Comparative Biochemistry and Physiology Part A 135(2003) 377–382

1095-6433/03/$ - see front matter� 2003 Elsevier Science Inc. All rights reserved.doi:10.1016/S1095-6433(03)00076-X

Energetic cost of a meal in a frequent feeding lizard

Sebastian Iglesias*, Michael B. Thompson, Frank Seebacher

School of Biological Sciences (A08) and Institute of Wildlife Research, University of Sydney, Sydney N.S.W. 2006, Australia

Received 13 September 2002; received in revised form 5 March 2003; accepted 6 March 2003

Abstract

Specific dynamic action(SDA) describes the rise in metabolism following feeding in animals and represents theenergetic cost of digesting and assimilating a meal. The overall energetic cost of feeding may depend on whether or notan animal is post-absorptive at the time of feeding. The aim of this study was to compare the energetic cost of SDAdue to feeding frequently compared with infrequently in the eastern water skink,Eulamprus quoyii. For similar quantitiesof food, repeated feeding incurred an energetic cost equal to 8.8% of the metabolizable energy of the meal(25 220 J),while single feeding incurred an energetic cost of 9.4% of the metabolizable energy of the meal(26 072 J). Experimentallizards maintained a rise in that was on average 1.8 times greater than the of the unfed controls over a 50-h˙ ˙VO VO2 2

interval as a result of feeding frequently. This prolonged rise in metabolism resulting from frequent feeding does notresult in a higher energetic cost of SDA compared with that resulting from infrequent single feeding.� 2003 Elsevier Science Inc. All rights reserved.

Keywords: Energy; Feeding cost; Specific dynamic action; Metabolic rate;Eulamprus quoyii; Assimilation; Calorimetric effect

1. Introduction

Energy must be harvested by an organism fromits environment to sustain life(Purves et al., 1992).Animals obtain energy by consuming food, butthere are energetic costs associated with feeding;the energy spent in procuring a meal and specificdynamic action(SDA), which is the energetic costassociated with digestion and assimilation. SDArefers to the rise in metabolic rate that occurs inall animals as a result of the consumption of ameal (Kleiber, 1961). SDA incorporates the costof digestion, absorption, and assimilation of foodas well as prey handling and gastrointestinal

*Corresponding author. Tel.:q61-2-9351-5610; fax:q61-2-9351-4119.

E-mail addresses:mrsiglesias24@hotmail.com(S. Iglesias),thommo@bio.usyd.edu.au(M.B. Thompson).

growth(Cruz-Neto et al., 2001; Secor, 2001). SDAcan be measured as the change in the rate ofoxygen consumption( ) of an organism afterVO2

feeding(Jobling, 1981; Wang et al., 2001). Typi-cally reaches a peak soon after feeding,VO2

depending on size and composition of the meal,before gradually decreasing to prefeeding values(Jobling 1981; Secor and Phillips, 1997). Theduration of SDA and the peak achieved helpVO2

to define the energetic cost of feeding and are,therefore, important ecological factors that can bequantified experimentally(Jobling, 1981).The largest factorial changes in metabolism

following feeding occur in large reptiles that feedinfrequently (Wang et al., 2001). For example,SDA in the python(Python molurus) elicits anincrease in that is greater than that resultingVO2

from muscular exercise(Secor and Diamond,1997).

378 S. Iglesias et al. / Comparative Biochemistry and Physiology Part A 135 (2003) 377–382

Comparative studies of SDA have used theresponse of fasted animals to a single meal todetermine the energetic cost of digestion andassimilation(Jobling, 1981; Janes and Chappell,1995; Sievert and Bailey, 2000). While this methodmay offer an accurate estimation of SDA foranimals that feed infrequently, such as large sit-and-wait predators, it may be less accurate foranimals that feed frequently. Most lizards, forexample, typically never have completely emptyguts when active(Halliday and Adler, 1987; Greer,1989) and may therefore experience SDA for longperiods. Hence, it was the aim of this study tocompare the SDA in a lizard resulting from fre-quent feeding to that resulting from a single meal.The concept of frequent and infrequent feeding

