Biology of Marine Birds - Chapter 13 pptx

Metabolized energy can be allocated to the following components of the chicks’ energy budget:1 resting metabolism at thermoneutrality i.e., the energy required for maintaining some basal

Development in Seabirds

G Henk Visser

CONTENTS

13.1 Introduction 439

13.2 Growth Patterns of Seabird Chicks in Relation to Taxon and Parental Feeding Strategy 442

13.2.1 Interspecific Variation in Growth Rates 442

13.2.2 Intraspecific Variation in Growth Rates 443

13.3 Energetics of Growth 444

13.3.1 Introduction 444

13.3.2 Components of the Chicks’ Energy Budget 444

13.3.3 Methods to Determine Energy Budgets in Free-Living Chicks 446

13.3.3.1 Periodic Chick Weighing 446

13.3.3.2 The Time-Energy Budget 447

13.3.3.3 The Measurement of Water Influx Rates and Subsequent Conversion to Energy Intake 450

13.3.3.3.1 The Doubly Labeled Water Method: Some General Principles 450

13.3.3.3.2 Applications of the DLW Method in Adult Seabirds: The Need for Standardization 451

13.3.3.3.3 Applications of the DLW Method in Seabird Chicks 452

13.3.4 Energy Budgets of Growing Seabird Chicks: The Importance of Asymptotic Body Mass, Duration of the Nestling Period, and Latitude 452

13.4 Development of Temperature Regulation 455

13.5 Physiological Effects of Food Restriction 457

13.6 Toward the Construction of Energy Budgets of Entire Family Units during the Peak Demand of the Brood 458

Acknowledgments 459

Literature Cited 459

13.1 INTRODUCTION

Chicks of most seabird species grow up on land situated in close proximity to the sea It is presumed that the nature of their food supply has not allowed the evolution of the self-feeding precocial mode

of development in seabirds (Lack 1968) Although there are marked interspecific differences with respect to developmental mode, in the majority of seabird species, chicks stay in or close to their nest until fledging, being parentally fed and brooded For example, chicks of pelicans, frigatebirds, gannets, and boobies are born naked with their eyes closed, being totally dependent on parental food and warmth (Figure 13.1) Chicks that hatch in this developmental state have been classified 13

as being altricial by Nice (1962; see Table 13.1) Chicks of tropicbirds (Figure 13.2) hatch withtheir eyes closed, but are covered in down (being classified as being semialtricial-2; Nice 1962),whereas tern, auk, murre, and jaeger chicks hatch with a downy plumage with their eyes open, andare able to walk (semiprecocial: Figure 13.3) In contrast, chicks of some murrelet species (Synth- liboramphus spp and Brachyrhamphus spp.) leave the nest shortly after hatching, being fed at sea

by their parent (precocial-4; Nice 1962, Eppley 1984, Gaston 1992, Starck and Ricklefs 1998a)

Chicks of Common Murre (Uria aalge), Thick-billed Murre (Uria lomvia), and Razor-billed Auk (Alca torda) do so after having attained about 25% of adult body mass (Daan and Tinbergen 1979,

Gaston 1985, Starck and Ricklefs 1998a) Obviously, early nest desertion by the chick potentiallyreduces parental traveling time and enables exploitation of remote feeding areas (Ydenberg 1989).However, this strategy can only be achieved with the co-evolution of some specific physiologicaladaptations of the chick to minimize and compensate for its heat loss (e.g., Eppley 1984)

FIGURE 13.1 A newly hatched altricial Lesser Frigatebird chick (Fregata ariel) (Photo by R W and E A.

Parental Brooding

Parental Feeding

Precocial-1 Contour feathers Open Leave nest None None

From Nice 1962.

Seabirds have developed different feeding strategies, ranging from in-shore feeding to off-shorefeeding (see Chapter 6) In species that feed in-shore (e.g., some pelicans, cormorants, gulls, andsome terns), the chick can be fed several times a day and one parent can remain at the nest tobrood it This foraging mode may not necessitate chicks developing homeothermy at an early age(Klaassen 1994) Small seabird chicks frequently receive food as whole particles, such as fish (e.g.,

in most tern species), or as a predigested food mash (e.g., in young chicks of the Black-legged

Kittiwake [Rissa tridactyla], cormorants, boobies) In pelagic seabirds (e.g., albatrosses, petrels,

fulmars, boobies, and some terns), however, due to the long travel distances to their food source,parents are often gone for one or more days on a foraging trip Therefore, until the achievement

of homeothermy by the chick, feeding rates of the chick may be somewhat reduced over those ofnear-shore feeding birds After the chick(s) achievement of homeothermy, both parents can leavethe nest, potentially resulting in a doubling of the amount of food brought to the brood (Ricklefsand Roby 1983) While foraging, most pelagic seabirds store the food in their stomach to carry it

back to the colony, although a few species carry fish in the bill (e.g., White Terns, Gygis alba;

puffins) Procellariiform birds are unique in the sense that parents partly concentrate the food caught

FIGURE 13.2 A 1-day-old Red-tailed Tropicbird chick They are the only Pelecaniform chicks to hatch with

a full coat of down (Photo by E A Schreiber.)

