Initiation of Behavioral Responding for Heat in a Cold Environment

RETROGRADE AMNESIA AND ECS

817

TABLE 2 RETEST LATENCIESFOLLOWINGDIFFERENTINTENSITIESOF TRANSCORNEAL(TC) AND TRANSPINNATE(TP) ECS ADMINISTEREDIMMEDIATELY AYlXR LEARNING

ECS intensity

Median retest latency (see)

N TC

16

Interquartile range (see)

300

--

5 mA

p as compared with FS sham no-FS sham ns

<0.01

ns

<0.01

<0.01

<0.01

<0.01

<0.01

<0.01

<0.01

<0.01

<0.01

ns TP

16

300

284-300

TC

32

109

51-171

10 mA

<0.05 TP

32

185

93-300

TC

16

116

77-164

20 mA

ns TP

16

96

TC

8

300

68-159 --

FS sham

--

<0.01

--

<0.01

ns TP

8

300

TC

8

16

-13-16

no-FS sham

<0.01

--

<0.01

--

ns TP

8

15

11-22

Only at 10 m A ECS intensity was there a significant difference between the effects of TC and TP ECS on retest latency. At this intensity some animals exhibited tonic extension while others did not. In order to determine whether the occurrence of tonic extension bore any relation to the differential effects of ECS, two additional groups of 48 and 47 animals were given 10 mA TC or TP ECS, respectively, 45 sec after learning. This brought the total N in the TC10 condition to 80, and that in the TP10 condition to 79; these two groups had median retest latencies of 95 and 205 see., respectively (p < 0.0005). I n Table 3 the retest latencies of the two 10 m A groups are compared with respect to the method of ECS administration (TC10 vs TP10) and the presence ( + ) or absence (--) of

RETEST LATENCY AS A FUNCTION OF (1) ROUTE OF ECS ADImNJSTRATION,AND (2) CONVULSIVE PATTERN

TC10 --transcorneal ECS (+) --tonic extension

N

tonic extension. The difference in the incidence of tonic extensor convulsions between the TC and TP conditions failed to reach significance (z ~ ---- 3.67, d f = 1 ; p > 0.05). The presence or absence of tonic extension following ECS affected retest performance only when the current was administered through the ears. The retest latencies of the TP10(+) group were significantly lower than those of the TP10(--) group, whereas there was no difference between the TC10(+) and TCI0(--) conditions (p > 0.1). However, the latencies of the TC1G(+) group were significantly lower still than those of the TP10(+) group; in other words, even when both treatments elicited full tonic extensor seizures, TC ECS still produced greater impairment of retention than did TP ECS. DISCUSSION

TABLE 3

Condition

p

TP10 transpinnate ECS ( - ) - - n o tonic extension Median retest latency (see)

Interquartile range (see)

TC10(--)

39

98

56--134

TC10(+)

41

78

41-171

TP10(+)

49

140

69-242

TPI0(--)

30

293

141-300

p

ns

<0.02 <0.01

Weissman [7] has reported that rats which received lowintensity TP ECS showed retrograde amnesia only if the ECS had elicited a full tonic extensor convulsion. Subsequently, Jarvik and Kopp [5] found that when TC ECS was administered to mice, large deficits in retention could be demonstrated even at low, subconvulsive intensities. The results of the present experiments suggest that this discrepancy was due largely to the method of ECS administration. At low current intensity TC ECS disturbed retention just as much when it did not elicit tonic extension as when it did; this confirms an earlier report [Doffman and Jarvik, 2] of seizure-amnesia dissociation in undrugged animals. On the other hand, the amnesic potency of TP ECS was much diminished if it failed to elicit a tonic extensor seizure. Even when TP ECS did elicit tonic extension, however, the retention deficit it produced was still not as great as that caused by convulsive TC ECS of the same intensity. While no systematic attempt was made in the present experiments B

818

DORFMAN AN[) JARVIK

to compare the convulsions elicited by the two treatments, there is evidence [1, 6] that the characteristics of maximal seizures in both animals and humans are relatively independent of the eliciting stimuli, and casual observation of the treated subjects in the present experiments revealed no gross differences in seizure pattern. These results support the hypothesis that the amnesic effects of ECS are due directly to the passage of electric currents through the brain and suggest that the path of the transcranial current can influence the degree of the retention deficit. Hayes [3] has shown that the intracerebral potential gradient during ECS applied to the temporal areas deviates significantly from a straight line joining the stimulating electrodes, primarily due to the high electrical impedance of the cranial bones; this deflects the current laterally to penetrate the skull via the foramina of the cranial nerves and the emissary veins. Such a mechanism might well operate also in the case of TP ECS, or, alternatively, the bulk of the current

may enter the skull through the nearby foramen magnum. In the case of a potential gradient applied to the eyes, however, the posterior orbit with its thin wall and multiple foramina affords a convenient sink for the current, assuring that much of it will flow directly through the anterior cerebrum. It is therefore possible that high current flux it1 rostral brain structures is what makes the TC ECS particularly effective in disrupting retention. The differential effectiveness of TC and TP ECS was not demonstrable when supramaximal electroshock intensities were employed. This may reflect increased current spread within the brain at higher intensities, and makes it unlikely that differences in the method of ECS administration can explain the dissonant results obtained by previous investigators using supramaximal current. Nevertheless, the present experiments do suggest that by manipulating the parameters of the stimulating current it may be possible to delineate an amnesic locus of action of ECS.

REFERENCES 1. Chatrian, G. E. and M. C. Petersen. The convulsive patterns provoked by Indoklon, Metrazol, and electroshock: Some depth electrographic observations in human patients. Electroenceph. clin. Neurophysiol. 12: 715-725, 1960. 2. Dorfman, L. J. and M. E. Jarvik. A parametric study of electroshock-induced retrograde amnesia in mice. Neuropsychologia, 1968 (in press). 3. Hayes, K. J. The current path in electric convulsion shock. Archs NeuroL Psychiat. 63: 102-109, 1950. 4. Jarvik, M. E. and R. Kopp. An improved one-trial passive avoidance learning situation. PsychoL Rep. 21: 221-224, 1967. 5. Jarvik, M. E. and R. Kopp. Transcorneal electroconvulsive shock and retrograde amnesia in mice. J. comp. physiol. Psychol. 64: 431-433, 1967.

6. Toman, J. E. P., E. A. Swinyard and L. S. Goodman. Properties of maximal seizures, and their alteration by antieonvutsant drugs and other agents. J. Neurophysiol. 9: 231-239, 1946. 7. Weissman, A. Effect of electroconvulsive shock intensity, and seizure pattern on retrograde amnesia in rats. J. comp. physioL Psychol. 56: 806--810, 1963. 8. Weissman, A. Drugs and retrograde amnesia, lnt. Rev. Neuro, biol. 10: 167-198, 1967. 9. Wilcoxson, F. and R. A. Wilcox. Some Rapid Approximate Statistical Procedures. Lederle Laboratories, Pearl River, New York, 1964.