can be problematic(Secor, 2001). While an animalthat feeds weekly may be considered a frequentfeeder when compared with an animal that feedsmonthly, the same animal may be considered aninfrequent feeder compared with an animal thatfeeds daily. In this study the term frequent feeding,as applied to the feeding habit of lizards, describesfeeding while under the influence of an SDAelicited by a previous meal.The omnivorous lizardEulamprus quoyii (fam-

ily: Scincidae) was used as an experimental modelas it readily feeds in captivity(Weigel, 1988). Wemeasured the inE. quoyii after they consumedVO2

mealworms(Tenebrio molitor), either as a singlemeal or as several small meals, to determinewhether the rise in was sustained as a resultVO2

of feeding behaviour, and to enable the calculationof the relative contribution of SDA to the restingmetabolism ofE. quoyii.

2. Materials and methods

2.1. Animal maintenance and collection

Twenty-eight adultEulamprus quoyii were cap-tured by noosing(Simmons, 1987) in coastalbushland near Maroubra Beach, Sydney,(338589S,1518169E) in May, 2001. Lizards were housedindividually in 450 mm=350 mm glass aquaria ina temperature regulated room at 208C. A 40 Wlight bulb, placed at one end of each aquarium,created a temperature gradient, with 418C at thewarm end and 218C at the cool end. The heatinglights were on from 09.00 h to 18.30 h. Roomlights were set to a 12:12 photophaseyscotophasecycle (08.00 h to 20.00 h).

2.2. Assimilation efficiency of energy

Mealworms(100 g) were dried using a Dynavacfreeze drier and ground using a mortar and pestleprior to analysis of energy content using bombcalorimetry(Gorecki, 1975).Nine lizards (29.0"0.8 g, mean"S.E.) were

fasted for 4 days to ensure a post-absorptive stateand placed in glass aquaria within constant tem-perature cabinets maintained at 308C to equilibrateovernight. The lizards were fed six mealwormsmarked with indigestible nylon thread(1.29"0.001 g, mean"S.E.), and water was avail-able ad libitum. Lizards were observed regularlyover 3 days and all egested matter was collected.The gross energy of the egested matter was deter-mined using bomb calorimetry. Gross energyintake, assimilation rate and digestive efficiencyof energy were calculated using the equationsoutlined in Spencer et al.(1998). The term ‘diges-tive efficiency’ in this experiment is actuallyassimilation efficiency(AE) as the urinary prod-ucts were included with undigested residues in theanalysis(Spencer et al., 1998).

2.3. Respirometry

The rate of oxygen consumption( ) wasVO2

measured using flow-through respirometry(With-ers, 1977). Perspex tubes,(38 mm diameter, 270mm long) were used as respirometry chambers.Half of each chamber was covered with a card-board sheath to provide a darkened area, whichallowed lizards to settle better and significantlyreduced their activity while inside the chambers.The chambers were placed in an incubator at 308C. Lights were on for the duration of eachexperiment to minimise the effect of circadianrhythms on in these lizards(Robert andVO2

Thompson 2000; Iglesias, 2001), and to allow forvisual monitoring of activity via a fish-eye lens inthe incubator.The lizards were habituated to the chambers

prior to experimentation for 5 h every 4 days, for4 weeks. At the end of each 5-h session the lizardswere fed while inside the chambers and returnedto their enclosures. During the habituation periodthe time needed for the lizards to achieve a steady

after entering a chamber was determined toVO2

be 30–60 min. Animal handling was minimisedduring experimental procedures and lizards werefed inside the chambers.

379S. Iglesias et al. / Comparative Biochemistry and Physiology Part A 135 (2003) 377–382

Fig. 1. Mean (mlO h ) of Eulamprus quoyii after ay1VO2 2

single feeding of six mealworms(�; ns6) and in unfed con-trols (m; ns6). Error bars are"standard error.

Air was drawn through the chambers by a flowcontroller (17 ml min ) and passed throughy1

Drierite to absorb water and Carbabsorb to� �

absorb carbon dioxide and finally again throughDrierite . Oxygen concentration of the air was�

measured using a two-channel Ametek N-37Moxygen sensor and Ametek S-3Ay11 oxygenanalyser.Measurements of were made at 308C,VO2

which is within the range of normal activity bodytemperatures ofE. quoyii (preferred body temper-ature is 28.88C; Bennett and Alder, 1986), andwhich has been used in previous studies of SDAin lizards (Roberts, 1968; Robert and Thompson,2000; Secor, 2001).