FIGURE 13.3 A Sooty Tern chick hatches with its eyes open and able to walk The down on this just-hatched

chick has not dried out yet (Photo by E A Schreiber.)

into stomach oil This substance mainly consists of wax esters with a very high energy density,and stomach oil formation has been interpreted as a strategy to increase the amount of energy perfeeding trip (Ricklefs et al 1985, Roby 1991, Roby et al 1997) This physiological developmenthas enabled procellariiform birds to exploit more remote feeding areas.

Recent reviews in the literature include growth patterns of birds in general (Starck and Ricklefs1998b), developmental plasticity (Schew and Ricklefs 1998), energy budgets during growth (Weath-ers 1992, 1996), and development of temperature regulation (Visser 1998) It is the aim of thischapter to partly update the information presented in these reviews with special emphasis on seabirdchicks In addition, current knowledge of several aspects of postnatal development on seabird chicks

is integrated into the text to provide insight into patterns of evolutionary and geographical sification Some recent methodological developments are evaluated in an attempt to provide guide-lines for the standardization of future work on the construction of energy budgets in seabird chicksand adults

diver-13.2 GROWTH PATTERNS OF SEABIRD CHICKS IN RELATION TO

TAXON AND PARENTAL FEEDING STRATEGY

13.2.1 I NTERSPECIFIC V ARIATION IN G ROWTH R ATES

Starck and Ricklefs (1998b) present growth parameter estimates based on logistic growth curvesfor a mixed data set of altricial and precocial land and seabird species (n = 557 species) Thelogistic growth curve, which has a sigmoid shape, enables the description of the development ofbody masses (M, g) of chicks as a function of age (t, d):

where A represents the asymptotic body mass of chicks (g), ti the time of inflection point of thecurve (d), and kl the logistic growth rate constant (d–1) For the seabird subset (Figure 13.4),

FIGURE 13.4 Relationship between the logistic growth constant (kl , d –1 ) and asymptotic body mass (A, g)

in seabird chicks Drawn diagonal line represents the general relationship between A and kl in birds (Equation 13.2 of this chapter, Starck and Ricklefs 1998b).

asymptotic body masses range between 38.5 g for the Least Tern (Sterna antillarum, Schew 1990) and 15,500 g for the Emperor Penguin (Aptenodytes forsteri, Stonehouse 1953), and the logistic

growth constants range between 0.019 d–1 for the Amsterdam Albatross (Diomedea amsterdamensis,

Jouventin et al 1989) and 0.38 d–1 for the fastest growing species, the Black Tern (Chlidonias niger, Schew 1990) Assuming independence of the data points for all 557 species, the relationship

between both parameters was described by:

(Figure 13.4; Starck and Ricklefs 1998b) Equation 13.2 was used to predict kl for each seabirdspecies listed by Starck and Ricklefs (1998b) Next, its residual value was calculated with thegeneral equation:

Residual value = 100 · (observed value – predicted value)/predicted value (13.3)The analysis revealed that, after correction for differences in asymptotic body masses, mostseabird families exhibit relatively low growth rate constants, which is particularly the case for theFregatidae (average level relative to prediction: –58.0%, which differs significantly from zero, aftercomparison of the standard error of the residuals, t3 = –14.4; p <0.001), Hydrobatidae (–56.9%; t4

= –19.0; p <0.0001), Diomedeidae (–28.3%; t9 = –2.81; p <0.02), Phaethontidae (–4.2%; t2 = –6.1;

p <0.03), and Procellariidae (–20.9%; t17 = –3.0; p <0.01) In the Spheniscidae, average relative

growth rates were significantly above prediction (+32%; t10 = 2.5; p <0.03); and in the Sulidae

(+2.9%; t4 = 0.18; p = 0.86), Alcidae (+6.9%; t11 = 1.0; p = 0.33), and Laridae (+8.2%; t31 = 1.4;

p = 0.16), residual values were higher than prediction, but these differences were not significant.