Physiology and Behavior. Vol. 3, pp. 819-825. Pergamon Press, 1968. Printed in Great Britain

Effects of a Few Hours a Day of Enriched Experience on Brain Chemistry and Brain Weights M A R K R. R O S E N Z W E I G , W I L L I A M L O V E A N D E D W A R D L. B E N N E T T

University o f California, Berkeley, California, U.S.A. (Received 29 M a r c h 1968) ROSENZWE1G, M. R., W. LOW ANO E. L. B F . ~ r r . Effects of a Few Hours a Day of Enriched Experience on Brain Chemistry and Brain Weights. PHYSIOL.Bmtnv. 3 (6) 819-825, 1968.--Instead of exposing rats to environmentalcomplexity (EC) for 24 hr a day as in previous experiments, we exposed some groups to EC for either 2, 2½, or 4½ hr a day. In five experiments, brain values of the various EC groups were compared with those of littermates kept continuously in the isolated condition (IC) for the duration of the experiment. A few hours a day of EC were found to produce significant changes in activities of acetylcholinesterase and cholinesterase and in weights of brain samples; these changes were quite similar to those produced by 24 hr a day. In Experiments I and II, 2½ hr and 4½ hr groups were in standard colony (SC) conditions when they were not in EC. Each of these experiments also included a 24-hr EC group, and the experiments lasted for eight weeks. Experiment III included a 2-hr EC group and an SC group; the main differences from IC were found to be due to EC and not to SC. In Experiments IV and V, the enriched-experience groups spent 2 hr a day in EC and 22 hr in IC, and the experiments lasted only one month. Two hr a day removal from IC to EC over a 30-day period produced significant cerebral effects. Acetylcholinesterase Brain acetylcholinesterase Brain weight Impoverished experience Effectsof experience on brain Rat Cerebral cortex

HAVING found that exposure to an enriched environment over an 80-day period leads to significant changes in several aspects of brain chemistry and brain anatomy in the rat [1, 8], we have been attempting to determine how the cerebral effects vary with the duration of treatment. Exposure of only 30 days has been found to produce significant effects that were in some cases larger than the 80-day effects [9]. In all of the preceding experiments, the treatment was maintained throughout the 24-hr day, but in the present experiments the duration within experimental days was varied. A specific reason for testing the efficacy of a few hours of daily environmental enrichment was the finding that 2 hr a day of operant conditioning produced only rather small cerebral effects. Would 2 hr a day exposure to an enriched environment yield measurable changes in the brain? Five successive experiments will be reported here.

Enriched experience Neuroanatomy Brain chemistry

age and kept under colony conditions until about 60 days of age when they were assigned among the following four groups: (a) 24-hr EC (Environmental Complexity or Enriched Condition). These animals were housed in a large EC cage provided with "toys" such as metal boxes, ladders, and running wheels. The arrangement of the toys was changed daily, and some toys were rotated among a larger stock. For 30 min a day the animals were placed in a Hebb-Williams maze device, with the pattern of barriers being changed daily. (b) IC (Impoverished Condition). Here the animals lived singly in cages with solid side walls; the cages were placed in an isolation room. Groups a and b thus received standard treatments of our laboratory. (c) 4½ hr EC. These animals lived most of the day in colony cages, three rats to a cage, in the same room as the EC cages. During 4 hr a day, they were placed in an EC cage, furnished with toys as for group a. For 30 min a day, they were placed in the Hebb-Williams apparatus. This 30 rain could precede or follow the period in the EC cage. (d) 2½ hr EC. These animals had the same schedule of EC cage and Hebb-Williams exploration as group c, except that the daily period in the EC cage lasted only 2 hr. Food and water were provided ad libitum to all groups except when animals were in the Hebb-Williams apparatus.

EXPERIMENTS I AND H METHOD

Subjects and Experimental Groups The subjects of Experiment I were 40 male rats of the Berkeley $8 strain. They were weaned at about 25 days of

XThis investigation was supported by Research Grant GB--5537 from National Science Foundation. It also received support from the United States Atomic Energy Commission. 2Presented at 1968 meetings of Federation of American Societies for Experimental Biology. 819

820

ROSENZWEIG, LOVE AND BENNET-I

Seven litters provided four rats each, and in these cases one animal per litter was assigned at random to each of the four groups. The remaining 12 rats came, three each, from four litters. They were divided into three groups of four, each group being matched as closely as possible in body weight. The rats within a group were then assigned to experimental conditions in the same manner as were the quadruplets above. The subjects of Experiment II were 44 males of the S~ strain, four rats from each of eleven litters. They were weaned at about 25 days of age and were then housed in colony cages, four to a cage. At about 60 days of age they were assigned randomly among the four experimental conditions as were the quads of Experiment I. All behavioral procedures replicated those of Experiment I. Experiments I and II were designed and the behavioral procedures were run by W. Love. Brain Analyses The animals of Experiment I were assigned to experimental conditions on September 13, 1966, and were maintained there for 55 days until November 7, when they were decapitated; those of Experiment II were run from December 14, 1966 to February 8, 1967 (56 days). Decapitation and all subsequent procedures were done under code numbers that did not reveal the previous behavioral treatment of the animals. Following routine procedures, we dissected the brains into these standard sections: occipital, somesthetic, motor, remaining dorsal, and ventral cortex, and rest of brain (or subcortex). (For a description of the brain sections, see Rosenzweig et al. [10]). The brain samples were analyzed colorimetrically for activities of acetylcholinesterase (ACHE) and cholinesterase (ChE). The methods were adapted from the procedure of Ellman et al. [3] by adding inhibitors and specific substrates. F o r analysis of ACHE, acetylthiocholine is employed as the substrate and promethezine is used to inhibit ChE activity. F o r analysis of CHE, we inhibit ACHE activity by BW284C51 and use butyrylthiocholine as the substrate. RESULTS

Tissue Weights The results of Experiments I and I I were closely similar, so they will be combined here for economy of presentation. Table 1 gives the results for wet weights of brain tissue. The mean weights of the IC group and the standard deviations

are given for all brain regions. The difference between means of each enriched experience (EC) group and the IC group is given in percentage terms, i.e., 100 (EC-IC)/IC. The absolute weight for any EC group can thus be found, if it is desired, by multiplying the IC value by the appropriate factor from the table; e.g., to find weight of the occipital sample for the 24-hr EC group, multiply the corresponding IC value of 67.6 mg by 1.068, obtaining 72.2 mg. To determine the significances of differences, Duncan's Multiple Range Test was applied to the results of analyses of variance. To conserve space, the means and standard deviations are not given for the various EC groups in Tables i through 3, but only for the IC group, since the latter group will appear in all five experiments. The means differ significantly between the EC and IC conditions, as can be seen from the percentage differences in the tables. On the other hand, the standard deviations do not vary systematically among the groups, so those of the IC group may be taken as representative of the other conditions as well. In Tables 4 and 5 where there are only IC and 2-hr EC groups, means and standard deviations are given for both; it will be seen that there is no consistent difference in SDs, although the usual pattern of differences among the means is found. The differences between the 24-hr EC and IC groups, where treatment extended from about 60 to 116 days of age, are similar to those we have reported for experiments extending from 25 to 105 days of age [1, 7]. Cortical effects are seen in Table 1 to be largest in the occipital and smallest in the somesthetic region, as we have previously reported [1, 2]. Total cortex shows a 4 per cent increase in bulk (p < 0.001). The rest of the brain, after cortex has been removed, shows little effect of environmental treatment. The 4½-hr and 2½-hr EC groups reveal effects very similar to those of the animals that remained in EC throughout the day and night. In no case does any of the differences among EC groups reach statistical significance. Thus a few hours of daily exposure to the enriched environment (the rest of the day being spent in colony conditions) appear to affect the brain as much as do 24 hr. In previous experiments in which $3 rats were run in EC for 80 days, we have found that the increase in cortical weight with EC occurs despite a decrease of about 11 per cent in terminal body weight in comparison with IC littermates. In the present experiments, the 24-hr EC group ended the experiment weighing 7.3 per cent less than the IC group (p < 0.01). The 4½-hr and 2~-hr EC groups sustained

