The Effect of Early Experience on Water Maze Spatial Learning and Memory in Rats RICHARD C. TEES KRISTIN BUHRMANN JORDAN HANLEY The University of British C o l ~ m b i a Vancouver, British Columbia, Canada

In the first of two experiments on spatial competence, groups of light-reared (LR) and dark-reared (DR) rats were compared using a “latent learning” variation of the Morris Water Maze task. On their initial test, the LR rats benefited more than DR rats did from viewing the room/pool from a platform in the correct location. Further, visually experienced rats remember the location of the platform more than DK rats when retested one month later. In a second experiment, in which a proximal cue as well as location was varied from trial to trial, LR rats again proved to be more competent than their DR counterparts. This second task also revealed significant benefits related to stimulation history in the case of a third group of animals raised in enriched or complex environment (CR) conditions. The results are discussed in terms of the nature of the impact of early experience on the ability to acquire and remember spatial concepts.

In a variety of mammals, including the rat, investigations of the deleterious effects of early, unimodal sensory deprivation have tended to focus on specific discriminative capacities within an affected modality (e.g., Tees, 1986). However, several investigators (e.g., Gottlieb, 197 I ) have made proposals that also predicted deprivation within a single modality will result in intereference with respect to the ontogenesis of other sensory systems. In this regard, most theoretical positions (e.g., Gibson, 1969) on the development of spatial cognition seemed to assign an important role to visual (and other kinds of) cumulative experiences and, for that matter, to cortical mechanisms. Although the evidence supports the idea that spatial competence does depend on cortical and hippocampal mechanisms (see Goodale & Carey, 1990), the evidence is less clear as far as experience is concerned. Dark-reared (DR) rats have been found to be more spatially oriented than their light-reared (LR) counterparts when tested on a Krech hypothesis maze Reprint requests should be sent to Richard C. Tees, Department of Psychology, 2136 West Mall, The University of British Columbia, Vancouver, British Columbia, Canada V6T 1Y7. Received for publication 27 November 1989 Revised for publication 2 May 1990 Accepted at Wiley 24 May 1990 Developmental Psychobiology 23(5):427-439 (1990) 0 1990 by John Wiley & Sons, Inc.

13$O4.OO CCC 0012-1630/90/05O427-

428

TEES, BUHRMANN, AND HANLEY

(Camboni, 3964). On the other hand, rats blinded shortly after birth peiformed less ably than late-blinded rats on an auditory localization task (Spigelman & Bryden, 1967). We (Tees, Midgley, & Nesbit, 1981) have looked at groups of sighted and blinded LR and DR rats tested on a series of Hebb-Williams Maze problems and their reversals using both appetitive and aversive reinforcement conditions. Other LR and DR animals were also compared on a 17-arm radial maze over a 36-day period in which variations in the task were introduced. Blindness at the time of testing had a significant (adverse) impact on the performance of both LR and DR animals on all problems, but a significantly greater effect on the DR animals. Overall, our visually inexperienced animals were found to be less able than experienced controls in acquiring and storing information about spatial location. We also have found that visually inexperienced DR animals are less effective than their visually experienced LR counterparts in demonstrating crossmodal transfer involving location of signals in two modalities (Tees & Buhrmann, 1989). Such is not the case when identical crossmodal transfer tests involve the duration rather than the location of auditory and visual signals (Tees & Symons, 1987). Lack of visual experience does retard the neural development of the visual cortex (see Juraska, 1990),and there is also strong evidence that the visual cortex (beyond O c l ) does play an important part in the learning of complex mazes and in spatial orientation. While our results on the Hebb-Williams and radial arm maze certainly are consistent with the speculation of the function of network underlying spatial ability in the rat, including structures that are adversely affected by lack of early visual experience, there is some evidence to suggest that spatial competence is not seriously affected by a lack of early visual experience when measured in the Morris (1981) place navigation task in which rats learn to escape cool water by swimming to the location of an invisible, submerged platform in a pool. For instance, Sutherland and Dyck (1984) reported that neonatally enucleated rats were very much comparable to LR animals enucleated at adulthood in learning to locate the invisible platform on the standard Morris Maze task. In recent years, variations of the Morris Water Maze task have proved to be specially sensitive and useful in assessing spatial learning and memory in rats with a variety of neurological manipulations. The task has been successfully employed to study the onset of spatial mapping ability in young rats (Rudy, Stadler-Morris, & Albert, 19871, its decline in aged animals (Gage, Kelly, & Bjrklund, 19841, and the adverse effects of hippocampal and neocortical damage (e.g., Sutherland, Kolb, & Whishaw, 1982). The idea that some variations of the task would not be sensitive to a lack of visual experience seemed unlikely. We have utilized (Buhrmann, Tees, & Symons, 1989) one variation of the task to demonstrate that early complex rearing improves performance on that variation. One potentially valuable paradigm has been provided by Sutherland and Linggard (19x2) who gave a simple demonstration of latent spatial learning. Naive rats were placed on the escape platform in a correct location and simply allowed to view the distal cues available. Compared to rats who did not receive this exposure/ training, or who were preexposed to an incorrect placement, these rats learned more rapidly to swim to the platform when given the opportunity, suggesting that they had learned something about the location of the platform by simply standing on it. In recent publications, several investigators have provided experimental evidence (and debate) on the extent to which rats can learn about the location of