These values indicate that growth rates are particularly low for many pelagic seabird species, andtend to be higher in species that feed in-shore such as the Laridae Highest relative growth ratesare observed in the Spheniscidae This has been interpreted to be an adaptation to the short Antarcticsummer enabling the chicks to leave the colony before the onset of the winter (Volkman andTrivelpiece 1980) In Section 13.3.4 we explore the energetic consequences of variations in growthrate for free-living chicks and their parents

13.2.2 I NTRASPECIFIC V ARIATION IN G ROWTH R ATES

Chicks of most seabird species are generally fed a high-quality diet rich in protein and energy,with the possible exception of chicks of some petrel and albatross species that are mainly fed squidwith a relatively low energy content (Prince and Ricketts 1981, see also Section 13.3.2) However,the quantity of food delivered by parents can be less predictable (see Chapters 6 and 7) In somespecies there is tremendous variation in postnatal growth rates owing to changes in the availablefood supply For example, residual values for the logistic growth constants of Wedge-tailed Shear-

waters (Puffinus pacificus) range between –72 and –9% (n = 13 studies), Black-legged Kittiwakes between –6 and +52%, Common Terns (Sterna hirundo) between –27 and + 74% (n = 21 studies), and Atlantic Puffins (Fratercula arctica) between –45 and +22% (n = 27 studies) In addition, these

four species exhibit marked intraspecific differences in calculated asymptotic body mass valueswhich vary between 424 and 750 g, 335 and 421 g, 100 and 133 g, and 265 and 400 g, respectively.These large differences may reflect differences in body condition of the fledglings at this stage.There is considerable evidence that growth retardation results from reduced food availability.For example, in a year with poor food availability, parental foraging trips of the Magellanic Penguin

(Spheniscus magellanicus) lasted 20% longer than normal, chick feeding rates were reduced, and

average body mass in 5-day-old chicks was 30% lower than in years with normal food availability(Boersma et al 1990) Consequently, there were large differences between years with respect tothe number of chicks fledged per nest, and in a 5-year study values ranged from 0.02 chicks per

nest to 0.60 (Boersma et al 1990) In some cases, variation in food availability is related to ElNiño–Southern Oscillation (ENSO) events (e.g., Schreiber and Schreiber 1993, Schreiber 1994).

In other cases, intraspecific variations in chick growth and nesting success have been attributed toindividual differences in timing of egg laying within a particular year (Brooke 1986, Catry et al.1998), age and breeding experience of the parents (Brooke 1986, Coulson and Porter 1985),individual quality, genetically determined (Brooke 1986), weather conditions (especially windspeed; Konarzewski and Taylor 1989), differences in food availability between colonies within aseason (Frank 1992, Frank and Becker 1992), and between years (Boersma et al 1990, Danchin

1992, Crawford and Dyer 1995) One of the key parameters for the interpretation of inter- andintraspecific variations in growth rates in seabird chicks seems to be the amount of energy collected

by parents during chick rearing per unit of energy spent (see Section 13.6)

On the longer term, food restrictions and the subsequent growth retardation can potentiallyresult in reduced survival (e.g., in the Common Murre, Harris et al 1992), tolerance to starvation

(e.g., in the Lesser Black-backed Gull [Larus fuscus], Griffiths 1992), and reduced recruitment rate

(e.g., in the Black-legged Kittiwake, Coulson and Porter 1985)

13.3 ENERGETICS OF GROWTH

13.3.1 I NTRODUCTION

One of the key factors needed for interpreting seabird life histories is the construction of energybudgets of free-living chicks and their parents (Drent et al 1992) It is assumed that during theevolution of avian life histories, chicks have developed an array of adaptive responses, for instance:

1 In single-chick broods, the total amount of food required until independence (TME, kJ).However, in multi-chick broods, sibling competition may select for rapid growth andactive food solicitation, potentially resulting in an increase of the TME

2 Peak level of daily metabolized energy (peak-DME, kJ d–1)

3 Duration of the growth period (tfl, d) in order to reduce the risk of predation (Lack 1968),and (in polar environments) to complete the reproductive cycle before the onset of winter(Obst and Nagy 1993)

It has to be noted that minimizing the duration of the development period may require increasinggrowth rate and therefore daily energy requirement (Weathers 1992, and Section 13.3.4) Theseadaptations may enable parents to maximize their lifetime reproductive output

13.3.2 C OMPONENTS OF THE C HICKS ’ E NERGY B UDGET

Of all energy ingested by a chick (gross energy intake: GEI, kJ d–1) only a part can be metabolized(metabolizable energy intake: MEI, kJ d–1); the remainder is excreted in the form of feces and urine(FU, kJ d–1) The assimilation coefficient (Q, dimensionless) is defined as:

Davis et al 1989), 2.9 to 4.9 kJ g–1 for zooplankton (Clarke and Prince 1980, Montevecchi et al.