2.4. Measuring SDA in response to a single meal

Twelve lizards weighing 19.2–34.3 g(26.9"4.6g, mean"S.D.) were fasted for 4 days to ensure apost-absorptive state. Two lizards in each of sixmeasurement periods(or experimental runs) wereplaced in individual respirometry chambers andallowed to settle for at least 1 h before the start ofrecordings. Each measurement period was startedat 1300 h on separate days. One of the two lizardsin each measurement period was fed six meal-worms (1.31"0.01 g, mean"S.E.) immediatelyafter an initial measurement of was madeVO2

( at 0 h after feeding or post-absorptiveVO2

metabolic rate). for each animal was meas-VO2

ured again at 5, 10, 15, 20, 25, 48 and 72 h afterfeeding. The other lizard in each measurementperiod served as an unfed control(ns6) whose

was measured at times corresponding to thoseVO2

of the fed animals.

2.5. Measuring SDA in response to repeatedfeeding

Sixteen lizards weighing 23.3–34.7 g(28.8"4.1 g) were fasted for 4 days. Four lizardsin each of four measurement periods were placedindividually into respiratory chambers, asdescribed for the single feeding experiment. Eachmeasurement period started at 13.00 h and lastedfor 60 h.Two lizards in each measurement period were

unfed controls(ns8). The other two lizards(ns8) were fed six mealworms initially and then threemealworms at 20 h and at 35 h after the initial

meal (total meal size was 2.59"0.03 g,mean"S.E.). The times at which lizards were fedwere chosen based on the timing of the decline inthe of lizards after a single feeding of sixVO2

mealworms. The of each lizard was recordedVO2

every 5 h during the 60-h interval.

2.6. Statistical analysis

Body mass of the lizards in each experimentwas compared using ANOVA prior to experimen-tation to determine if there was a significantdifference in lizard size between and within treat-ments. ANCOVA with repeated measures wasperformed to compare the variation in of theVO2

experimental and control groups over time in boththe single feeding and frequent feeding experi-ments. Mean of experimental groups wereVO2

compared with the control groups at equivalenttimes with ANCOVA. Whenever ANCOVA wasused the assumption of homogeneity of slopes wastested using a general linear model(GLM). Allstatistical analyses were performed usingSYSTAT

statistical analysis package.P-0.05 was consid-ered significant. All values are given asVO2

means"standard error.

2.7. Estimating the magnitude of SDA

The magnitude of the SDA in each experimentis equal to the area under the curve(Figs. 1 and2) of the experimental group(i.e. total oxygenconsumed) minus the area under the curve of the

380 S. Iglesias et al. / Comparative Biochemistry and Physiology Part A 135 (2003) 377–382

Fig. 2. Mean (ml O h ) of Eulamprus quoyii after repeated feeding of a total 12 mealworms(�; ns8) and in unfed controlsy1VO2 2

(m; ns8). Black arrows indicate the times at which lizards were fed. Error bars are"standard error.

control group(i.e. oxygen consumed due to resting) (Jobling, 1981). Areas were calculated usingVO2

Simpson’s approximation, which is exact whenapproximating the area under a curve of a poly-nomial of order 3 or less(Stein and Barcellos,1992).

3. Results

3.1. Assimilation efficiency of energy (AE)

The average dry matter content of mealwormsin this study was 41.0"1.1%. Gross energy con-tent of mealworms was 27.7"0.4 kJ per gram ofdry matter. The AE ofE. quoyii fed six mealwormswas 85.7"1.3%.