EARLY EXPERIENCE AND SPATIAL LEARNING

429

a goal without actually engaging in goal-seeking behavior (e.g., Jacobs, Zaborowski, & Whishaw, 1989; Keith & McVety, 1988; Sutherland, Chew, Baker, & Linggard, 1987). Much of the debate centers on what Morris (1981) called “instantaneous transfer,” which implied the rat was fully capable of making computations that permitted it to generate an accurate “plot” through novel space based only on knowledge acquired from different views of that environment. While the issue of whether normally reared rats can display “true” latent spatial learning with certain kinds of limited experiences is of considerable significance, we were interested in answering a different question, i.e., whether or not visually experienced rats learn more or less from limited exposure than do visually inexperienced animals. Thus, we employed a variation of the latent learning paradigm to test the idea of whether lack of visual experience can influence performance in the water maze. In Experiment I, our variation involved preexposing animals (who had not yet been given the opportunity to swim) to a pool and its environment repeatedly from a platform in what would turn out to be a correct location (or an incorrect location). In addition, we were interested in whether or not a month’s delay after one day of training would affect visually inexperienced rat’s memory for the correct location more than visually experienced animals. A great deal of evidence has accumulated that suggests that lack of early visual stimulation history can affect attention, memory, and perception of temporal and spatial relationships involving visual events (e.g., see Tees, 1986, 1990 for reviews).

Experiment I

Method Subjects and Rearing Conditions Light-reared (LR) subjects consisted of 18 male Long-Evans rats (rattus noruegicus) from 5 litters, and dark-reared (DR) subjects consisted of 18 males from 6 DR litters. The general rearing condition for DR and LR animals has been described previously (Tees, 1968). Subjects were born in plastic cages (42 x 15 x 24 cm) and, at 25 days of age, were weaned and transferred to individual cages (24 x 18 x 18 cm). The LR rats were housed in a colony room on a 12-hrlight : 12-hr-dark schedule. DR litters were transferred to a DR colony room on the day of birth and were housed there for the entire experiment. All subjects received continuous access to food and water throughout the experiment and were handled for a short period once a day for the week before training began. Preexposure/ training began at 120 days of age. At that time, DR rats were also adapted to moderate illumination (1-2 cd/cm2) for 30 min. prior to beginning experimental testing.

Apparatus The rats were tested in a water maze (Morris, 1981) consisting of a circular rigid-walled plastic pool with a diameter of 1.4 m and a height of 0.3 m. Four notches were cut onto the outside edge of the rim of the pool to mark the geographic north (N), south (S), east (E), and west (W) “poles” of the maze. On this basis,

430

TEES, BUHRMANN, AND HANLEY

the pool surface was divided into four quadrants of equal area, NE, NW, SE, and SW. The pool was filled to a depth of 21 cm with cool water (22”C), and black watercolor paint powder (Alphacolor) was added to the pool to render the water opaque. The platform consisted of a black plastic container, 19 cm tall and 9 cm in diameter. The top surface of the invisible container/platform was covered with black mesh to provide the rat with good purchase and was located 2 cm under the surface of the water. The pool resided in a 3 x 3 m laboratory room with a number of conspicuous extra maze visual cues available. The room illumination was 1-2 cd/cm2.

