Reprinted for prirale circularion Irom PHyslorocrcAl Zodrocy, Vol. 44, \o. l, January L971, oo.

2g-32 ' '-' Copyright @ i971 by rhe Lniversity of Chicago. Ali rights reseived. prihred in 0. S. i.

EVAPORATIVE THERMOREGULATION

IN

TURTLES1

M. L. RIEDESEL, J. L. CLOUDSLEY-THOMPSON,z AND J. ANNE CLOUDSLEY-THOMPSON Department of Bioiogy, University of New Mexico, Albuquerque, New Mexico

87106

has been found experimentally that when the body temperature of young African tortoises, Testudo sul'cata Miller, rises to 40 or 41 C, copious salivation takes place, wetting the head, neck, and front legs. This produces evaporative cooling which pre-

It

iocity ranged from 25 to 35 m/minute and the water vapor pressure ranged from 10 to 16 mm Hg.

Special methods were needed in experiments involving the heating or cooling of the head. A day or two before each experiment, a collar comvents any further rise (Cloudsley- prised of a cork ring was fitted round Thompson 1968, 1970). In the work the neck of each turtle while the anidescribed below, the physiology of a mal was anesthetized with ether. This similar phenomenon was investigated was necessary to prevent withdrawal in the box turtle, Terrapene ornata. of the head into the shell during the period of the experiment. Immediately MATERIAL AND METHODS before experimentation, oral and cloBox turtles, Terrapene ornata ornata acal thermistor probes were inserted (Agassiz), were collected from various and a third probe was attached to the localities in New Mexico and Okla- top of the head. Electrocardiograph probes were inserted at the bases of homa. A total of nine specimens was limbs and the turtle strapped to a studied. Between experiments, the ani- the metal clamp. In preliminary experimals were housed in a constant temperature room at 27 C, illuminated ments the head was exposed to the rays of an infrared radiant heat lamp between 05oo and 20oo hr and prowhile the body was shielded, but the vided with water and vegetable food. Body weight was regained within 24 heat produced in this way tended to be either minimal or so excessive as to hr after each experiment. Temperatures were measured with cause blistering. In the reported experiments, therefore, the lamp was rethermistor probes connected to a teleplaced with an electric heating coil thermometer (Yellow Springs Instruembedded in plastic. This was partially ment Co.) and heart rates were recorded by means of a physiograph insulated from the head of the turtle (B & M Instrument Co.). The air ve- by means of a pad of cotton. In any case, the presence of a probe in the l This study was supported by NSF grant mouth made it difficult to see the exact GB 5339. moment at which salivation started. 2 Present address: Department oI Zoology, In other experiments, turtles were Birkbeck College, Malet Street, Londor W.C.1 E?HX. This work was carried out during tenure placed in an ambient temperature of of a National Science Foundation Senior Visiting 45 C and their heads cooled by dripScientist Fellowship, for rvhich appreciative acping ice water onto a pad of cotton knowledgment is hereb5' given.
28

EVAPORATIVE T}IERMOREGULATION

IN

TURTLES

29

bound around them. Heart rate and temperature measurements were made

AND URINATION

at

S-min intervals.
RESULTS
BODY TLMPERATURT AND SEART RATE

The relationship between heart rate and body temperature was established in turtles no. I (223.2 g) and no. 2 (196.5 S) by placing them in an ambient temperature of 50-55 C and recording both parameters simultaneously while their cloacal temperatures rose from 24.5 to 42 C and again as they cooled. The relationship was similar in all turtles, but some individuals showed consistently lower heart rates at all temperatures than did others. For any given body temperature the heart rates in a particular individual were similar whether the animal was being cooled or heated. Some typical data plotted in figure 1 show a direct relationship between heart rate and cloacal temperature. A Qro value of near 2.0 describes the heart-rate data. In contrast to heart rate, the thermoregulatory response of the turtles was not predictable.
TURTLE

Ib,I

Evaporative cooling in Terrapene ornata is engendered either by salivation or by urination. In the former, a considerable amount of frothy saliva was produced from the mouth and nostrils. This was smeared onto the front legs, which were used to wipe the mouth, and it also ran over the head and neck. Urine was rubbed onto the back legs, which frequently became coated with a pad of cool damp mud. Urination usually occurred before salivation (frothing), but sometimes afterward, in an erratic fashion. Salivation was therefore adopted as the indicator of thermoregulation. In order to determine the temperature at which salivation began, turtles were placed in various high ambient temperatures and their cloacal temperatures recorded at intervals of 2-5 min until 10 min after salivation (frothing) was first noted. The results (table 1) do not describe a specific environmental or body temperature for salivation. Weight loss prior to salivation was presumably due to urination plus respiratory water loss, as there was no evidence of defecation during the experiments. Ambient temperatures of 65 and 75 C represent acute thermal stress, and at both temperatures the turtles with the highest weight loss (nos. 8 and 9) started frothing at the lowest cloacal temperature.
RELATIONSEIP BETWEEN BODY TEMPERATI'RE AND EVAPORATIIIE COOLING

CLOACAL

TEMP 'C

Frc.