3.2. SDA in response to single feeding

The body mass of the experimental group(mean: 26.9 g) was not significantly different fromthat of the controls(mean: 27.0 g;F s0.002,1,10

Ps0.96) at the start of the experiment.VO2

changed significantly over time in both the exper-imental and control groups(experimental:F s7,39

6.437, P-0.01; control: F s2.62, Ps0.031;7,39

Fig. 1),with mean of experimental lizardsVO2

being significantly higher than those of the unfedcontrols(F s8.70,Ps0.02).1,9

At 0 h the mean in the experimental lizardsVO2

(3.62"0.26 ml O h ,ns6) and in the unfedy12

controls (3.10"0.73 ml O h ,ns6) were noty12

significantly different(F s0.54,Ps0.48). VO1,9 2

of the experimental group rose significantly com-pared to the controls 5 h after the meal(F s1,9

40.00,P-0.01). of the experimental groupVO2

began to decrease after 20 h(Fig. 1), althoughremained significantly elevated comparedVO2

with the unfed controls after 25 h(F s22.10,1,9

P-0.01). After 48 h, of digesting lizards didVO2

not differ significantly from the unfed controls(F s0.895,Ps0.37). Peak in the experi-VO1,9 2

mental group was 7.95"0.73 ml O h (ns6)y12

and occurred 15 h after the meal and coincideswith a peak in the controls at the same time dueto circadian rhythms in the resting metabolic rate.

3.3. SDA in response to repeated feeding

The body mass of the experimental group(mean: 28.8 g) was not significantly different fromthat of the controls(mean: 28.9 g;F s0.002,1,10

Ps0.96) at the start of the experiment. variedVO2

significantly over time in both experimental andcontrol groups(experimental:F s4.55, P-13,97

0.01; control:F s4.14,P-0.01; Fig. 2). How-13,97

ever, mean of experimental lizards wasVO2

significantly higher than those of the unfed con-trols as a result of feeding(F s41.94, P-1,13

0.01).At 0 h the mean in the experimental groupVO2

(2.42"0.17 ml O h , ns8) and the controlsy12

(2.33"0.3 7 ml O h ,ns8) were not signifi-y12

cantly different. of the experimental groupVO2

rose significantly compared with the controls 5 hafter the meal(1.29"0.01 g) (F s40.00, P-1,9

0.01). rose to a peak of 6.67"0.76 ml OVO2 2

h after 15 h and remained elevated for morey1

than 45 h (i.e. from 5 to 50 h after feeding).

381S. Iglesias et al. / Comparative Biochemistry and Physiology Part A 135 (2003) 377–382

returned to values not significantly differentVO2

from the unfed controls at 55 h(F s0.60,Ps1,13

0.45), which was 20 h after the last meal consumedby this group. of the lizards in the experi-VO2

mental group was on average 1.8"0.1 times largerthan the of the unfed controls during this 50VO2

h period.

3.4. Energetic cost feeding

Body mass of the lizards in the experimentaland control groups for both the single feeding andfrequent feeding experiments were not significant-ly different from each other(F s0.413, Ps3,24

7.45). The exact duration of the SDA resultingfrom the single consumption of six mealworms isdifficult to estimate because the of the lizardsVO2

remained significantly elevated compared withtheir prefeeding values 25 h after the meal and thenext measurement did not occur until 48 h afterthe meal. Based on our results and the standardpattern of SDA observed in other animals, includ-ing the closely related lizardEulamprus tympanum(Jobling, 1981; Robert and Thompson, 2000),

in E. quoyii was assumed to return to pre-VO2

feeding values after 30 h. The energetic cost ofconsuming six mealworms was therefore calculat-ed for an SDA of 30 h duration(as described inSection 2.6). The total amount of oxygen con-sumed over 30 h by a lizard of similar size tothose used in our experiments after consuming sixmealworms is 61.10 ml O or 1227 JwAssuming2

that 1 l of oxygen releases 20.08 kJ,(Schmidt-Nielsen, 1990)x.The total volume of oxygen consumed by the

same lizard after assimilating two separate mealsof six mealworms, where the second meal is notconsumed until the first has been completelyassimilated, must therefore equal twice the ener-getic cost of the single meal. Thus, 2454 J areexpended due to SDA, which is 8.1% of the grossenergy content of the meal(30 415 J), or 9.4% ofthe metabolizable energy available(26 072 J) toE. quoyii after feeding in this fashion(i.e.infrequently).The total amount of oxygen consumed by lizards

eating 12 mealworms as three separate meals(6at 0 h, 3 at 20 h and 3 at 35 h), over 50 h isapproximately 110.22 ml O or 2213 J expended2

over 50 h(continuous). Thus, 7.52 % of the grossenergy (29 421 J) of the meal or 8.8% of the

metabolizable energy available(25 220 J) to E.quoyii after feeding in this fashion(i.e. frequently)is expended due to SDA.