Procedure Six LR and six DR rats were randomly assigned to one of three conditions. The groups formed differed in respect to the experience provided them during four days of pre-training “shaping.” On these four days, each rat was placed on the submerged platform for 10 sec, once per hour, for three consecutive hours. For Condition A (Correct Placement) rats, the platform was located in the northeast quadrant of the pool; for Condition B (Incorrect Placement) rats, the platform was located in the southwest quadrant; for Condition C (Control Placement) rats, the platform was located in a (47 x 27 x 21 cm) plastic maternity bin containing cool, blackened water in a laboratory sink in a separate laboratory room. On the following (5th) day, each rat was tested in the pool on eight separate trials. On trials 1-4 and 6-8, the invisible platform was located in the northeast quadrant 6-8 cm from the wall of the maze. The 5th trial of the day served as a probe trial and no platform was present at all. Each rat was released into the water facing the wall of the pool at one of the geographic poles in a pseudorandom sequence (S-W-N-E-N-S-E-W). The latency to reach the platform (to a maximum of 90 sec) was the primary measure but the movement through the pool of each rat was also recorded on every trial. If the platform was found, the rat was allowed to remain upon it for 10 sec. If the platform was not found within the 90 sec limit, the rat was removed from the pool. Intertrial intervals ranged from 3-5 minutes. Thirty to thirty-five days later, each rat was retested as it had been on the test day (51, and the latency to reach the platform in the northeast quadrant and the swim paths were recorded once again on the series of 8 trials.

Results A repeated measures ANOVA was used to assess differences in escape latency among the six groups of animals (DR and LR rats exposed under the three pretest conditions). The first analysis concerned performance on the first four trials in which the animals were released from each of four release points (S-W-N-E) for the first time. Not surprisingly, trial order had a significant impact on overall performance, F(3,90) = 60.9, p < .0001. Overall performance improved significantly from Trial I ( M = 74.8 sec) to Trial 4 ( M = 20.5 sec). In addition, the interaction between Condition (A, B, C ) and Rearing (LR, DR) was also significant in respect of overall performance on these four trials F(2,30) = 5.6, p < .001. The interaction between Trial, Rearing, and Condition F(6,90) = 3.1, p < .05 also proved to be significant. Further analysis, (Tukey’s HSD comparisons, p < .05),

EARLY EXPERIENCE AND SPATIAL LEARNING

43 1

indicated that one group performed significantly better than the other five groups on the first trial. On that first trial, the group of LR rats who received pretraining with the correct quadrant Northeast (NE) placement on the hidden platform (Condition A), found the platform significantly faster than LR animals who had been placed on the platform in the incorrect southwest (SW) quadrant (Condition B) or on sink/control platform (Condition C) experience. The performance of the DR groups who were exposed under three conditions was not significantly different from those LR animals pretrained under Conditions B and C, nor from one another. Multiple comparison of the 6 groups on Trials 2-4 (Tukey p < .05) revealed no further significant differences in performance (see Figure I ) between the groups. On the fifth trial, in which there was no platform present, all 36 of the animals swam to the NE quadrant first and continued searching/swimming for 90 sec until removal. There were no differences due to Rearing or Condition in the latency-toreach-the-location (LR M = 17.3, DR M = 18.8 sec) on the probe trial. A single analysis of performance on Trials 6-16 revealed several interesting points. (The second probe trial, #13, was not included.) The interaction between Rearing and Trial was significant, F(5,150 = 5.2, p < 0.0001. Trial itself was a significant factor, F(9,270) = 7.7, p < 0.0001. Significantly ( p < .05) poorer performance was observed overall on Trials 6 and 9 (see Figure 1). On Trial 6 (which followed the probe trial in which no platform was available) poor performance (M = 28.3 sec) was observed as far as the latency to find the invisible platform was concerned. Subjects took significantly longer to swim to the northeast location of the platform after the 90 sec swim in the empty pool. However, Rearing itself did not significantly affect performance on that particular trial. On the other hand, on Trial 9, the multiple comparisons (Tukey, p < .05) revealed that the difference in performance due to Rearing was significant on this first trial after a one-month delay. Retested LR animals performed reasonably well (M = 14.5 sec). Dark-reared animals did not (M = 38.5 sec). Comparison of performance on Trial 8 with that of Trial 9 (one month later) revealed that LR animals’ performance was unaffected by the interval (M = 13.2 vs. 14.5 sec) while the retention of DR rats was significantly affected (M = 13.5 vs. 38.5 sec). (See Figure 1). On the second probe trial (no platform) all 36 of the rats again swam into the NE quadrant first, and continued searching/swimming for the platform until removed after 90 sec. There were no significant differences due to Rearing in the latency-to-reach-thelocation (LR M = 12.3 DR M = 13 sec).