TABLE 1 DIFFERENCESBETWEENIMPOVERISHEDCONDITION(IC) AND ENRICHEDCONDITION(EC) LITTERMATESIN WEIGHTSOF BRAINTISSUE (EXPERIMENTS I AND II, Sa RATS, N = 21 PER GROUP) CORTEX Occipital

Somesthetic

IC Mean (mg) 67.6 54.8 SD 4.0 3.5 Percentage diffs., EC minus IC" 24 hr EC 6.8*** 3.3 4½ hr EC 9.0*** 2.3 2½ hr EC 9.2*** 1.4 *P < 0.05, **P < 0.01, ***P < 0.00l

Motor 26.8 2.5 4.2** 7.5* 8.7**

Rem. Dorsal 302.7 20.4 3.4 2.0 5.8**

Ventral 321.9 21.3 4.0 3.0 2.8

Total 773.9 35.5 4.0*** 3.3** 4.6***

REST OF BRAIN

TOTAL BRAIN

1089.7 50.3

1863.6 81.2

--0.3 -- 1.4 0.4

1.5 0.6 2.2

CORTEX REST 0.710 0.022 4.3*** 4.7*** 4.2***

BRIEF DAILY EXPERIENCE AFFECTS BRAIN MEASURES similar relative weight losses, 11.1 per cent (p < 0.001) and 6.8 per cent (p < 0.001), respectively.

Chemical Results ACHE. Our usual finding is that total activity of AChE increases throughout the brain in EC animals [1]. The increase in the cortex is less than the increase in weight of cortical tissue, so AChE activity per unit of tissue weight (AChE/weight) drops. In the rest of the brain, AChE/weight has usually been found to increase with enriched experience. Changes of these sorts were found in the present experiments, as Table 2 show, but most EC-IC differences fell short of statistical significance. In AChE/weight, only the 2½-hr EC group showed significant drops--3.4 per cent (p < 0.05) in the occipital area, 4.0 per cent (p < 0.05) in ventral cortex, and 3.6 per cent (p < 0.01) for total cortex.

821 exceeded I C in ChE/weight. Comparisons with standard colony groups showed that the main source of the EC-IC difference in total ChE activity was the drop in the I C group, whereas in the cases of cortical weight and total AChE activity, the main source was the rise in the EC group (see Fig. 5 of reference 1). In the present experiments, the per cent differences in total ChE activity was less than the weight difference in most regions. Significant differences in total ChE activity in favor of the EC groups were found in the following cases: 24-hr EC, ventral cortex, 5.3 per cent (p < 0.05); 4½ hr EC, occipital cortex, 6.9 per cent (p < 0.01); 2½ hr EC, occipital cortex, 7.8 per cent (p < 0.001), motor cortex, 7.8 per cent (p < 0.05), total cortex, 4.5 per cent (p < 0.01). In ChE/weight (where the usual 80-day finding is that EC > IC) the only significant changes were the following cases in which EC values fell below IC values for the 24-hr group: occipital cortex, 5.2 per cent (p < 0.001);

TABLE 2 ACETYLCHOL1NESTERASEACTIVITYPER UNIT OF WEIGHTASA FUNCTIONOF ENRICHMENTOR IMPOVERISHMENTOF ENVIRONMENT (EXa'ERIMENTSI AND II, Sa RATSN = 21 PER GROUP) CORTEX

IC Mean a SD

Occipital 4.7 0.23

Somesthetic 5.9 0.37

Percentage diffs., EC minus 24 hr EC --2.3 4½ hr EC --1.4 2½ hr EC -3.4*

IC: --0.2 --0.2 -2.4

Motor 6.4 0.36

Rem. Dorsal 6.3 0.29

Ventral 9.2 0.46

Total 7.3 0.29

REST OF BRAIN 16.3 0.50

TOTAL BRAIN 12.6 0.35

0.0 --1.1 -2.7

--1.2 --1.1 -2.1

--2.5 --2.9 -4.0*

--1.8 --2.1 --3.6**

0.5 1.9 0.8

--0.8 0.1 -1.0

CORTEX REST 0.451 0.013 --2.2 --3.9** -4.3***

Percentage dills., 24-hr EC minus IC, between groups run for 80 days, N : 44 per group: CORTEX Occipital --2.3*

Somesthetie Rem. Dorsal (incl. Motor) --0.9

0.2

Ventral

Total

REST OF BRAIN

TOTAL BRAIN

CORTEX REST

--0.8

--0.9

1.7"

0.4

--2.6**

*p < 0.05, **p < 0.01, ***p < 0.001. aAcetylcholinesterase activity expressed in units of millimicromoles acetylthiocholine hydrolyzed/min/mg.

In previous experiments we have also found the $3 strain to display rather small chemical differences between EC and IC groups. This is shown by the bottom line of the table which presents EC-IC percentage differences from four earlier $3 experiments. The pattern of reduced AChE/weight in cortex and increased values in the rest of the brain is present but weak. As we have usually found, EC is significantly lower than IC in the cortical/subcortical ratio of AChE/weight. In total AChE activity, only a few significant increases occurred in Experiments I and II. In the occipital cortex 4½-hr EC showed a 7.5 per cent rise (p < 0.01) and 2 ~ h r EC showed a 5.4 per cent rise (p < 0.05).

ChE. In earlier experiments in which animals were kept in EC or I C from 25 to 105 days of age, the EC-IC percentage difference in total ChE activity in the cortex has been found to exceed the percentage difference in weight, so EC also

somesthetic cortex, 4.7 per cent (p < 0.01); remaining dorsal cortex, 3.5 per cent (p < 0.05). The failure of total ChE activity to show larger percentage EC-IC differences than did tissue weight in these and other recent EC-IC experiments led us to try to determine the reason for the discrepancy. As we noted above, the ChE difference appears to be caused chiefly by the isolation of the IC animals, but in the last year or two we have been housing more rats at a time in the isolation room, and thus perhaps reducing the degree of isolation as compared with former conditions. Therefore in two recent experiments we housed only 20 IC rats in a separate isolation room and replicated earlier results: The percentage difference in total ChE activity of cortex between EC and IC exceeded the percentage difference in tissue weight.