1.-A

typical plot

of heart rate

versus

cloacal temperature.

Experiments were carried out, mainly on turtles no. 1 and 2, to ascertain the relationship between body temperature and evaporative cooling. The animals were weighed and then exposed to various high ambient temper-

30 u. L. RTEDESEL, J. L. cLoUDsLEy-TrroMpsoN,
TABLE
1

AND

J. A.

clorrDslDy-TrroMpsoN

DlrA conrcruo oN TURTLIs rN

INITIAL WEToET (e)
228.O
231.O

vARrous

Eor

ENvTRoNMENTs
CLoAcAl

TURTLE

No.

AMBIENT TEMPERATURT (c)

45

TIME To SerrverroN

(min)
/J

Tnup. nr Ousrr or Serwerro*

(c)

-66.6 4.3 8.1 6.2 9.8

BoDY WETGI{T Loss

cF..:
10.7 5.1

s7.8
40.4

45
45 55 55 55 65 65 65 65 65 65
7S

115

2t7.5
209.O
L27.S

90 36

34
39
ZJ

38.4 s6.4 J/.O

37.O

ll.1

21.s

21.9
17.8 21.7 7.9
46.O

210.5 232.O

+.J
3.6 1.4 8.9

215.5
196.0 222.O

23

36.9 39.3
JO-O

z3 40
26

37.9

t.4
3.4

r17.5
261.O

38.0
37.8

4.7 9.2
.5-l

46

2t7.5
t22.O
278.5 139.0

t?
?)

39.8
37.2

o(

1.1

387.5

/J /J /J 75

t7 t4
18

ao
5.0 10.8 6.9

72.9

t2.s
26.1 64.3
8S.O

36.8 36.0
52 -.t

atures for periods of 6 hr. After a 2-hr period to allow for equilibration, they were reweighed and their cloacal temperatures measured. This was repeated at hourly intervals for 4 hr. It was noticeable that once the body temperature had approached the ambient temperature during the first hour of exposure, it tended subsequently to continue to rise only slowly and, indeed, between some hourly readings at the

lower ambient temperatures it even dropped. The results, summarized in table 2, show a clear relationship between ambient temperature and rate of water loss. The mean depression of the cloacal temperature increased with ambient temperature, as did the rate of water loss. Some saliva and urine were lost without producing evaporative cooling. In the natural environment, however, soil and mud may
2

TABLE

Tgr lrnex TLMIERATURE,

wErcET Losr By ruRTr,Es, AND RELATED cALcutATroNs oF.DATA cot-LEcrED DuRrNc FwE 4-rrn srrlnv srATE EXpERTMENTS
Ossrnvro

Dlu
Calorica

CALCULATDD VALUES

Tunrrr
No.

Ambient Temp.

(c)

Cloacal Temp,

Ambient Less

Weight

(c)

(e/hr)
0.8

Loss

of Temp. Differential (cal)

Equiv,

Caloricb

of

Equiv. Weight Loss (callhr)

450
9?0

Estimated
Efficiencyc

of Cooliug (%')
108

lr2 lr2

40.5 41.0
44.O
(

2.6

495
931
1

(0.14)

t-o

l'2
lr2
3-9

(0.25)
5.8

(0.1s) r.7 (0.44)

2.4

95 80 70

100

1370 1170
1998

l.os)
6.5

(0.40)
3.1

45.3 45.5

t23s 12l9

(0.88)
6.4

(0.1s)

(0.44) J.J (0.2s)

6t

Norr.-Figures in parentheses are sE. a Based on specific heat of tissue (0.85) and mean body weight of animals (224 g). b Based on heat of vaoorization of water at 45 C (571 cal/eram), c Bced on caloric eqdivalent of temperature differei:tial; weight loss represents evaporation; and the assumption metabolic rate ard heat gain from tlc envimnment are equivaleat in 8ll experiments.

EVAPORATIVE TIIERMOREGULATION

IN

TURTLES

3l

accumulate on the legs to a greater extent than it did under experimental conditions, and thus change the efficiency of cooling.
CONTROI,

OI'

THERMOREGUI,ATION

in an attempt to describe the role of head and body temperatures in determining onset of frothing. In three of
the experiments we were able to induce

A few experiments

were conducted

frothing by applying heat to the head,

whereas on two occasions raising oral temperature to 33.2 and 35.2 C failed to result in frothing (table 3). Cooling the head of a preheated turtle did not stop frothing on one of two occasions.
DISCUSSION

temperature achieved by salivating (frothing) and by urinating on the legs. A function of the very large bladder in these animals may well be to store urine for use in emergency thermoregulation. In some species this bladder is known to retain a considerable quantity of watel for long periods (Mayhew 1968). Under natural conditions evaporative thermoregulation by salivating (frothing) or urinating probably occurs only rarely and in exceptional circumstances. High temperatures are normally avoided, and a certain amount of cooling is achieved by panting. Panting and the evaporation of urine and saliva probably account for the fact that we observed oral and cloacal temperatures fluctuating over long periods in many of our
experiments.