Experiment I1 In previous work, we were able to show that early enrichment or complex rearing could have a significant impact on dewhiskered rat using another version of the water maze task (Buhrmann, et al., 1989). In that version, in addition to the invisible platform located in a particular quadrant location, a visual proximal stimulus was attached, to provide a cue to the location of the invisible platform. Subsequent testing consisted of trials in which the cue and/or the location of the platform were removed or relocated. While Sutherland and Dyck (1984) reported that neonatally enucleated rats were very much comparable to LR rats binocularly enucleated at adulthood in trying to locate an invisible platform in a Morris Water Maze, they did report that these (DR) blind rats were significantly less effective

I

20 0

I

1

4) 0 ;

I

/

I

60.0

l

A

I

I

60.0

1

h

s

1

100.0

k

Probe

+

v -

Probe

I

0

-,

0-

(D-

0

N

;

Probe

20.0 a . 0

80.0

30 Day Delay

60.0 100.0

20.0

I

-1

o i

z:p---u-u

Mean Escape Latency (sec) 60.0

80.0

30 Day Delay

40.0

100.0

EARLY EXPERIENCE AND SPATIAL LEARNING

433

than blind (LR) rats when the platform’s location was shifted on probe trials in rediscovering the location of the platform. One of the ideas we have offered is that spatial cognition may be an example of a competence whose initial development involves the establishment of an organizational framework dependent on a variety of early experiences with signals in different modalities (Buhrmann et al., 1989; Turkewitz & Kenny, 1982). If the appropriate visual (and tactual) functioning is a necessary precursor to intermodal spatial ability, then complex rearing should produce a more spatially competent animal and testing in the Morris Water Maze should and could be sensitive to improvement as well as nondevelopment or decline in competence. In fact, the most consistent (although not universal) behavioral finding has been the superior performance of environmentally enriched rodents in land-based, complex spatial maze tasks (Juraska, 1990). While there has been controversy regarding this result as to the relative influence of various aspects of the environmental enrichment or complex rearing, visual pattern experience itself (Forgus, 1954) has been shown to influence maze performance. The second study allows us to investigate both visually inexperienced DR as well as normal LR and complex reared animals under identical conditions in the water maze. A lack of visual experience as well as augmented visual experience provided by complex rearing should have an impact on spatial competence. We might expect to see a pattern of results emerging with respect to probe trials that proved to be sensitive to early rearing conditions in other circumstances (Buhrmann et al., 1989).

Method

Subjects and Rearing Conditions The subjects were light-reared (LR) animals consisting of 15 male Long-Evans rats from 6 litters, dark-reared (DR) animals consisting of 15 males from 5 DR litters, and complex reared (CR) subjects consisting of 15 males from 6 CR litters. The rearing conditions for the DR and LR animals were described previously in Experiment 1 and elsewhere (Tees, 1968). The unique environmental conditions for complex-reared subjects began at the age of 20 days of age. Those litters assigned to the enriched or complex rearing condition were given daily access in groups of 6 for 1-2 hours in a large closed field o r maze. [This daily period of enriched environmental exposure has been reported to have an effective equivalence to continue exposure, at least in terms of some brain measures (Rosenzweig, Love, & Bennett, 1968).] The “open field” or maze was a tall, wire mesh box (180 x 92 x 62 cm). Its floor and two bridges (covered with Corncob Granules) were located approximately 40 and 100 cm above the floor. Wire mesh racks connected the bridges with one another and with the bottom of the apparatus. The floor of the open field is filled with an assortment of toys, many of which were changed daily (Greenough & Green, 1981). This daily exposure continued until testing began at 120 days of age.