ChE/AChE. Since both enzymatic activity per unit of weight and total enzymatic activity involve weight of the

822

ROSENZWEIG, LOVE AND BENNE[T

tissue, we have sometimes eliminated tissue weight as a factor in the results by employing the purely chemical measure of ChE/AChE. Since ChE/weight tends to decrease in the cortex with restricted experience, whereas AChE/weight tends to increase in the IC animals, the ratio of ChE to AChE has previously been shown to be significantly smaller in IC than in EC animals [1]. In the present experiment, 2½-hr EC exceeded IC in ChE/AChE in all cortical regions, and significantly so in ventral cortex (p < 0.01) and total cortex (p < 0.05). The other EC groups also exceeded IC in ventral cortex at beyond the 0.05 level. In the cortical/ subcortical ratio of ChE/AChE, both 4~-hr and 2½-hr EC exceeded IC at beyond the 0.05 level, but 24-hr EC did not differ significantly from IC. Thus the 2½-hr EC group showed most clearly differences from their isolated littermates in the usual EC-IC pattern. EXPERIMENT IlI Since the enriched-experience animals spent most of the day in the colony condition, three to a cage, it was necessary to test whether the colony experience might have produced the cerebral differences from the isolated animals. Although in other cases we have not found the differences between colony and isolated animals to be as large or significant as those observed in Experiments I and II, we did not have the precise control of Ss rats kept in colony conditions from 60 to 115 days of age. The present experiment was therefore designed to determine how 2-hr EC and colony animals would differ from IC littermates. METHOD

The subjects were 30 male $8 rats taken from 10 litters. At about 60 days of age, one animal was assigned at random to 2-hr EC, one to the standard colony (SC) condition, and one to IC. The EC animals were placed for 2 hr a day in an enriched environment cage; the other 22 hr they spent in colony cages, three rats to a cage. The SC animals were housed three to a cage in the same cage rack as the ECs. In order to place three rats in each colony cage, two additional rats were added to the EC and two to the SC group during the behavioral procedures. The IC rats lived in individual cages in the isolation room. The animals were assigned to the experimental conditions on October 26, 1967 and were decapitated 54 days later on December 19, 1967.

RESULTS

Tissue Weights Table 3, set up like Table 1, shows the 2-hr EC group to exceed the IC group significantly in weight of all cortical samples, as well as in weight of total brain and in the cortical/ subcortical weight ratio. In contrast, the colony group did not differ significantly from the isolated group in any measure. F o r all measures except remaining dorsal cortex, the colony group had values intermediate between the EC and IC values. EC was significantly greater than SC in motor cortex (p < 0.05), remaining dorsal cortex (p < 0.05), total cortex (p < 0.01), and in the cortical/subcortical ratio (p < 0.001). Thus, two hours a day in the enriched situation caused the EC to differ from the SC group, while 24 hr a day in the colony situation in the busy experimental room did not cause the SC to differ from the IC group. Terminal body weights of all three groups differed by less than one per cent, so the significant brain weight effects occurred in spite of identical body weights.

Chemical Results ACHE. In AChE activity per unit of tissue weight, the 2-hr EC group differed from their IC littermates in the usual EC-IC pattern, EC being lower than IC in the cortex and higher in the rest of the brain. SC did not depart as far from I C as did EC, nor did the SC-IC differences conform to the usual pattern of EC-IC differences. While 2-hr EC fell below I C in all five cortical areas, SC exceeded I C in three of the five cortical areas. In AChE/weight of total cortex, 2-hr EC fell below IC by 2.8 per cent (SC fell below IC by 1.0 per cent); in rest of brain, 2-hr EC exceeded IC by 2.0 per cent (SC by 1.6 per cent), and in the cortical/subcortical ratio EC was 4.7 per cent (p < 0.05) below IC (SC, 2.4 per cent below IC, NS). In total AChE activity, the 2-hr EC group also differed more from the IC group than did the SC group. Two-hr EC differed from IC by the following percentages: total cortex, 3.5; rest of brain, 3.2 (p < 0.05); total brain, 3.3 (p < 0.05). The SC-IC percentage differences were as follows: total cortex, 0.3; rest of brain, 2.9 (p < 0.05); total brain, 2.2 (p < 0.10).

ChE. ChE activity per unit of weight again showed no differences among the three groups for any brain region in

TABLE 3 DIFFERENCESIN WEIGHTSOF BRAIN TISSUEBETWEENLITTERMATESIN IMPOVERISHED(It), COLONY (SC) AND ENRICHED (EC) CONDITIONS (ExPERIMENT III, Sa RATS, N = 10 PER GROUP) CORTEX Occipital IC Mean (nag) SD

73.6 6.1

Somesthetic 56.4 3.0

Percentage dills., EC or SC minus IC: 2-hr EC 11.1"* 6.2* 24-hr SC 5.2 4.8 *p < 0.05, **p < 0.01, ***p < 0.001

Motor 31.1 2.3 7.0* 1.2

Rem. Dorsal 295.0 17.6 5.3* --1.1

Ventral 340.5 35.8 6.4* 2.2

Total 796.6 46.3 6.4** 1.4

REST OF BRAIN

TOTAL BRAIN

1093.4 54.7

1890.0 97.2

1.2 1.2

3.4* 1.3

CORTEX REST 0.729 0.022 5.3*** 0.2

BRIEF DAILY EXPERIENCE AFFECTS BRAIN MEASURES this experiment. In total activity of ChE, the 2-hr EC group had significantly larger values than IC in several measures (occipital, remaining dorsal and total cortex, and in the cortical/subcortical ratio), whereas SC did not differ significantly from IC on any measure. Furthermore, 2-hr EC significantly exceeded SC in occipital cortex, total cortex and total brain.

ChE/AChE. In total cortex, 2-hr EC exceeded IC by 3.8 per cent (p < 0.10) and SC, by 2.4 per cent ('NS). In the rest of the brain, SC departed from the usual pattern by falling below IC by 3.3 per cent (p < 0.10). Because of this, the cortical/subcortical ratio of SC exceeded that of IC by 5.8 per cent (p < 0.05); 2-hr EC exceeded IC by 4.2 per cent (p < 0.10). Except for the cortical/subcortical ratio, 2-hr EC showed the usual pattern of differences from IC more clearly than did SC. Most of the chemical results, although rather weak, point in the same direction as the brain weight results: The differences between brains of 2-hr EC and IC littermates must be attributed more to the 2 hr a day of enriched experience than to the remaining 22 hr a day in the colony situation. EXPERIMENT IV Experiment IV was designed to focus more narrowly on the 2-hr daily EC period by excluding any other social component of experience. For this reason, except for the 2 hr in EC, the experimental animals lived the other 22 hr of the day in individual colony cages. (Since the 2-hr EC and IC groups to be reported here were part of a larger design on interaction of drugs with EC, the 2-hr EC group received an injection of saline solution each day about 20 min prior to the EC session.) METHOD

The subjects were 22 male $3 rats taken from 11 litters. At about 60 days of age, one animal chosen at random from each litter was assigned to the saline-EC group and another animal from each litter, to the IC group. The IC rats were housed continuously in individual cages in the isolation room. The 2-hr EC animals were housed for most of the day in individual cages in a large experimental room. About 9 a.m. each day they were weighed and injected intraperi-

823 toneally with 0.003 ml saline per gram of body weight. Upon injection the rats were placed in groups of 5 or 6 in holding cages. At least 20 rain after the last rat had been injected, they were placed in the enriched environment. For 90 min, the group of 11 was in a large cage provided with "toys" that were changed daily. For 30 min the rats were placed in groups of 5 or 6 in a Hebb-WiUiams apparatus used as a field for exploration. On some days the open field was given first, and on other days, second. At the end of the 2 hr, the rats were returned to their individual home cages. The behavioral phase of Experiment IV extended from June 30 through July 31, 1967. At the end of the 31 days of differential experience, the rats were killed by decapitation. Tissue samples were taken in the same way as for the preceding experiments, except that no motor cortex sample was removed; the remaining dorsal cortex for this experiment thus includes the motor area. RESULTS

