J. therm. Biol. Vol. 10, No. 1, pp. 41-45, 1985 0306-4565/85 $3.00+0.

00
Printed in Great Britain. All rights reserved Copyright ,( 1985 Pergamon Press Ltd

ALTERED T H E R M O R E G U L A T I O N IN THE IGUANA

DIPSOSA URUS D O R S A L I S FOLLOWING EXERCISE
J. G. CANNON* and M. J. KLUGER
Department of Physiology, The University of Michigan Medical School, Ann Arbor, MI 48109, U.S.A.

(Received 4 April 1984; accepted 16 May 1984)

Abstract--1. Seven desert iguanas ran on a motorized treadmill for 20-min periods. Before, during and
after exercise, the iguanas were in a thermal gradient which allowed them to thermoregnlate behaviourally.
2. For several hours following exercise, the iguanas selected warmer ambient temperatures, resulting
in small, but statistically significant, increases in body temperature.
3. The increases in temperature were proportional to the exercise intensity.
4. These changes were not observed if exercise was preceded by administration of the antipyretic drug,
sodium salicylate.
5. These data support the hypothesis that exercise causes a change in central thermoregnlatory control
which may be similar to fever caused by infection.

Key Word Index--Thermoregulation; desert iguanas; Dipsosaurus dorsalis; exercise; fever; hyperthermia.

INTRODUCTION In exercising mammals, the high rates of heat

production and heat elimination may make detection
During exercise, mammalian core temperature can of a change in the thermoregulatory set-point follow-
increase by several degrees Celsius in a manner ing exericise difficult. Other consequences of exercise,
proportional to the exercise intensity and relatively such as fluid shifts and dehydration, sometimes com-
independent of ambient temperature (Nielsen 1938). promise heat dissipation mechanism (Greenleaf and
Many physiologists have hypothesized that this
Castle, 1971; Nielsen et al,, 1971) and thus lead to an
occurs because exercise somehow causes a resetting apparent, but not actual, shift in the set-point.
of the thermoregulatory "set-point" (Brooks et A poikilotherm such as the desert iguana
al., 1971; Hammel et al., 1963; Jequier, 1970; Dipsosaurus dorsalis may be a better animal model to
Myers et al., 1977). Tl-.is hypothesis has been study changes in the thermoregulatory set-point since
repeatedly tested by comparing body temperature it does not produce or store significant amounts of
thresholds at which effector responses (such as sweat-
metabolic heat. This animal depends upon environ-
ing and peripheral vasodilation) are initiated during mental heat sources to regulate body temperature.
exercise with other forms of hyperthermia. Much Thus, the body temperature attained through behav-
(Nadel, 1977; Wyndham, 1973), but not all (Johnson
ioural selection of a preferred ambient temperature
and Park, 1981), of the data from these experiments
reflects the status of the lizard's thermoregulatory
favour the viewpoint that no change in the thermo-
system (Bernheim and Kluger, 1976, 1977) and is not
regulatory set-point actually occurs during exercise.
complicated by fluid losses from sweating, internal
Thermoregulation after exercise has not been
fluid compartment shifts or changes in energy sub-
extensively studied. In 1973, Haight and Keatinge
strate availability. A large body of evidence has
found that human core temperature remained
shown that the primary effector mechanism of poi-
elevated for many hours following long-duration
kiiotherms is behavioural selection of an appropriate
exercise. Furthermore, sweat responses to heat and
ambient environment. Thus, an increase in the body
metabolic responses to cold occurred at warmer
temperature of a free-ranging lizard in a steady
temperatures after exercise implying that the set-
thermal gradient represents a repositioning of the
points of these subjects were elevated. The authors animal in a warmer part of the gradient.
suggested that endogenous pyrogen (EP), the medi- In the present study, 7 desert iguanas were exer-
ator of fever and other non-specific host defence cised on a monitorized treadmill. Cloacal tem-
responses during infection, may be responsible for perature was monitored before, during and after
these results. Support for this theory comes from a exercise to determine if changes in thermoregulation
recent investigation showing that a factor is present occur compared to time control periods. Two tread-
in human plasma after exercise which causes fever and mill speeds were employed to ascertain if observed
other responses associated with EP (interleukin 1) elevations in body temperature correlated with exer-
when injected into rats (Cannon and Kluger, 1983). cise intensity or were merely non-specific stress re-
sponses to handling and the novelty of the protocol.
*Present address: Department of Medicine, Tufts Univer- Finally, to test the hypothesis that observed increases
sity School of Medicine, 136 Harrison Avenue, Boston, in body temperature following exercise may be medi-
MA 0211 I, U.S.A. ated by a mechanism similar to that causing fever
41
42 J. G. CANNONand M. J. KLUGER