1984, Simons and Whittow 1984, Clarke et al 1985), 4.2 to 10.3 kJ g–1 for fish (depending on itsfat content; Montevecchi et al 1984), and 39 to 41.7 kJ g–1 for the oil component of procellariiformdiets (Warham et al 1976, Simons and Whittow 1984, Obst and Nagy 1993) Some seabird speciesare known for having highly specialized diets (e.g., feeding exclusively on fish [terns] or krill [somepenguins]), whereas other species (like most Procellariiformes) exhibit a nonspecialized aquaticdiet (squid and other zooplankton, krill, fish, and trawler offal)

Assimilation coefficients have been determined in seabird chicks of several species and fordifferent diets (Table 13.2) Average values for fish, krill, and zooplankton diets are 0.80 (SD =0.035, n = 10), 0.75 (SD = 0.014, n = 2), and 0.75 (SD = 0.040, n = 3), respectively It is interesting

to note that the measured values in chicks are in close agreement with those reported for adultbirds fed fish or invertebrates (0.77 and 0.74, respectively; Castro et al 1989), which suggests thatdigestion efficiency in seabird chicks is high Little information is available on the development ofthe assimilation efficiency as a function of the chicks’ ages An increase in assimilation coefficientfrom 0.8 at 11 to 12 days of age to a value of 0.88 at 20 to 21 days of age was reported in Double-

crested Cormorant (Hypoleucos auritus) chicks (Dunn 1975) A similar trend with age was found

in Jackass Penguin (Spheniscus demersus) chicks (Heath and Randall 1985), but not in chicks of the Cape Gannet (Morus capensis, Cooper 1978) and Common and Sandwich Tern (Sterna sand- vicensis, Klaassen et al 1992).

TABLE 13.2

Assimilation Coefficients for Seabird Chicks in Relation to Food Type

Spheniscidae

Jackass Penguin (S demersus) Zooplankton 0.71 Heath and Randall 1985

Procellariidae

White-chinned Petrel (Procellaria aequinoctialis) Fish 0.78 Jackson 1986

White-chinned Petrel (P aequinoctialis) Zooplankton 0.74 Jackson 1986

White-chinned Petrel (P aequinoctialis) Krill 0.76 Jackson 1986

Sulidae

Phalacrocoracidae

Double-crested Cormorant (Hypoleucos auritus) Fish 0.85 Dunn 1975

Laridae

Black-legged Kittiwake (Rissa tridactyla) Fish 0.80 Gabrielsen et al 1992

Alcidae

Note: In chicks of some diving petrels, prions, and storm petrels, the digestion efficiencies of the dietary wax

component was near 0.99 (Roby et al 1986).

It appears as if there are marked differences between species with respect to assimilationefficiency as a function of diet type For example, assimilation coefficients in White-chinned Petrel

(Procellaria aequinoctialis) chicks were relatively insensitive to diet type, and values ranged from

0.74 for a squid diet to 0.78 for a fish diet (Jackson 1986) In contrast, assimilation coefficients ofJackass Penguin chicks varied from 0.68 and 0.87 for these two diets (Heath and Randall 1985).The high flexibility of the digestive system of White-chinned Petrel chicks is interpreted to be anadaptation to their nonspecialized diets (squid, krill, fish, and trawler offal; Jackson 1986; see also

Brown 1988)

Metabolized energy can be allocated to the following components of the chicks’ energy budget:(1) resting metabolism at thermoneutrality (i.e., the energy required for maintaining some basalphysiological functions within the chicks’ body; RMR, units kJ d–1), (2) heat increment of feeding(i.e., the energy required to warm and digest the food; HIF [also referred to as specific dynamicaction of food; SDA], units kJ d–1), (3) temperature regulation (to compensate for heat losses fromthe chick to its environment: TR, units kJ d–1), (4) activity (e.g., walking, preening, calling, andbegging; A, units kJ d–1), (5) biosynthesis-related heat production (the energy required for synthe-sizing new tissue such as fat and protein; S, units kJ d–1), and (6) tissue energy (energy deposited

as protein and fat; TE, units kJ d–1):

At a given level of metabolizable energy intake of a chick (e.g., the maximum level that can beprovided by the parents), growth is highest at low levels of energy expenditure Growth is zero ifMEI equals energy expenditure, and growth is negative if MEI is lower than the level of energyexpenditure Under the latter conditions, body tissue (e.g., fat or protein) is used to produce energyfor supporting other physiological functions

13.3.3 M ETHODS TO D ETERMINE E NERGY B UDGETS IN F REE -L IVING C HICKS

Four different methods have been used to determine the chick’s level of MEI under free-livingconditions:

1 Determination of gross energy intake based on periodic weighing of the chick in thefield (e.g., Prince and Walton 1984, Ricklefs et al 1985, Obst and Nagy 1993)

2 Determination of energy expenditure of the chick based on the extrapolation of laboratorymeasurements to field conditions, with an added component for energy deposited intissues (see Ricklefs and White 1981)

3 Measurement of water influx rates and subsequent conversion to gross energy intake(Gabrielsen et al 1992, Konarzewski et al 1993)

4 Measurement of the level of energy expenditure directly in the field with doubly labeledwater method, with an added component for growth energy (see Klaassen et al 1989,

Klaassen 1994, Visser and Schekkerman 1999)