4. Discussion

The lizard controls in both experimental treat-ments experienced significant variations inVO2

over time. These variations are the result of anendogenous circadian rhythm in the resting meta-bolic rate ofE. quoyii which persists in the absenceof an external light cue, but is dampened byconstant light(Iglesias, 2001). All lizards showedsimilar variations in through time as a resultVO2

of these circadian rhythms. However, wasVO2

significantly greater in the digesting lizards com-pared with the unfed controls in both treatments,which is indicative of an SDA effect.The energetic cost incurred byE. quoyii feeding

frequently (8.8% of metabolizable energy) wassimilar to that calculated for lizards eating thesame amount of food in an infrequent fashion(9.4% of metabolizable energy). Thus, any cost ofup-regulating digestive processes from a fastingstate contributes little, if anything, to the SDAresponse ofE. quoyii.The cost of up-regulation of digestive processes

from a fasted state may not be as great in reptilesthat feed frequently compared with those that feedinfrequently and this could explain the much largerSDA responses observed in the infrequent feeders(Secor and Diamond, 1995; Secor, 2001; Wang etal., 2001). Frequent feeders avoid atrophy anddown-regulation of their intestine and hence couldavoid the cost of reversing these processes uponfeeding (Secor, 2001). The cost of up-regulationcontributes little to SDA inPython molurus (Over-gaard et al., 2002), however, and the reason forthe difference in SDA response between frequentfeeders and infrequent feeders may lie elsewhere.Frequent feeding results in a sustained rise in

metabolic rate( ) in E. quoyii where after˙ ˙VO VO2 2

feeding almost doubled compared with fastinglizards at rest. This rise was sustained for 50 hwith frequent feeding under experimental condi-tions. Thus, lizards in the field may experiencesignificant and prolonged elevation in metabolicrate as a result of digestive state. Higher metabolicrates equate to higher energy requirements(Ben-nett and Dawson, 1976) but, since the energeticcost of feeding frequently is similar(or smaller)

382 S. Iglesias et al. / Comparative Biochemistry and Physiology Part A 135 (2003) 377–382

than single feeding, lizards still maintain a netenergy gain despite the prolonged rise in metabo-lism. The increase in resulting from frequentVO2

feeding is directly tied to food consumption andis only sustained as long as food is being consumedby the lizard. In this fashion, metabolic rate andhence energy requirements in lizards may be reg-ulated by food availability to a certain extent.In summary, the energetic cost of digestion and

assimilation (SDA) is affected to only a smalldegree by feeding habit(frequent or infrequent)in E. quoyii. The feeding habit adopted byE.quoyii in the field (frequent) may minimise thecost of SDA experienced over long periods, despitealso resulting in a sustained rise in duringVO2

this time. Our study has demonstrated and quanti-fied the affect of feeding habit on the of aVO2

lizard under experimental conditions. Furthermeasurements of the metabolic response to feedingand feeding regime in other lizards are necessaryto determining if the SDA response ofE. quoyiiis typical of all small frequent feeding reptiles.

Acknowledgments

We thank K. Robert, J. Herbert, J. Sparrow, K.Rogers and C. Brown for all their help and advice.We also thank G. Packard for advice on statisticsand S. Abubla without whom this study wouldnever have taken place. This research had theapproval of the NSW National Parks and WildlifeService, and the University of Sydney AnimalCare and Ethics Committee(LO4y5-2001y1y3380).

References

Bennett, A.F., Alder, H.J., 1986. Thermal relations of someAustralian skinks(Sauria: Scincidae). Copeia 1986, 57–64.

Bennett, A.F., Dawson, W.R., 1976. Metabolism. In: Gans, C.,Dawson, W.R.(Eds.), Physiology. Biology of the Reptilia,Vol. 5. Academic Press, New York.