during infection, the antipyretic drug sodium salicy- prodded form time to time to simulate the exercise
late was administered before exercise in an attempt to protocols as much as possible. At the end of this
block the rise in body temperature. period, the lizards were returned to the thermal
gradient and left undisturbed.
MATERIALS AND METHODS
Experiment 3. Effect o/'
sodium salicylate on body
Seven desert iguanas, mean body wt 70 + 5 g temperature of rested and exercised lizards
(SEM), were maintained in individual 0 . 2 m At 9:30 a.m., groups of lizards (less than 4 on any
high x 0.2 m wide x 1.2 m long enclosures with sand- given day) were given an i.p. injection of 7.5 mg
covered floors. The enclosures were illuminated sodium salicylate in sodium chloride solution (saline).
between 6a.m. and 6p.m. with fluorescent lights. This dose of sodium salicylate has been shown to
Between 8:30 a.m. and 4:30 p.m., 250 W heat lamps block fever in infected lizards (Bernheim and Kluger,
at one end of each enclosure created a thermal 1976). The animals were then either exercised at
gradient ranging from 55~C under the lamps to 25~C 0.54km h ~ for 2 0 m i n and then returned to their
at the other end. When all lamps were off, ambient thermal gradient, or not exercised (controls) and
temperature was a uniform 22 + 2~C. The animals returned to their thermal-gradients. For the next
were fed canned dog food, romaine lettuce, dandelion several hours, body temperature was monitored in
flowers and bananas. these groups of undisturbed lizards.
A custom-built treadmill provided a 0.2 × 0.6 m Statistical comparisons were made by analysis of
running surface and continuously variable speed con- variance. Significant differences between any two
trol up to 2.5 km h ~. The running surface was groups were then determined by Tukey's method for
surrounded with 0.2 m high cardboard walls and one equal sample sizes.
corner was irradiated with a 250 W heat lamp. The
animals were kept moving by lightly stroking their RESULTS
tails against the grain of the skin or gently tapping the
upper posterior hindlimb area. Preliminary experi- Experiment 1. Responses 0t lizards to low- and high-
ments indicated that the lizards could maintain a pace speed exercise
of 0.65 km h -~ for 2 0 m i n which is similar to the Figure 1 illustrates the temperature variations of
exercise performance of D. dorsalis reported by two animals exercised for 2 0 m i n at 0.54km h -I
others (John-Alder and Bennett, 1981). At higher compared to the same animals when they were un-
speeds, respiratory movements became exaggerated. disturbed. The cloaca[ temperature of the lizards
The animals would soon lock their legs, expand their oscillated over a range of several degrees as the
thorax and abdomen and close their eyelids. Once animals shuttled between warmer and cooler parts of
adopting this defensive posture, the animals could the thermal gradient. When hourly averages were
not be coaxed into further movement. computed, however, these rapid temperature
A c o p p e r - c o n s t a n t a n thermocouple sheathed in fluctuations were damped out, illustrating a tightly
PE-160 polyethylene tubing was inserted ca. 5 c m regulated circadian pattern which peaked at
into the cloaca and taped to the tail of each lizard at 10-11 a.m. and slowly declined over the next several
8 a.m. Temperature was recorded at 30-s intervals on hours (Fig. 2, Q).
a Honeywell Electronik 112 multipoint recorder.
9o.m lOom 11 a m noon
Experiment 1. Responses q/" lizards to low- and high- 41
speed exercise
Each animal exercised on the treadmill for 20 min 39,~i i /\ ,,,,"~ /" 14 It I I

