–2, Wildlife Ethics Committee permit #4/2001. Research protocols were approved by Prevention of Cruelty to Animals Act 1985 (Australia), Chancellor's Animal Research Committee (UCSC: #Cost01·01).
Table Appendix. Summary of FMR data for Australian sea lions (mean ± SE). Equations from Nagy & Costa (1980) were used to calculate total body water (TBW) determined by 18O and water influx. We used the two-pool model presented in Speakman (1997) to calculate CO2 production