The work described above indicates that, in addition to behavioral thermoregulation and panting, box turtles show physiological control of body
Bonv lwo EEAD TEMpERATURES AND HEART

The calculated efficiency of cooling

presented in table 2 represents rough estimates. However, the magnitude of
3

TABLE

RATE rN EEAD-EEATTNG oR HEAD-coorrNG EXpERrtr4ENTs CONDUCTED UNTII- ONSET OR CESSATION OF T'ROTIIING

TEMPERATURE

(C)

TURTLE

No,

Ambient

Head Surface IFIF

Head-heating Experiments

Hlenreonrs/
MrN

CouurNrs

DURATION oF ExPERIMENT

(min)

2t.o
24.0

23.7 24.4 23.7 23.9

25.5

26.1 26.2 29.2 29.9

s2 .O

23.O

29.1

31.0

42.O

26.1 22.6
24.2

29.4
33.2

NA

2t.s

40.0
39.O

45.5 45.2

45 52 54

56 70 66

I ..,, I ....

ZZ.5

33.8
35.3

NA NA

Z,t-s

27.3

35.6

40.s

min to frothing 75 min to frothing 80 min and no frothing 95 min to frothing 125 min and no frothing
105

110

80
80 100

t2s

[Iead-cooling Experiments

? .... 1 ....

45.0

37.4 38.0

37 .6

3s.6
37,6

28.1

20.o

19.5

85 14

82

90

min of

90

continuous
45.0 38.4
28.4

t7.o

20-o

78

frothing frothing
stopped

45

after 35 min

Norn,-I = initial; F = final; NA = not

available,

32 v. L. RTEDESEL, J, L. cLouDsLEy-THoMpsoN, AND J. A. cLouDSLEy-THoMpsoN

the differential between cloaca and rium with the temperature at

ambient temperatures is impressive. Panting can represent more efficient cooling than frothing and urinating, but evaporation from the body surfaces must be important in minimizing the rate of heat transfer from environment to body surface to core. The existence of physiological thermoregulation implies thermosensitivity, and a temperature-sensitive center manifested by thermally induced blood pressure changes has been demonstrated in the brain of the turtle
Pseud.emys elegans

the surface of the head. Future studies involving the recording of temperatures in the brain may disclose a thermoregulatory center in Terrapene ornata. The data presented here suggest the existence of a primi-

tive

thermolegulation system which may, or may not, involve a single specific neuroregulatory center.
SUMMARY

by Rodbard, Samson, and Ferguson (1950). Warming the brain promotes a rise in blood pressure, cooling it results in a fall. The degree of response is proportional both to the intensity and the site of the stimulus, the most sensitive region being on the level of the third ventricle. The data in table 3 may be affected by the temperature of both the head and the body. The circulation of blood to and from the head is extremely efficient, as evidenced not only by the length of time necessary to raise oral temperature when the head was being heated, but by the fact that the cloacal temperature rose nearly as quickly. It is possible therefore that blood from the body may influence the brain temperature before coming into equilibCr,oupsr,on-Tso rpsoN, J, L. 1968. Thermoregulation in tortoises. Nature (London) 217:575. 1970. On the biology of the desert tortoise Testudo sulcata Miller (London) 160:17-33.

1. A direct correlation has been established between heart rate and cloacal temperature in Terrapene ornata.

2. Bvaporative cooling is engendered by salivation (frothing) and

urinating. 3. Salivation begins when the body temperature reaches 32.3-40.4 C. 4. Evaporative water loss depends upon the thermal stress. 5. Heart rate is correlated with the temperature of the body rather than of the head, but salivation (frothing) may depend upon both. Blood from the body may influence brain temperature before coming into equilibrium with the temperature at the surface of the head. 6. It is suggested that a function of the enlarged bladder of turtles may be to store urine for emergency thermoregulation.

LITERATURE CITED
Bnowrv (ed.), Desert biology. Vol, 1. Academic,

New York.
Rooneno, S., F. Seaasor, and D. Frncuson. 1950. Thermosensitivity of the turtle brain

in

Sudan.

l,

Zool,

Merunw, W. W.

ians and reptiles. Pp. 195-356

1968. Biology

of desert amphib-

in G.

W.

as .manifested by blood pressure Amer. J. Physiol. 100:402-408.

changes.