Brain Weights The tissue weight results for this experiment are given in Table 4. Even though 22 hr a day were spent in isolation by both groups, 2 hr a day of enriched experience for 31 days produced significant differences from the IC group. The increased braia weight of the 2-hr EC group was accompanied by a 3 per cent reduction in terminal body weight, in comparison with the IC littermates. Chemical Measures ACHE. On the enzymatic measures, the results did not resemble the earlier experiments as closely as they did in tissue weights, but all of the significant differences that occurred were in the usual directions. In AChE/weight, the enriched-experience animals showed less of a drop in cortical values and more of a gain in the rest of the brain (3.7 per cent, p < 0.05), leading to a cortical/subcortical ratio (--4.0 per cent, p < 0.01) that closely approximated the earlier results. There was a greater gain in total AChE activity throughout the brain than in the earlier experiments. ChE. As in the earlier experiments in this paper, there were no significant changes in ChE/weight, although activity per unit of weight was greater in EC than in IC for most cortical areas. Total ChE activity increased significantly in occipital cortex (11.4 per cent, p < 0.001), remaining dorsal

TABLE 4 WEIGHTSOF BRAINTISSUEAND PERCENTAGEDIFFERENCES BETWEENIMPOVERISHEDCONDITION(IC) AND 2-HR ENRICHED CONDITION (EC) LrrlW.RMATES(ExPERIMENTIV, Sa RATS,N ~ 11 PER GROUP) CORTEX Occipital IC Mean (rag) SD 2-hr EC Mean (rag) SD

Somesthetic Rem.Dorsal (incl. Motor)

Ventral

Total

REST OF BRAIN

TOTAL BRAIN

CORTEX REST

71.9 5.6

55.6 2.9

319.8 15.7

314.2 7.8

761.5 16.8

1022.0 43.9

1783.5 58.8

0.746 0.020

80.4 5.2

57.2 4.7

336.3 18.1

320.2 16.3

794.0 36.1

1013.0 42.9

1807.1 75.9

0.784 0.020

Percentage diffs., EC minus IC: 11.8"* 2.8

5.2

1.9

--0.9

1.3

5.1"*

*p < 0.05, **p < 0.001

4.3*

824

ROSENZWEIG, LOVE AND BENNETT

cortex (6.8 per cent, p <0.05), and in total cortex (5.1 per cent, p < 0.05). ChE/AChE. Here there was a small positive EC-IC difference in total cortex and a negative difference in the rest o f the brain, yielding a significant difference in the usual direction for the cortical/subcortical ratio (3.8 per cent, p .< 0.05). Thus only 2 hr a day of EC yielded several chemical as well as weight components of the usual EC-[C pattern of effects. EXPERIMENT V In this experiment the difference in treatment of groups was confined to the 2 hr of EC, since the isolated group lived in the same room and in the same type of colony cages as the EC group. The only differences between EC and IC were that the ECs had 2 hr a day of enriched experience, a daily injection of saline solution and daily handling. (It will be recalled that in the previous experiments the ICs lived in an isolation room in cages with solid side walls.) Evidence of strain generality of the effect was also obtained, since Experiment V employed rats of the Berkeley Sa strain instead of the $8 strain used in the previous experiments. The 22 animals were assigned to the experimental condition on October 27, 1967, at the age of 60 days, and were decapitated one month later.

Chemical Measures ACHE. As usual, the 2-hr EC group fell below the IC group in cortical AChE per unit of weight, but only in the somesthetic region was the difference significant (p < 0.05); for total cortex, the difference o f --2.9 per cent reached only the 0.10 level of confidence. In total activity of ACHE, the results again showed the expected pattern, with the 2-hr EC group exceeding the IC group. The only statistically significant results occurred in the occipital region (10.9 per cent, p < 0.00]) and in total brain (2.8 per cent, p <: 0.05).

ChE. Here the 2-hr EC animals showed the expected greater activity per unit of weight than the ICs, but the differences did not attain statistical significance. In total ChE activity, 2-hr EC exceeded IC significantly in several measures: occipital cortex (14.8 per cent, p < 0.001), ventral cortex (7.0 per cent, p -< 0.05), total cortex (7.0 per cent, p < 0.001), rest of brain (2.8 per cent, p < 0.05), total brain (4.0 per cent, p < 0.01), and the cortical/subcortical ratio (4.1 per cent, p < 0.01). ChE/AChE. The 2-hr EC group exceeded the IC group in several measures: ventral cortex (5.8 per cent, p - ( 0 . 0 5 ) , total cortex (4.3 per cent, p < 0.05), and the cortical/subcortical ratio (4.0 per cent, p < 0.05).

TABLE 5 WEIGHTS OF BRAIN TISSUE AND PERCENTAGEDIFFERENCES BETWEEN IMPOVERISHEDCONDITION (IC) AND 2-HR ENRICHED CONDITION (EC) LITrERMATES (EXPERIMENT V, Sx RATS, N -~- 11 PER GROUP)

CORTEX Occipital IC Mean (mg) SD 2-hr EC Mean (mg) SD Percentage diffs.,

67.3 7.3

Somesthetic Rem. Dorsal (incl. Motor) 53.0 4.4

76.0 56.0 6.2 3.8 EC or SC minus IC: 12.9'** 5.8*

Ventral

Total

276.6 17.5

278.6 22.9

675.5 32.1

291.1 12.9

288.6 14.6

711.8 29.7

5.2*

3.6

5.4**

REST OF BRAIN

TOTAL BRAIN

CORTEX REST

908.3 44.0

1583.8 74.1

0.744 0.017

927.6 31.9

1639.4 58.8

0.767 0.018

2.1

3.5*

3.1'*

*p < 0.05, * * p < 0.01, ***p < 0.001 RESULTS

Brain Weights The brain weight results are given in Table 5. Although the S~ strain has smaller brain weights than the Ss strain, as we have reported previously [6], the EC-IC effects for the S~s in Table 5 are rather similar to those o f the Sas in Table 4. As we haze found repeatedly [2, 7], the percentage EC-IC differences are considerably larger in the occipital than in the nearby somesthetic cortex. Thus the fact that 2 hr a day in EC can significantly increase cerebral weights is found for the Sx as well as the S, strain. Furthermore, it is found even when the EC and IC groups lived in the same room and in the ,same type of home cage. There was no difference in terminal body weights between the EC and IC groups.

In considering this experiment, it must be recalled that the IC animals--unlike our other experiments---did not live in a separate isolation room but rather lived in the same experimental room as the EC animals. Only during the 2-hr day EC period and the brief injection procedure did the experience of the EC and IC groups differ, yet this was enough to produce significant differences in brain weights and brain chemistry: DISCUSSION

The results of these experiments emphasize how readily the brain can be affected by experience. Only 2 hr a day of enriched experience given over a one-month period is sufficient to produce significant increases in weight o f the cerebral cortex. It also increases significantly the total activity of

aNote added in proofi" Recentlywe have run four further 30-day experiments in which the IC rats not only lived in similar cages in the same room with the EC's, but in which the IC as well as the EC animals received daily saline injections. Animals of the $1, Ss and of the inbred Fischer strain were used. All four experimeats, in which only the brief daily F_X2period differentiated the treatment of the two groups, have yielded the usual pattern of EC-IC differences in brain weights.