at 0 . 5 4 k m h ~ (low speed) or at 0.65km h ~ (high

speed) and was then returned to its thermal gradient 3z "~ ,; / i i ',,j ~;\\
enclosure. On a given day, no more than 3 lizards ,' --...&_
were subjected to any exercise protocol and only one L 35
LIZARD A
\J
animal was placed on the treadmill at a time. Exercise #
periods began at 9:30, 9:50 and 10: 10a.m.

Experiment 2. Time controls

~E EXERCISE
Two time controls were used. Undisturbed time
controls consisted of 7 lizards which were immedi- 1lore
9Qm lOQm noon
ately returned to the thermal gradient enclosures after ~ 41 "
insertion of the cloacal thermocouples and left un-
disturbed. A second time control group consisted of "~ 39
o / i
7 lizards placed on the exercise treadmill, but not -_ -g
L iI[,
exercised. This control group was necessary since in 37 ] /\i
Experiment l it was found that the lizards lost a great \j '
deal of heat while exercising on the treadmill (see 35
LIZARD e
Results). To ensure that the post-exercise tem-
perature elevation was not merely a compensatory
response to this heat loss while exercising, lizards Fig. 1. The individual responses of 2 lizards to 20 min of
were placed on the stationary treadmill for 2 0 m i n exercise at 0.54 km h-~ ( ) are compared to undisturbed
with the heat lamp off. The animals were handled and thermoregulatory behaviour (. . . . ).
 Exercise and body temperature in lizards 43

II Ip<ool
i .0.5 - ." •. ,

-=,,=,
I-. 0.0

q
~ -0.5

9a.m. lOo.m 11a.m noon 1 p.m. 2 p.m. 3 p.m.

Fig. 2. The change in lizard cloaeal temperature from 9 a.m. baseline (hourly means _+SEM, n = 7). The
following interventions were performed at approx. 10a.m.: (i) O, undisturbed in thermal gradient
enclosures, body temperature (Tb) at 9 a.m. = 37.8 + 0.2°C; (ii) &, exercise at 0.54 km h -t for 20 min, Tb
at 9 a.m. = 37.6 + 0.3°C; (iii) II, exercise at 0.65 km h-J for 20 min, Tb at 9 a.m. = 37.3 + 0.3°C.

The cloacal temperature of the lizards dropped 3 h period in which post-exercise responses were
progressively during the 20 min exercise period even recorded (Fig. 3).
though a heat source was available to them (Fig. 1,
10 a.m.). Apparently, keeping pace with the treadmill Experiment 3. Effect of sodium salicylate on body
and avoiding the prodding finger of the experimenter temperature selection of rested and exercised lizards
took precedence over short-term thermoregulation. Administration of sodium salicylate did not
The precipitous fall in temperature appeared to be an significantly alter the thermoregulatory patterns of
artefact of the protocol, therefore, the temperature non-exercised (rested) lizards. The 3 h average tem-
data collected during the 20 min on the treadmill was perature after injection of sodium salicylate was
not included in the time-averaged data which follows. slightly, but not statistically significantly, higher than
After exercise, the lizards behaviourally regulated during the same period without injection of sodium
at higher temperatures. Figure 1 presents examples of salicylate (see left-hand side of Fig. 4). Sodium
the two strategies employed by the lizards to elevate salicylate administered immediately before exercise
their temperature. Lizard A spent brief periods in the completely abolished the post-exercise rise in body
warm end of the gradient, but returned there often, temperature, and actually caused a suppression of
resulting in more rapid oscillations of body tem- body temperature for at least 3 h following exercise
perature. In this case maximum temperatures did not (right-hand side of Fig. 4).
exceed those observed during control periods, how-
ever minimum temperatures did not fall as low as 0.4
during control periods, resulting in a higher average P<O.05 ~
body temperature. Lizard B, on the other hand,
0.3
moved to the warmer end of the gradient and
remained there, resulting in a dramatically higher
maximum temperature over control measurements. 0.2
The temperature elevation after exercise is more
readily discernable when the data are plotted as
0.1
time-averaged changes from baselines as in Fig. 2.
Following low intensity exercise, the cloacal tem-