Cruz-Neto, A.P., Andrade, D.V., Abe, A.S., 2001. Energeticand physiological correlates of prey handling and ingestionin lizards and snakes. Comp. Biochem. Physiol. A 128,503–511.

Gorecki, A., 1975. The adiabatic bomb calorimeter. In: Gro-dinski, W., Klekowski, R.Z., Duncan, A.(Eds.), Methodsfor Ecological Bioenergetics. Blackwell ScientificPublications, Oxford, pp. 281–287.

Greer, A.E., 1989. The Biology and Evolution of AustralianLizards. Surrey Beatty & Sons, Chipping Norton, NSW.

Halliday, T., Adler, K., 1987. The Encyclopedia of Reptilesand Amphibians. Equinox(Oxford) Ltd, London.

Iglesias, S., 2001. Specific dynamic action(SDA) in the lizardEulamprus quoyii. Honours Thesis, University of Sydney,NSW, Australia, unpublished.

Janes, D.N., Chappell, M.A., 1995. The effect of ration sizeand body size on specific dynamic action in Adele penguinchicks,Pygoscelis adeliae. Physiol. Zool. 68, 1029–1044.

Jobling, M., 1981. The influences of feeding on the metabolicrate of fishes: a short review. J. Fish. Biol. 18, 385–400.

Kleiber, M., 1961. The Fire of Life: An Introduction intoAnimal Energetics. Wiley, New York.

Overgaard, J., Andersen, J.B., Wang, T., 2002. The effects offasting duration on the metabolic response to feeding inPython molurus: an evaluation of the energetic costs asso-ciated with gastrointestinal growth and upregulation. Phy-siol. Biochem. Zool. 75, 360–368.

Purves, W.K., Orians, G.H., Heller, H.C., 1992. Energy,enzymes and catalysis. Life: The Science of Biology. Sin-auer Assocs. Inc, Sunderland, pp. 113–133.

Robert, K.A., Thompson, M.B., 2000. Influence of feeding onthe metabolic rate of the lizard,Eulamprus tympanum.Copeia 2000, 851–855.

Roberts, L.A., 1968. Oxygen consumption in the lizardUtastansnuriana. Ecol. 49, 809–819.

Schmidt-Nielsen, K., 1990. Animal Physiology: Adaptationsand Environment. Cambridge University Press, New York.

Secor, S.M., 2001. Regulation of digestive performance: aproposed adaptive response. Comp. Biochem. Physiol. A128, 565–577.

Secor, S.M., Diamond, J., 1995. Adaptive responses to feedingin Burmese pythons: pay before pumping. J. Exp. Biol. 198,1313–1325.

Secor, S.M., Diamond, J., 1997. Determinants of the postfeeding metabolic response of Burmese pythons, Pythonmolurus. Physiol. Zool. 70, 202–212.

Secor, S.M., Phillips, J.A., 1997. Specific dynamic action of alarge carnivorous lizard,Varanus albigularis. Comp. Bioch-em. Physiol. A 117, 515–522.

Sievert, L.M., Bailey, J.K., 2000. Specific dynamic action inthe toad,Bufo woodhousii. Copeia 2000, 1076–1078.

Simmons, J., 1987. Herpetological collecting and collectionsmanagement. SSAR Herpetol. Circ. 16, 1–70.

Spencer, R., Thompson, M.B., Hume, I.D., 1998. The diet anddigestive energetics of an Australian short-necked turtle,Emydura macquarii. Comp. Biochem. Physiol. A 121,341–349.

Stein, S.K., Barcellos, A., 1992. Estimating the definite inte-gral. Calculus and Analytic Geometry. McGraw-Hill, NewYork, pp. 278–285.

Wang, T., Busk, M., Overgaard, J., 2001. The respiratoryconsequences of feeding in amphibians and reptiles. Comp.Biochem. Physiol. A 128, 535–549.

Weigel, J., 1988. Care of Australian Reptiles in Captivity.Reptile Keeper’s Assoc, Gosford, NSW.

Withers, P.C., 1977. Measurement of VO , and VCO , and2 2

evaporative water loss with a flow-through mask. J. Appl.Physiol. 42, 120–123.