BRIEF DALLY EXPERIENCE AFFECTS BRAIN MEASURES

825

AChE and ChE and the ChE/AChE ratio in total cortex. F o r tissue weight and for the chemical measures, the cortical/ subcortical ratio is significantly greater for 2-hr EC than for IC. Experiment HI demonstrated that these effects must be attributed specifically to the two hours a day in EC and not to the 22 hr a day in the colony condition. Experiments IV and V showed that the EC effects occur even when the remaining 22 hr a day are spent in isolation. We are not yet sure why 2 hr of differential experience should produce as large or larger cerebral effects as do 24 hours. Two opposed hypotheses can be suggested: (a) Twenty-four hr of exposure to the enriched environment produces an overload of information which tends to inhibit learning. (b) A few hours is sufficient to absorb all the information afforded by a new cage arrangement, and longer exposure produces only habituation. Some observations that we have made of the animals' behavior suggest that hypothesis (b) is more likely to be correct. When they are exposed to the complex environment for only 2 hr a day, the animals display almost continuous activity for the 2 hr, interacting with their cagemates and with all of the furnishings of the cage. Although activity persists during the two hours, its rate is greatest during the first half hour. The rate of activity of the 2-hr or 2½ hr groups is greater than that observed at any time among the 24-hr groups, except for brief bursts of activity when the latter animals are stirred up by rearrangement of the objects within the cage. These observations suggest that the 24-hr groups are not overstimulated but rather that they habituate rapidly to the enriched environment. Explicit tests will be necessary to permit a decision among these or other hypotheses. F o r example, suppose that different groups were to be given different numbers of short periods a day in environments that were so clearly distinctive that experience in one would provide little transfer to experi-

ence in another. Would the cerebral EC effects grow as a monotonic function of the number of different daily EC periods ? Or would there be signs of overload in that at some point the magnitude of effects would decline as the number of EC periods was increased ? Behavioral Effects o f Brief Exposure to Enriched Conditions As part of our overall program of attempting to relate cerebral and behavioral measures, we have begun to measure effects of two hours a day exposure to enriched experience on subsequent problem-solving behavior in several standardized tests. Studies of Hymovitch [4] and Nyman [5] indicate that a few hours of daily experience can significantly improve subsequent problem-solving behavior. In his Experiment I, Hymovitch gave some rats 2½ hr a day of "free-environment" experience from about 25 to about 70 days of age. When tested on the Hebb-Williams Maze (pretraining begun at 80 days of age), the "free-environment" rats were superior to their restricted controls. Nyman reported that ten days of enriched experience (given to various groups at 30-40, 50-60, or 70-80 days of age) improved scores on an Alternation Maze (testing begun at 86 days) and on the HebbWilliams Maze (testing begun at 110 days of age). One hr a day of enriched experience was enough to produce a significant effect on several measures; eight hours of enrichment per day produced significant effects on both mazes and for all ages of exposure. Further research will be needed to determine whether the function relating duration of daily enrichedexperience to cerebral effects is similar to the function relating duration of daily experience to problem-solving scores. We wish to acknowledge the assistance of Hiromi Morimoto, Marie Hebert, Clarence Turtle, Don Gassie, Ann Muto, Anne Betancourt and Diane Rein.

REFERENCES

1. Bennett, E. L., M. C. Diamond, D. Krech and M. R. Rosenzweig. Chemical and anatomical plasticity of brain. Science, N.Y. 146: 610-619, 1964. 2. Bennett, E. L., D. Krech and M. R. Rosenzweig. Reliability and regional specificity of cerebral effects of environmental complexity and training. J. comp. physiol. Psychol. 57: 440--441, 1964. 3. Ellman, G. L., K. D. Courtney, V. N. Andres, Jr., and R. M. Featherstone. A new and rapid determination of acetylcholinesterase activity. Biochem. Pharmacol. 7: 88-95, 1961. 4. Hymovitch, B. The effects of experimental variations on problem-solving in the rat. J. comp. physiol. Psychol. 45: 313-321, 1952. 5. Nyman, A. J. Problem solving in rats as a function of experience at different ages. J. genet. Psychol. 110: 31-39, 1967. 6. Rosenzweig, M. R. Effects of heredity and environment of brain chemistry, brain anatomy, and learning ability in the rat. Symposium on Physiological Determinates of Behavior:

7. 8.

9.

10.

Implications for Mental Retardation. Kans. Stud. Educ. 14 3-34, 1964. Rosenzweig, M. R. Environmental complexity, cerebral change, and behavior. Am. Psychol. 21: 321-332, 1966. Rosenzweig, M. R., E. L. Bennett and M. C. Diamond. Effects of differential environments on brain anatomy and brain chemistry. In: Psychopathology of Mental Development, edited by J. Zubin and G. Jervis. New York: Grune & Stratton, 1967, pp. 45-56. Rosenzweig, M. R., E. L. Bennett and M. C. Diamond. Effects of experience on brain chemistry and brain anatomy. Proceedings International Symposium on Recent Advances in Learning and Retention. Atti Accad. Naz. Lincei, 1968 (in press). Rosenzweig, M. R., D. Krech, E. L. Bennett and M. C. Diamond. Effects of environmental complexity and training on brain chemistry and anatomy: a replication and extension. J. comp. physiol Psychol. 55: 429-437, 1962.

Physiology and Behavior. Vol. 3, pp. 827-830. Pergamon Press, 1968. Printed in Great Britain

Initiation of Behavioral Responding for Heat in a Cold Environment' H. J. C A R L I S L E

Department of Psychology, University of California, Santa Barbara (Received 29 M a r c h 1968) H. J. Initiation of Behavioral Respondingfor Heat in a Cold Environment. PHYSIOL.B~-/AV. 3 (6) 827-830, 1968.----Clipped and unclipped rats were exposed to a cold (0°C) environment, and given the opportunity to learn a leverpressing response to obtain radiant heat. Unclipped rats failed to learn the response, while clipped rats learned after several hours of exposure. Hypothalamic temperature rose initially in both the clipped and unelipped animals in response to a decreasing ambient temperature, while subcutaneous temperature fell. Hypothalamie temperature was maintained for many hours in unclipped rats, but began to fall rapidly after approximately 30 minutes of cold exposure in clipped subjects. Hypothalamic temperature typically fell below 37°C prior to initiation of responding, but this result was not invariant. A decrease in peripheral temperature is considered a sufficient condition for learning in this situation. CARLISLE,

Thermoregulation

Behavior

Hypothalamus

Operant conditioning

Nnxw rats and pigs placed in a cold environment, and given the opportunity to obtain radiant heat by depressing a lever, typically wait several hours before abruptly commencing to work at a steady rate [3, 8, 12]. Weiss and Laties [12] have proposed that the critical factor in the initiation of working for heat is a fall in body temperature. Evidence in favor of this view is the observation that animals cold-exposed [12] or made hypothermic [10] immediately prior to the test wait less time befole commencing to work, while cold-acclimatized animals wait a longer time [9]. In addition, direct measurement of subcutaneous temperature shows a drop of 8.2°C, on the average, before commencement of stable responding for heat [12]. It has been noted [10], however, that subcutaneous temperature measurements need not reflect a fall in deep internal temperature. The purpose of this study is to measure the time course of central as well as peripheral, temperature changes during initial exposure to cold, and during subsequent initiation of operant responding for heat. Specifically, hypothalamic temperature is taken as the measure of central temperature. While hypothalamic temperature need not reflect the absolute level of internal temperature, or even temperatures in other parts of the brain [7], it does parallel intraperitoneal temperature [1]. Furthermore, hypothalamic temperature is of critical importance in thermoregulatory responses [4].

water. Body weight ranged from 390-470 g. Surgery was performed using sodium pentobarbital anesthesia (50 mg/kg), and a K o p f stereotaxic instrument. A small bead-type thermistor affixed to an electrode was implanted in the preoptic area or anterior hypothalamus of thirteen rats, and fixed to the skull with stainless-steel screws and dental acrylic. Electrical connections were made with Winchester plugs permanently affixed to the cranium. Two weeks were allowed for recovery before testing commenced.