13.3.3.1 Periodic Chick Weighing

A method used frequently to assess levels of food intake in seabird chicks is based on periodicchick weighing (expressed in grams per unit of time; e.g., Ricklefs 1984, Ricklefs et al 1985,Schreiber 1994, 1996, Philips and Hamer 2000) Chicks are weighed regularly (e.g., at 2- to 12-hintervals) to monitor changes in their body mass It is assumed that body mass decreases with time,

a process that can be approached mathematically (e.g., by taking initial body mass, age, body sizeindex, and time into account; Philips and Hamer 2000) If the chick exhibits a positive change inbody mass, it is assumed that it was fed exactly between two weighings The food intake level at

the assumed feeding time is calculated as the difference between backward and forward lation of the mass loss curves of a recently fed and fasting chick, respectively The value obtainedrepresents the amount of food eaten by the chick (in grams per unit of time) Next, to convert thisvalue to gross energy intake (GEI), an assumption must be made with respect to the mass-specificenergy content of the food (see Section 13.3.2) Finally, MEI can be calculated on the basis ofEquation 13.5, after assuming a specific value for the assimilation coefficient of the diet (see Section13.3.2 and Table 13.2).

extrapo-Although this method is very easy to apply, it can only be used in chicks that are fed mealsthat are heavy relative to their body mass In addition, apart from weighing and extrapolation errors,the calculated MEI level is subject to several other accumulating methodological errors The firstpotential error is caused by the uncertainty with respect to the exact feeding time This error islarger if weighings are done with a lower frequency However, a high weighing frequency may, insome cases, interfere with chick begging, or parental feeding behavior, although some species arenot bothered by it In some sedentary seabird chicks, this problem can be circumvented by contin-uous weighing on an electronic balance, as employed in albatross chicks (Prince and Walton 1984,Huin et al 2000)

The second potential error relates to the conversion of mass change to GEI This error isprobably smallest in species with a specialized diet, facilitating accurate estimation of the mass-specific energy content of the diet The error is probably largest in procellariiform chicks because

of the large difference in energy density of separate components of their diet, which ranges fromabout 4 kJ g–1 for predigested food to about 40 kJ g–1 for the stomach-oil component (see Section13.3.2) Another complication is the large variation in the relative quantity of the oil componentbetween meals within a species (e.g., values determined for different birds from the same colony

during one observation day ranged between 20 and 83% in Wilson’s Storm-Petrel [Oceanites oceanicus]; Obst and Nagy 1993), between species (e.g., see Roby 1991), and the relative difficulty

to estimate the fraction of the oil component in a chick’s diet (Roby et al 1997)

The third potential error of this method is the conversion of GEI to MEI (Equation 13.5), afterassuming a specific value for the assimilation coefficient As discussed in Section 13.3.2, theseaverage values range from 0.77 for krill and zooplankton to about 0.99 for stomach oil (Roby et

al 1986) Because of its high mass-specific gross energy content and its high assimilation cient, stomach oil is the most important component in energy budgets of procellariiform chicks,and it may contribute up to about 80% of their energy budgets (Roby 1991, Obst and Nagy 1993).The overall error of using the “chick weighing” method to estimate MEI can be as high as about25%, depending on the number of assumptions made (Weathers 1992, 1996)

coeffi-13.3.3.2 The Time-Energy Budget

The “time energy budget” method differs fundamentally from the “chick weighing” method in thesense that in the former method, metabolizable energy intake is estimated on the basis of measure-ments on energy expenditure (the components RMR, HIF, TR, A, and S; Equation 13.6) with anadded component of the tissue energy (TE; Equation 13.6) As a first step, levels of oxygenconsumption (and carbon dioxide production) are determined in resting chicks while housed in asmall respiration chamber (indirect calorimetry; Weathers 1996) Next, metabolic rate (MR) can

be calculated after assuming a specific energy equivalent per unit oxygen consumed or carbondioxide produced Typically, levels of energy expenditure are determined at different ambienttemperatures to reveal the lowest level of energy expenditure at thermoneutrality (RMR), and thethermoregulatory costs (TR) at temperatures below the lower critical temperature (LCT, units °C)

At each temperature, thermal conductance can be calculated being the metabolic level per degreetemperature difference between the chick’s body and its environment (see Visser 1998) The thermalconductance is assumed to be minimal at ambient temperatures below LCT

To facilitate extrapolation of laboratory measurements to field conditions, the thermal ment of a chick must be characterized in its habitat (Bakken 1976, Klaassen 1994) This is mosteasily accomplished in chicks that live in deep burrows (e.g., procellariiform chicks, by measuringburrow-air temperatures), and it is most difficult in mobile chicks that live in sparsely vegetatedcolonies When fully exposed, a chick experiences cooling effects of wind, compensated for by thechick elevating its metabolism These effects can be strongly diminished by the chick positioningitself in vegetation (i.e., an energy-saving mechanism) In contrast, when fully exposed, a chickmay experience the heating effects of solar radiation (enabling the chick to reduce thermoregulatorycosts) Both effects can be integrated when employing heated taxidermic mounts, or (partly) withtemperature measurements using black spheres (Gabrielsen et al 1992, Klaassen 1994).