LJ
peratures of the lizards as a group was significantly
elevated over control (P < 0.05) approx. 90 min after
ll~ 0.0
(C) (O)
exercise. The elevation in cioacal temperature after 0.54 km h-1 0.65kin h "1
higher intensity exercise was of greater magnitude -O .1 EXERCISE EXERCISE
(0.9°C, P < 0 . 0 1 ) and duration. The correlation
between running speed (either 0, 0.55 and -O.2
0 . 6 5 k m h -u) and mean 3 h change in temperature
was 0.55 (P = 0.01).
-0.3 (A) (B)
Experiment 2. Time controls UNDISTURBED STRESSED
Cloacal temperature fell progressively while the CONTROL CONTROL
PERIOD PERIOD
lizards were on the motionless treadmill without
access to the heat lamp. This procedure led to erratic Fig. 3. Average change in cloaca] temperature for 3 h
thermoregulatory behaviour when the animals were following: (A) lizards left undisturbed in the thermal gra-
dient chamber; (B) lizards placed on the treadmill, but not
returned to their thermal gradients. However, this exercised; ((2) lizards exercised for 20 min at 0.54 km h-':
handling, introduction to a novel environment, and and (D) lizards exercised for 20 min at 0.65 k h- L Data are
cold stress did not cause the lizards to increase their normalized to 9 a.m. baseline temperatures. *Significant
temperature over undisturbed control values for the different from (A) and (B).
44 J . G . CANNONand M. J. KLUGER

04 experiments did not affect body temperature of non-

exercised lizards, this dose did affect thermo-
0.2 I No drug
e treatment
regulation after exercise. It is, however, possible that
the salicylate was not "blocking a fever" since body
temperature actually fell after the low-intensity exer-
cise and the lizards were not able to perform the
ik oo
<3 - o 2
!~i.[ ii high-intensity exercise after salicylate treatment. The
drug may have affected acid-base balance or some
aspect of arachidonate metabolism which only indir-
ectly influenced temperature regulation.
-0 4

$ahcylate
ium
While these behavioural and pharmacological
results must be viewed with some caution, they do
-0.6
support the "elevated set-point hypothesis" suggested
for mammals. The fact that a post-exercise "hyper-

RESTED EXERCISED thermia" can be elicited in an animal which exhibited

a net loss of heat during exercise argues that a change
in central thermoregulatory drive has occurred rather
Fig. 4. Three-hour average change in lizard cloacal tem-
perature (n = 7) after control or exercise periods with and than a change in peripheral thermoregulatory
without prior i.p. injection of 7.5 mg sodium salicylate. capability.