Apparatus and Procedure The water maze was the same as that used in Experiment 1. A blue racket ball (painted with white stripes) was used as a proximal cue and was attached by a 20 cm string t o a sheetmetal anchor which allowed the cue to be placed on the platform

434

TEES, BUHRMANN, AND HANLEY

Table 1 Outline of the Trials Sequence for the Water Muze Task. ( A schematic drawing showing the quadrant position uppears in Figure 2.) ~~~

Trial Number

~~

Trial Type

Description

CI/PI CIIPl CIIP1 ClIPI

Cue Cue Cue Cue

C3lP1

located located located located

on on on on

top of top of top of top of

platform platform platform platform

in in in in

quadrant quadrant quadrant quadrant

1 (Northwest). I. 1. 1.

6

C3lPI

Cue located in quadrant 3 (Southeast). Platform located in quadrant I (Northwest). Cue located in quadrant 3 . Platform located in quadrant 1.

7 8

C3lP3 C3lP3

Cue located on top of platform in quadrant 3 . Cue located on top of platform in quadrant 3 .

9

10

C3lP3 C3lP3

Cue located on top of platform in quadrant 3 . 7-day delay from T8. Cue located on top of platform in quadrant 3 .

11 12

-1P3 -1P3

N o Cue. Platform located in quadrant 3 . No Cue. Platform located in quadrant 3 .

13 14

-1P4 -1P4

N o Cue. Platform located in quadrant 4 (Northeaat). No Cue. Platform located in quadrant 4.

or to float freely depending on the nature of the conditions of the trial. The entire apparatus resided in an IAC sound-attenuating chamber (Model 402-A) with numerous conspicuous distal visual cues available. The animals in each of the rearing conditions were tested on 16 trials conducted over 3 days. On the first day, each rat was given two 90-sec habituation trials with no platform or cue present. On the second day, testing began, which consisted of 8 trials (See Table 1 for a summary). For the first 4 trials, the platform was located in the northwest quadrant and a proximal cue was placed on top of it. On Trials 5 and 6, the cue was placed in the southeast quadrant while the platform remained in the northeast quadrant. On Trial 7, the platform was moved beneath the cue in the southeast quadrant. On the second test day, one week later, Test Trials 9 through 14 were run. The conditions of Trials 9 and 10 were replications of Trials 7 and 8. On Trials 1 I and 12, the cue was removed from the pool. On Trials 13 and 14, the platform was moved to the northeast quadrant. As in the first experiment, animals were released in each trial from one of four poles in a pseudorandom sequence. Escape latency (and whether or not the animal touched the cue during Trials 5 and 6 when the cue was located in an incorrect quadrant) were recorded. When the rat encountered the platform it was permitted to remain there for 30 sec. If the platform was not located within 90 sec, the trial was terminated and the rat was removed. The intertrial intervals were again 3-5 min. This sequence of trials allowed the testing of a number of different aspects of spatial memory and perception. First trials occurred in which the cue was present and tested the animal’s attention to local and distal visual aspects of the environment (e.g., Trials 1-4). The shifting of the proximal cue tested the animal’s reliance on the local cue to

EARLY EXPERIENCE AND SPATIAL LEARNING

435

100 I

90 -

@

80 70-

m- Complex Reared

Pool Quadrants

8

?

Light Reared

-6 Dark

’,

,~ , ,,’

60 -

01 0

1

Legend

,

I

,

I

1

2

3

4

1/1

111

1/1

1/1

.