environ-To estimate the tissue energy component of the energy budget in relation to the chicks’ age, it

is necessary to determine their growth curve in the field (see Section 13.2.1), as well as their specific energy content of the body The latter component is often estimated with the generalequation:

mass-slope indicates that as birds get heavier, they have a higher energy density per gram of body mass)

As can be seen, seabirds exhibit large differences in developmental patterns, and energydensities are particularly high in some pelagic seabird species (Ricklefs et al 1980, Obst and Nagy

1993, Ricklefs and Schew 1994) Therefore, the use of group-specific estimates of the regressions

to estimate energy accumulation is suggested, instead of the use of Weathers’ (1996) Equation13.10 for birds in general To estimate the biosynthesis-related heat production in chicks, a synthesisefficiency value of 0.75 has traditionally been assumed (Ricklefs 1974), which has been used forthe construction of energy budgets of most species listed in Table 13.3 More recently, Weathers(1996) advocated the separation of tissue growth into a fat component (with high synthesis effi-ciency) and a protein component (with low synthesis efficiency) Although this approach is morecorrect, it appears as if little error is made when employing a value of 0.75 (Konarzewski 1994,Ricklefs et al 1998)

There are a number of potential methodological errors involved in this time energy budgetmethod that merit attention First, it is virtually impossible to account for the costs associated withlocomotion or activity of the chick (e.g., see Dunn 1980) Probably these costs are lowest inindividual procellariiform chicks that spend “about 90% of their time resting and sleeping” (Simonsand Whittow 1984, Brown 1988), but the costs can be much higher for chicks growing up in densecolonies where frequent social interactions occur (Figure 13.5) Second, the extrapolation oflaboratory measurements to field conditions for estimating the costs for temperature regulation isrelatively difficult, especially because of the difficulty in accounting quantitatively for energy-savingmechanisms such as huddling, sheltering, or exposure to solar radiation Third, poikilothermicchicks are frequently brooded by a parent, which considerably reduces its energy expenditure level(Klaassen 1994) Fourth, the extrapolation procedure is very sensitive to the assumed body tem-perature of the chick under field conditions Although chicks of most species tend to keep up abody temperature of about 40°C, chicks of some species enter into torpor periodically betweenfeeding bouts (Pettit et al 1982, Boersma 1986) Occurrence of torpor has been interpreted to be

an energy-saving mechanism to minimize costs of temperature regulation

Some of the aforementioned difficulties with the extrapolation of laboratory-based ments to field conditions also apply to adult birds However, the energetic importance of energy-saving mechanisms in chicks is magnified because (1) chicks exhibit a larger surface area per unitbody mass than adults, and (2) per unit of body surface, minimal thermal conductances are higher

measure-in chicks than measure-in adults (Visser 1998) The average error of the time-energy budget method hasbeen estimated to be on the order of 25% (Weathers 1992)

TABLE 13.3

Development of the Mass-Specific Energy Content of Chicks during Growth

of Some Seabird Species

Family/Species

Intercept (a)

Slope

Spheniscidae

Gentoo Penguin (Pygoscelis papua) 3.26 7.57 Myrcha and Kaminski 1982

Chinstrap Penguin (P antarctica) 3.18 7.08 Myrcha and Kaminski 1982

Sandwich Tern (S sandvicensis) 3.97 4.68 Drent et al 1992

Black-legged Kittiwake (Rissa tridactyla) 4.24 4.51 Gabrielsen et al 1992

Alcidae

Note: Model used E = a + b · (M/A), where E represents the mass-specific energy density of the body

(kJ g –1 ), M the actual body mass of the chick (g), and A the asymptotic or fledging body mass (g).

FIGURE 13.5 In a colony of Sooty Terns, nests are very close together Here adult and large chicks are

constantly interacting, possibly raising the energy budget of birds in these colonies (Photo by R W Schreiber.)