Temperature Measurement Temperature-sensing elements were VECO 32A7 thermistors. The elements, 0.013 in. in dia., were soldered to 40 gauge, teflon-insulated copper wire, and insulated with Stoner-Mudge and Insl-X varnish. Each unit was calibrated in a constant-temperature bath against a precision mercury thermometer. The time constant of the element was determined by measuring the time to reach 63 per cent of the final value after moving the unit rapidly from one water bath into a second of a different temperature. The time constant so measured varied between 0,2 and 0.3 sec for the thermistors used in this study. Subcutaneous temperature was measured in some animals by means of a thermistor inserted into polyethylene tubing, and surgically implanted beneath the skin at the nape of the neck. A flexible cable connected the animal to a wheatstone bridge, the output of which was continuously monitored using Leeds-Northrup recording potentiometers.

METHOD

Animals

Behavioral Test Apparatus

Twenty-four male rats of the Sprague-Dawley strain were used. Eleven animals were unoperated controls. The animals were individually caged, and fed Purina lab chow pellets and

The behavioral test apparatus has been described previously [5]. It consisted, essentially, of a 9-in. dia. hardware-

aThis study was supported by U.S. Public Health Service Grant MH-12414. 827

828

CARLISLE

cloth cage with Plexiglas rod flooring. Depression of a Plexiglas lever (defined as a "response") mounted in the cage activated two infrared (IR) heat lamps (defined as a "reinforcement") mounted at the sides of the cage. Each reinforcement was at an intensity of 350 W (175 W per lamp) for a duration of 3 sec. This produced an increment in subcutaneous temperature of 0.20°C in clipped rats, and 0.15°C in unclipped rats. Depression of the lever while the lamps were on did not prolong thermal reinforcement. The numbers of lever presses and reinforcements were recorded on a cumulative recorder and digital print-out counter.

Procedure Three tests were conducted. First, the variation of hypothalamic temperature at a neutral ambient temperature was measured in seven rats for at least 8 hr in order to observe the normal temperature variation. Second, six implanted rats and three normal animals were tested at 0°C with intact pelage for 6-8 hr in order to assay the protective value of fur and to observe behavioral responses. Three of the implanted animals had subcutaneous thermistors. Finally,

0~

.25"

[]

TA

--N~-

~8

eight implanted and eight normal rats were clipped the night prior to a test, and exposed to 0°C for at least 6 hr. Six of these rats (three implanted and three control rats) were given 1 hr in the test cage at a neutral ambient temperature prior to cold exposure. Hypothalamic and, in five cases, subcutaneous temperatures were monitored, as was the rate of working for IR heat. Testing was repeated at weekly intervals for five control and five implanted rats. Rate of responding for heat, as noted elsewhere [3, 8, 12], commences abruptly after several hr of cold exposure. The point at which responding commenced was defined as the point at which the subject began to work at the rate of at least two bar-presses/min, and maintained this rate for one hr with no breaks in responding longer than 10 min. The implanted subjects were perfused at the end of testing. Frozen sections cut at 40~ were obtained, and stained with cresyl violet. The loci of the thermistors were 1-2 mm lateral to the midline between A7 and A8.2 in the Atlas of de Groot [6]. Lesions induced by the implants were approximately 0.5-0.75 mm in dia. No apparent defects in regulation were noted as a consequence of the locus of the lesion in these animals.

~

~

~A--~.~

oli'w, . / o ¢ r---

i

r TIME

IN

r

~ 2

HOURS

r----~r

n 3

4

TIME

• 25 ° ~

O°

TA

5

~ ....

t 7

6

HOURS

g~

I

4o 7

IN

407"~z5" 9:

T~

',' T$ \/

34 jI =

sz~

4

I~ j~

28

;

/

~V,/~, ~ #,m

/V

'.~

I

""%'~J

r-----r g 3 TIME

,J' lj ~"

,~

Z

5

r 4

IN

t~

-IO m

0 s ---T

]

~0

Lo I

2

3 TtME

IN

4 5 HOURS

-r . . . .

n

6

•

HOURS

FIG. 1. Temperature of the hypothalamus (Thy) and subcutaneous tissue (Ts) during cold exposure after 1 hr at a neutral ambient temperature (Ta) for rat No. 5. Average number of responses/rain is indicated for 10 rain periods: (a) unclipped; (b, c, and d) 3 successive weekly tests with fur removed. Ambient temperature was redttced from 25-0°C in 10 rain.

BEHAVIORAL THERMOREGULATION

829

RESULTS

Neutral Ambient Temperature The mean (n -- 7) low and high hypothalamic temperatures (Thy) recorded from unclipped rats during daylight hr were 37.0 and 38.9°C. The lowest recorded value was 36.7°C, while the highest was 39.7°C. One rat was tested for 24 hr; a high of 39.3°C was observed at 0040 hr, and a low of 37.1°C at 0510 hr. This same rat was subsequently tested for only 8 hr; the low was 37.3°C and the high was 39.3°C. Thy remained within the same range when the animals were clipped. Subcutaneous temperature (Ts) was approximately 1.5°C lower than Thy in unclipped rats, and 2.0°C lower in clipped rats. Observation of the animals showed that low temperatures were associated with sleep, while eating and other motor activities were correlated with high temperatures. This result is comparable to that of Abrams and Hammel [1, 2], and will not be dealt with further. The low average Thy of 37°C will be considered as the lower limit of the normal temperature variation in the absence of thermal stress.

was 71-155 min. The eight implanted subjects commenced to work at a steady rate for I R heat after a mean exposure time of 152 rain. The range of latencies was 27-230 rain. The rate of bar-pressing, averaged over the total exposure time, was 2.39/min, while the average rate of reinforcement was 1.43/min for the implanted rats. Thy fell to a mean (n = 8) of 35.6°C, while Ts fell to a mean (n = 5) of 32.5°C prior to initiation of responding. The range of Thy was 32.4 to 37.5°C, while the range of Ts was 27.3-35.6°C. Thy of the rat that learned with the shortest latency (27 rain) was 36.4°C at the inception of responding. Three subjects had a Thy of 37.0 to 37.5°C when responding commenced.

40

--~s"

,

38 3 36

"

. •

• -

39

'°1

Cold Tests with Pelage The nine animals with fur, in all cases, failed to work for I R heat during cold exposure. The average rate of leverpressing was 0.16/min, and the rate of reinforcement was 0.12/min. The rate of lever-pressing by six naive, clipped rats during 1 hr in the test cage at a neutral ambient temperature was 0.29/min, and rate of reinforcement was 0.18/ rain. Rate of working for heat by unclipped rats in the cold is thus not significantly different from that observed in clipped rats at a neutral ambient temperature. Response rate is higher than reinforcement rate, indicating some leverpresses were made while the lamps were on. A normal level of Thy was observed in unclipped rats during cold exposure. The mean (n = 6) range of Thy was 37.4 to 38.3°C. The lowest recorded value was 36.7°C. The mean (n = 3) range of Ts was 33.9 to 36.2°C. The lowest value was 32.0°C. Rate of responding, Thy, and Ts for one rat are shown in Fig. la.