,

, \

,

Reared

I I

i’’,.,

,

,

,

,

,

,

,

,

,

8 6 7 8 9 1 0 1 1 1 2 1 3 1 4 1 6 3/1 3/1 3/3 3/3 3/3 3/3 -/3 -13 -14 -14

Trial cue posltlon/platforrn positlon

Fig. 2. Latency to reach the submerged platform on each trial for complex-reared (CR), lightreared (LR), and dark-reared (DR) rats.

find the platform (e.g., Trials 5 and 6). The repositioning of the platform tested the speed to which new distal (and old local) visual cues could be utilized to navigate to a new location (i.e., Trials 7 and 8). The week delay represented a test of retention (e.g., Trial 9). Trials 11 and 12 (no local cue) represented a probe of the role played by the local visual cue and 13 and 14 represented a test of the subject’s ability to locate the new shifted location of the platform without the presence of a local visual cue.

Results A repeated measures ANOVA was used to assess differences in escape latencies between the 3 groups (LR, DR, CR) of animals across the 8 trials of the first day of testing. Early stimulation history did have a significant impact on overall performance [F(2,42) = 55.7, p < .000l)]. There were also a significant main effect for Trial [F(7,294) = 37.0, p < .001] as well as a Rearing Condition x Trial interaction [F(14,294) = 6.1, p < .0001]. Posthoc testing (Tukey’s HSD, all p’s = .05) revealed that CR animals had significantly faster escape latencies than either LR or DR rats on the first trial (see Figure 2). Performance improved significantly over the next 3 trials for animals in all of the groups and there were no differences due to rearing conditions on Trials 2 , 3, and 4. On Trial 5, the proximal visual cue moved to another (SW) location, while the invisible platform remained in its original (NE) location. The performance of the DR rats was significantly inferior to that of the CR and LR groups (Tukey’s HSD, p = .05). The performance of the CR and LR rats was not adversely affected by the change ( M ’ s = 8.1 + 13.4 sec) while that of the DR rats was disrupted ( M = 65.7 sec). This disruption was also evident in terms of the percentage

436

TEES, BUHRMANN, AND HANLEY

(DK = 8696, LR = 60%, CR = 33%) of animals coming into contact with the proximal visual cue placed in the SW quadrant [X(2) = 8.8, p < .021. On Trial 6, administered 3-5 minutes later under identical conditions, the CR and LR rats ( M s = 9.3 and 14.7 sec) continued to reach the platform faster than the DR rats ( M = 5 I .7 sec). On Trial 7 the platform was shifted to the quadrant containing the proximal cue. The CR rats found the platform more rapidly ( M = 15.1 1 sec) than the LR rats ( M = 31.0) who in turn were faster than the DR rats ( M = 61.6 sec). On Trial 8 (repeat) only the DR rats ( M = 22.8 sec) continued to be significantly slower than both the CR and LR rats ( M = 9.3 and 9.6 sec). One week later, Trials 9-14 were given and the ANOVA-revealed effects of the Rearing [F(2,42) = 14.8, p < .001], Trials [F(5,210) = 12.0, p < .001], and the interaction between Rearing X Trial [F(10,210) = 4.8, p < .001] were all significant. Post-hoc testing (Tukey’s HSD, allp’s = .05) revealed that, overall, CR and LR had significantly faster escape latencies than DR rats on Trials 9 and 10. On Trials 11 and 12, the visual cue was removed and again CR and LR rats were significantly faster than DR rats in reaching the platform safely. Finally, on Trials 13 and 14 the platform was moved to quadrant 4 and no proximal cue was available. N o significant differences were found. Obviously, it took animals longer to relocate the platform on Trial 13 ( M = 29.8 sec). Their performance on Trial 14 (repeat) was significantly better ( M = 13.4 sec). To summarize, this version of the water-based test of spatial learninghemory yielded significantly better performance by CR animals than LR rats on Trials 1 and 7. Moreover, DR rats performed less well than either LR or CR animals on Trials 1 . 5 , 6, 7, 8, 9, 10, 11, and 12.