Acknowledgements--We thank Orin Gelderloos for loaning

us the lizards, John Faulkner and Timothy White for use of
DISCUSSION the small animal treadmill and Jeffrey Tatro and Wendy
These experiments show that the desert iguana Scales for critically reviewing the manuscript.
responds to treadmill exercise with a gradual rise in This work was supported by NIH Grant AI 13878.
body temperature which reaches a maximum
elevation approx. 90--120 min following activity. It
must be emphasized that D. dorsalis thermoregulates
almost exclusively by behavioural means (DeWitt, REFERENCES
1967). Therefore, this temperature elevation after
exercise is clearly not a consequence of impaired Bernheim H. A. and Kluger M. J. (1976) Fever and
antipyresis in the lizard Dipsosaurus dorsalis. Am. J.
effector mechanisms for heat dissipation; the animal
Physiol. 231, 198-203.
must actively seek out a warmer environment. Fur- Bernheim H. A. and Kluger M. J. (1977) Endogenous
thermore, as indicated by the fall in temperature pyrogen-like substance produced by reptiles. J. Physiol.,
during exercise (Fig. 1), the lizards actually lost more Lond. 267, 659-666.
heat during exercise (due to the inability to maintain Brooks G. A., Hittelman K. J., Faulkner J. A. and Beyer B.
proper position under the heat lamp) than they E. (1971) Tissue temperatures and whole-animal oxygen
produced (through increased metabolic rate). There- consumption after exercise. Am. J. Physiol. 221, 427-431.
fore, the ensuing temperature elevation cannot be Cannon J. G. and Kluger M. J. (1983) Endogenous pyrogen
explained as the load-error in a negative feedback activity in human plasma after exercise. Science 220,
control system. 617-619.
DeWitt C. B. (1967) Precision of thermoregulation and its
This series of experiments represents a trade-off of
relation to environmental factors in the desert iguana,
one set of uncertainties for another. While this poi- Dipsosaurus dorsalis. Physiol. Zool. 40, 49-66.
kilothermic model is advantageous because it circum- Greenleaf J. E. and Castle B. L. (1971) Exercise temperature
vents complications of effector origin, it is possible regulation in man during hypohydration and hyper-
that the lizard's behaviour may be influenced by a hydration. J. appl. Physiol. 30, 847-853.
variety of factors and not the exercise per se. How- Haight J. S. J. and Keatinge W. R. (1973) Elevation in set
ever, since the rise observed after exercise was pro- point for body temperature regulation after prolonged
portional to the exercise intensity and not evoked by exercise. J. Physiol., Lond. 229, 77-85.
handling and environmental stresses alone, this con- Hammel H. T., Jackson D. C., Stolwijk J. A. J., Hardy J.
D. and Stromme S. B. (1963)Temperature regulation by
cern is diminished. Further, it is not likely that
hypothalamic proportional control with an adjustable set
thermoregulation after exercise was impaired either point. J. appl. Physiol. 18, 1146-1154.
through fatigue or some hyperkinetic state since no Jequier E. (1970) Reduced hypothalamic set point tem-
c o m m o n pattern was observed; some animals were perature during exercise in man. Experientia 26, 681.
very active, others were not. The only c o m m o n facet John-Alder H. B. and Bennett A. F. (1981) Thermal de-
of their behaviour was that it resulted in an elevated pendence of endurance and locomotory energetics in a
averaged body temperature. lizard. Am..1. Physiol. 241, R342-R349.
The time-course of the rise in body temperature Johnson J. M. and Park M. K. (1981) Effect of upright
and blockade by the drug sodium salicylate are exercise on threshold for cutaneous vasodilation and
sweating. J. appl. Physiol. 50, 814-818.
similar to the development of a fever and antipyresis, Myers R. D., Gisolfi C. V. and Mora F. (1977) Calcium
respectively (Bernheim and Kluger, 1976). Exercising levels in the brain underlie temperature control during
human subjects release an endogenous pyrogen-like exercise in the primate. Nature 266, 178-179.
substance into their plasma (Cannon and Kluger, Nadel E. R. (1977) Problems with Temperature Regulation
1983) and it is possible that exercising lizards release During Exercise, pp. 1-141. Academic Press, New York.
a similar pyrogen (Bernheim and Kluger, 1977). Nielsen B., Hansen G., Jorgensen S. O. and Nielsen E
Although the dose of sodium salicylate used in these (1971) Thermoregulation in exercising man during dehy-
 Exercise and body temperature in lizards 45

dration and hyperhydmtion with water and saline. Int. J. Tukey J. W. (1949) Comparing individual means in the
Biomet. 15, 195-200. analysis of variance. Biometrics 5, 99-114.
Nielsen M. (1938) Die regulation der korpertemperatur bei Wyndham C. H. (1973) The physiology of exercise under
muskelarbeit. Skand. Arch. Physiol. 79, 193-230. heat stress. A. Ret'. Physiol. 35, 193-220.