13.3.3.3 The Measurement of Water Influx Rates and Subsequent Conversion

to Energy Intake

Apart from metabolic water formation, food is the only other water source for most seabird chicks(Gabrielsen et al 1992, Konarzewski et al 1993) Water fluxes can be measured in free-livingchicks following labeling with a heavy hydrogen isotope (i.e., water enriched with respect to 2H

or3H) Its concentration decreases in relation to the water-influx level (Nagy and Costa 1980) andcan be measured on the basis of small blood samples (see Section 13.3.3.4) Food-water intake can

be calculated after correction for metabolic water formation (Gabrielsen et al 1992, Konarzewski

et al 1993) Finally, the amount of food eaten (in g) can be calculated after assuming a specificwater percentage in the chicks’ diet This method has been validated in chicks of the Black-leggedKittiwake (Gabrielsen et al 1992) by comparing the quantity of gross energy provided to chickswith concomitant determination of water influx with stable isotopes On average, the validationrevealed an excellent agreement of both methods (the isotope method tended to underestimate thetrue GEI level by only 2%) However, determinations on individual birds could differ by 34% atmaximum, which suggests that a considerable sample size is required for the construction of energybudgets The most critical steps of this method seem to be the estimation of the amount of metabolic-water formation, and the conversion of water fluxes to levels of food intake, the latter conversionbeing very sensitive to the hydration level of the food (plus attached water) The method cannot

be applied in chicks that have access to other sources of water

13.3.3.3.1 The Doubly Labeled Water Method: Some General Principles

The doubly labeled water (DLW) method has frequently been used to measure the rate of energyexpenditure of adult and young seabirds under free-living conditions Its principle is based onlabeling the bird’s body water pool with the heavy isotopes 2H and 18O, and the subsequentdetermination of their fractional turnover rates (kd and ko, respectively, units d–1; Lifson andMcClintock 1966, Nagy 1980, Speakman 1997, Visser et al 2000a) In the past, the radioactiveisotope 3H has also been used instead of the stable 2H isotope, but due to permit restrictions,nowadays most DLW experiments on free-living seabirds are performed with 2H It is assumedthat, following labeling, 2H leaves from the body water pool as water only, and 18O both as waterand as carbon dioxide Thus, the difference between 18O and 2H can be converted to a rate of carbondioxide production (rCO2, moles d–1):

where N represents the size of the body-water pool (moles), which can be determined on the basis

of isotope dilution (Speakman 1997) The rate of carbon dioxide production can subsequently beconverted to a level of energy expenditure, after assuming a diet-specific energy equivalent ofcarbon dioxide For example, for seabird species with diets rich in proteins, it can be assumed thatthe production of 1 l of carbon-dioxide per hour is equivalent to a level of energy expenditure of27.3 kJ h–1 (see Gessaman and Nagy 1988)

Concentrations of the heavy isotopes 2H and 18O in the body-water pool of animals are oftendetermined on the basis of three blood samples, stored in flame-sealed capillaries or vacutainers.The first sample is taken prior to administration of the dose, to determine the natural abundance

of both heavy isotopes in the bird’s body-water pool (typically 0.015 and 0.20 atom percent for 2Hand 18O, respectively) The second sample is taken after equilibration of the pulse dose (oftenreferred to as “initial blood sample”), and the animal can be released It needs to be recapturedafter 24 to 72 h to take the third blood sample (often referred to as “final blood sample”) to determinethe isotope concentrations at the end of the measurement period Thus, the calculated rate of CO2production is related to the level of energy expenditure by the bird between the taking of the “initial”and “final” blood sample

Another important assumption of the DLW method for its application in seabirds is that theheavy 2H and 18O isotopes in the water molecule exhibit the same physical kinetics as the normal

1H and 16O isotopes (Lifson and McClintock 1966, Speakman 1997) This is true with respect to

fecal and urine water loss (Lifson and McClintock 1966, Visser et al 2000b), but not with respect

to evaporative water loss: water molecules with the heavy 2H or 18O are less likely to evaporatethan those with the lighter isotopes 1H or 16O This process is called fractionation As a result, theestimated fractional turnover rates of both isotopes (based on samples of the body-water pool) aretoo low Theoretically, this fractionation effect also affects the calculation of water fluxes based ondeterminations of 2H turnover (see Section 13.3.3.3), but an error analysis has revealed that amaximum error of about 1% is made if a fractional evaporative water loss value of 0.25 is used(Visser et al 2000b) In contrast, the DLW method to estimate the rate of CO2 production is muchmore sensitive to the effects of fractionation Because this fractionation effect is larger for the 2Hisotope than for the 18O isotope, this process also affects the difference between the 18O and 2Hturnover rate (i.e., the calculated level of CO2 production) Therefore, to solve this fractionationissue mathematically, the water efflux is considered to consist of one route subject to fractionation(i.e., evaporative water loss), and another route not subject to fractionation (i.e., fecal and urinewater loss) After taking this effect into account, in combination with fractionation of the 18O isotope

in the CO2 molecule (Speakman 1997), Equation 13.8 can be rewritten as:

rCO2 = N/2.078 · (ko– kd) – (rG· 0.0249 · N · kd) (13.9)where rG (dimensionless) represents the fraction of the water efflux lost through evaporativepathways