Cold Exposure Without Pelage The eight control rats began to work for heat after an average exposure time of 126 min. The range o f latencies

38

x ~x -~-x ~

x'~.x.~x_ /x / x---~x 3

ou -r" I--"

35

L I

I

0

1

I

I

I

I

I

I

2

3

4

TIME

IN

HOURS

I

I 5

I

I 6

FIG. 2. Average change in hypothalamic temperature for 5 dipped rats during 3 successive cold tests. Arrows indicate the average elapsed time until the animals began to work for IR heat at a steady rate during tests 1-3.

• • ~ o

: t 40

. . . . . . . .

•

.

. .

• .

• .

•

.

"

• .

"

. •

~ os

•

" " -Is

=¢

. . . . .

$ 7 q

$;F ~

;;?..

'

-[

' I. I

"

" .

" •

.. 2

[

l" ' .... " " " " 2 ;

5 TIME

4 iN

5

6

HOURS

FIG. 3. Hypothalamic temperature during 4 successive tests for clipped rat No. 2. Average rate of responding/rain (RPM) is indicated for 10 min periods.

Five control and five implanted rats were tested at weekly intervals in order to determine the latency for responding during repeated tests. F o u r rats responded immediately upon being exposed to the cold on the fourth test; three rats started on the fifth, and three on the sixth test. In general, the animals waited less time on each successive test before commencing to work, and finally responded immediately during the fourth to the sixth exposure. Once an animal began to respond immediately after exposure on one test, it continued to do so on subsequent tests. The data for the five implanted rats was averaged for the first three tests, as shown in Fig. 2. Average Thy began to fall after approximately 30 min of cold exposure. On each successive test the subjects waited less time before commencing to work. The rate of fall of Thy is less rapid on the third than on previous tests prior to work. Figure 1 shows the change in Thy, "Is, and rate of responding for one subject during four successive tests at weekly intervals. The first test (a) was conducted without clipping the fur, while the three subsequent tests (b, c, and d) were conducted with the fur removed. Note that this animal waited less time before commencing to work on successive tests with fur removed. The pattern and time course of Thy during repeated tests for another rat are shown in Fig. 3. The subject remained at a neutral ambient temperature for 1 hr prior to cold exposure. Thy rose initially when the rat was placed in the testing cage. A n invariant rise in Thy was

830

CARLISLI,

then observed when ambient temperature was lowered, followed by a decrease in Thy until sustained responding commenced. This animal was atypical in that it waited longer on the third than on the second test to commence working for I R heat. DISCUSSION

A decrease in peripheral temperature has been considered by Weiss and Laties as the most critical event determining when a naive animal will begin to work for radiant heat [12]; a decrease in deep internal temperature was not ruled out. Baldwin and Ingram [3] have noted that two out of ten pigs learned to respond for I R heat in less than 30 min. Although neither central nor peripheral temperatures were measured, it is doubtful that central temperature could have fallen appreciably in an animal with a large body mass such as the pig. In addition, vasoconstriction induced by cold exposure would be expected to produce a fall in Ts. Baldwin and Ingram suggested that the pig learns to balance heat loss against minimal metabolic rate, which would imply that a complex control system is operating in this situation. The present results show that Thy of the rat is maintained, on the average, between 37 to 39°C at a neutral ambient temperature. Exposure to 0°C with intact pelage results in a less variable maintenance o f average Thy between 37.4 and 38.3°C. The initial response to a decreasing ambient temperature is a fall in "Is, but a rise in Thy, which is then maintained for many hours. The initial response to a decreasing ambient temperature in a clipped rat is also a rise in Thy and a decrease in Ts. I n this case, however, Thy is not maintained, and begins to fall after approximately 30 rain. Deep internal temperature fails to 35.6°C, on the average, prior to commencement of working for heat. However, three rats had a Thy above 37"C when they began to work during the initial cold exposure. A decrease in internal temperature below a level normally observed at a neutral ambient temperature, therefore, cannot be a necessary condition for learning in this situation. Ts begins to fall immediately after cold exposure, so that a

decrease in peripheral temperature could be a sufficient condition for learning. It cannot, however, be a necessary condition, since peripheral temperature falls in the unclipped rat during cold exposure, but learning does not occur. This may imply that a complex control system, sensitive to rate of heat loss relative to heat production, may be involved. Another alternative is that peripheral temperature does not drop far enough in the unclipped rat to initiate responding. The temporal variability in exposure time prior to learning is great, ranging from 27 to 230 rain for rats in this study, and from 5 min to 10 hr for pigs [3]. Hence, the latency for initiation of behavioral responding during the initial cold exposure would not appear to be an accurate or reliable measure of the speed of learning. If it is assumed that an increasing duration of cold exposure in clipped rats is related to an increasing degree of motivation, then this work would be in agreement with other studies noting that motivational variables are primarily related to the performance of a learned response, and not to the speed of acquisition of that response [11]. Clipped rats typically huddle and shiver during cold exposure, and periodically intersperse this behavior with exploration of the cage or grooming. The accidental depression of the lever during a bout of activity is reinforcing only after a variable period of cold exposure, and the rat subsequently learns to work for IR heat. The duration of cold exposure would appear to be a highly variable predictor of when this learning will occur. This study noted that rats typically wait a shorter time on each successive test after the initial session prior to working for heat, but commence to work immediately after being exposed to the cold by the sixth test. Weiss and Laties [12] have also noted a decrease in the latency for responding during repeated tests in the rat. It is not clear what factor or factors determine the latency for responding during these tests, since the lever-pressing response is learned during the first session. Perhaps the latency to onset of responding grows less and less with successive exposures because the rat must learn to use peripheral temperature as a discriminati~.e stimulus for responding on the lever.

REFERENCES

1. Abrams, R. and H. T. Hammel. Hypothalamic temperature in unanesthetized albino rats during feeding and sleeping. Am. J. PhysioL 206: 641-646, 1964. 2. Abrams, R. and H. T. Hammel. Cyclic variations in hypothalamie temperature in unanesthetized rats. Am. J. Physiol. 208: 698-702, 1965. 3. Baldwin, B. A. and D. L. Ingrain. Behavioral thermoregulation in pigs. Physiol Behav. 2: 15-21, 1967. 4. Bligh, J. The thermosensitivity of the hypothalamus and thermoregulation in mammals. Biol. Rev. 41:317-367, 1966. 5. Carlisle, H. J. Heat intake and hypothalamie temperature during behavioral temperature regulation. J. comp. physiol. Psych. 61: 388-397, 1966. 6. de Groot, J. The rat forebrain in stereotaxic coordinates. Verh. Kon. Med. Akad. Wet., B. Natuurkunde 52: 1-40, 1959.

7. Delgado, J. M. R. and T. Hanai. lntracerebral temperatures in free-moving cats. Am. J. Physiol. 211: 755-769, 1966. 8. Jakubczak, L. F. Behavioral thermoregulation in young and old rats. J. appL PhysioL 21: 19-21, 1966. 9. Laties, V. G. and B. Weiss. Behavior in the cold after acclimatization. Science, N.Y. 131: 1891-1892, 1960. 10. Panuska, J. A. and V. Popovic. Learning in hypothermic rats, J. AppL PhysioL 18: 1016--1018, 1963. 11. Stellar, E. Drive and Motivation. In Handbook of Physiology, edited by J. Field, H. W. Magoun and V. E. Hall. Sect. 1, vol. 3. Washington, D.C.: American Physiological Society, 1960, pp. 1501-1527. 12. Weiss, B. and V. G. Laties. Behavioral thermoregulation. Science, IV. Y. 133, 1338-1344, 1961.