13.3.3.3.2 Applications of the DLW Method in Adult Seabirds: The Need for

Standardization

After having collected and analyzed blood samples to determine N, ko, and kd, we only need toestimate rG in order to calculate the rate of carbon-dioxide production (Equation 13.9) Afterstudying small mammals in the laboratory, Lifson and McClintock (1966) assumed that an rG value

of 0.5 was appropriate (i.e., 50% of the total water efflux is lost through evaporative pathways).This value has been used for many terrestrial and aquatic species (see Speakman 1997) However,for other seabird species, rates of carbon dioxide production were calculated with Equation 13.8,thereby assuming that fractionation did not play a role (e.g., Nagy 1980) Speakman (1997)recognized the fact that in many free-living species, the assumption of a fractional evaporativewater loss value of 0.5 was too high, and he proposed to use a value of 0.25 to be applied inEquation 13.9

To illustrate the importance of assumptions concerning fractional evaporative water loss inadult seabirds, the levels of energy expenditure in adult Black-legged Kittiwakes were calculated

at two different types of behavior In both cases fractional 2H and 18O turnover rates were measuredbased on small blood samples If the bird was foraging to feed its young, the rate of water effluxwas high (about 450 g d–1), and the calculated level of energy expenditure using Equation 13.8 was

1654 kJ d–1 However, if Equation 13.9 were applied with the same 2H and 18O turnover rates, butafter assuming that (1) fractional evaporative water loss was zero (rG = 0), (2) fractional evaporativewater loss was 0.25, and (3) fractional evaporative water loss was 0.5, calculated levels of energyexpenditure were 1584 kJ d–1, 1487 kJ d–1, and 1389 kJ d–1, respectively This indicates that at highwater fluxes, the DLW method is very sensitive to the assumptions made, and calculated rates ofenergy expenditure could differ by about 19% In contrast, during incubation, water efflux wasmuch lower (about 55 g d–1), and the maximum difference in calculated levels of energy expenditurewas only 9% This clearly indicates that the DLW method needs to be validated in adult seabirdsunder different feeding conditions Pending these results, the author proposes the use of Equation13.9 in adult seabirds, with a fractional evaporative water loss value of 0.25

13.3.3.3.3 Applications of the DLW Method in Seabird Chicks

Application of the DLW method in growing chicks has been hampered by uncertainties with respect

to the routes of 2H (or 3H) and 18O loss from the chick’s body-water pool Both isotopes may notleave from the body-water pool exclusively as water (2H) or as water and carbon-dioxide gas (18O),but they may also be incorporated in growing tissues (Williams and Nagy 1985, Weathers andSullivan 1991) Differential rates of isotope incorporation could potentially result in an underesti-mation of the true rate of CO2 production in the order of 10 to 25%

In only two seabird species, the Arctic Tern (Sterna paradisaea; Klaassen et al 1989) and

Kittiwake (Gabrielsen et al 1992), has the DLW method been validated to estimate the rate of CO2production in growing chicks After assuming a fractional evaporative water loss value of 0.5, itwas found that the DLW method underestimated the “true” rate of CO2 production by about 10%

on average (Klaassen et al 1989), which suggests that differential rates of isotope incorporationdid play a role However, at a later stage it was demonstrated that the error of the DLW methodcould be significantly reduced after assuming a fractional evaporative water loss value of 0.25(Visser and Schekkerman 1999) This value was also found to yield the best results in growing

shorebird chicks (i.e., Northern Lapwing [Vanellus vanellus] and Black-tailed Godwit [Limosa limosa], both belonging to the Charadriiformes; Visser and Schekkerman 1999) and Japanese Quail (Coturnix c japonica; Visser et al 2000a) Therefore, to calculate energy expenditure in growing

seabird chicks, the author recommends the use of Equation 13.9, with a fractional evaporative waterloss value of 0.25 Because meals are often rich in fat and proteins, an energy equivalent of 27.33

kJ should be used per liter CO2 produced (Gessaman and Nagy 1988) Unfortunately, in the past,little effort has been made to standardize this conversion, and in many studies a lower value hasbeen employed

As a last step in the construction of an energy budget, the level of metabolizable energy intake

of a chick is calculated from its level of energy expenditure (as measured with DLW) plus an addedcomponent of tissue growth (see above) Of all methods described here, the DLW method is assumed

to be the most accurate (Weathers 1992, 1996)

13.3.4 E NERGY B UDGETS OF G ROWING S EABIRD C HICKS : T HE I MPORTANCE

OF A SYMPTOTIC B ODY M ASS , D URATION OF THE N ESTLING P ERIOD ,

AND L ATITUDE

Weathers (1992, 1996) identified two important components of energy budgets of chicks: (1) thetotal amount of energy metabolized until fledging (TME, kJ) and (2) the peak level of dailymetabolized energy during the growth period (peak-DME, kJ d–1) His analysis for birds in generalrevealed significant effects of body mass at fledging (M, g) and the duration of the fledging period(tfl, d) on the TME level