Discussion In Experiment I, the visually experienced LR (but not DR) rats did benefit from viewing the test pool and room from the correct platform location. However, their performance, while significantly better than the other LR and DR groups, cannot be characterized either as accurate or involving instantaneous transfer. Our study does support the notion that visually experienced LR rats can make somewhat better use of latently learned spatial information acquired while sampling environmental cues from a vantage point of the correct escape platform location (prior to any participation in swimming in the cool water of the pool). While these visually experienced animals did not show instantaneous transfer of place navigation, they did show a small but significant difference on their first opportunity to swim to the invisible platform, as have been demonstrated by others (Sutherland & Linggard, 1982; Keith & McVety, 1988; Keith, 1989). The effect is not large and it is understandable that some researchers have failed to find this effect in other tests in the water maze (Jacobs, Zaborowski, & Whishaw, 1989). There is certainly evidence available in other experimental (dry land) situations for both latent spatial and nonspatial learning (e.g., Chamizo & Mackintosh, 1989; Tees, 1986). While our results provided confirmation of latent spatial learning in the Morris Water Maze, the argument that these visually experienced LR animals recognized and used the information about the correct platform location and its relationship to distal cues in the room would have been more persuasive if we were able to demonstrate not only positive transfer, but also negative transfer for

EARLY EXPERIENCE AND SPATIAL LEARNING

437

animals being tested under Condition B (Incorrect Location) in comparison to animals (Condition C) placed on a platform in an entirely different environment. See Church & Meck (1983) and Tees and Buhrmann (1989) for further discussions. While we did not observe differences between LR and DR rats’ behavior (e.g., in rearing) during preexposure, we are currently beginning to do a more sophisticated analysis of the movements of differentially reared and lesioned animals during such preexposure. The LR and DR rats tested under all three conditions improved significantly over subsequent trials on that first day of testing. All of the animals showed evidence that they were able to remember (after 3-5 minutes) the correct location of the invisible platform. On the first probe trial (Trial 5 ) all of the rats navigated to their former location and began searching that quadrant first. There were no differences between LR and DR groups in terms of their behavior in that situation. However, there were differences due to early visual rearing conditions on a retest one month later. The performance then of the 3 DR groups was less accurate as far as remembering the correct location of the platform in comparison to animals in the LR groups. On the basis of performance on land-based mazes, Greenough, Wood, and Madden (1972) suggested that mice reared under enriched conditions were able to process or store information more quickly as the mice raised in standard colony conditions. We made a similar argument in connection with visually deprived rats (Tees, 1984; Tees, Midgley, & Nesbit, 1981), suggesting that visually deprived animals process and store information about spatial arrangements in the 17-arm mazes at a slower rate. In our earlier analysis (e.g., Tees, 1984), we showed that both reference memory and working memory are adversely affected by lack of visual experience using a 17-arm radial arm maze. In this initial experiment, the first trial one month after initial training seemed to provide evidence that the visually inexperienced animal’s reference memory for the correct location relative to distal cues in the environment was more fragile, more susceptible to interference, and/or less well developed than that of their visually experienced counterparts. Performance on subsequent trials made it clear that the DR animals are as able to remember as the LR animals minutes after a triallexperience. Sutherland and Dyck (1984), in an early, exhaustive analysis describing characteristics of place navigation for normally reared rats, showed that with 40 trials (over 8 days of training), rats remained accurate on a test of retention after a 14 day retention interval. They indicated they had retested some of the animals after 3 months with comparable success. Our test involved less initial training with the excellent retention after 30-35 days. As Sutherland and Dyck suggested, spatial map information is retained with no apparent decline for weeks or months at least in normally reared rats. Overall, the results of this experiment indicate that the latent spatial learning paradigm is capable of illustrating some deficits in spatial competence and memory in visually deprived rats. They also reveal DR rats remain spatially competent enough to be successful in this test of place learning. Perhaps their effectiveness is not surprising in light of early emergence of spatial competence in the 23-dayold rat when tested in water maze (Rudy et al., 1987), prior to much in the way of visual experience. In our second experiment our DR rats had significantly more difficulty not only initially but also with most of the “switches” in the test conditions. Moving

438

TEES, BUHRMANN, AND HANLEY

the proximal cue, moving the platform, and removing the proximal cue had an adverse effect on their ability to navigate directly to the platform’s location relative to their CR and LR counterparts. (They also were less accurate after a week’s delay.) Further experimentation is needed to answer the question of whether the presence of the proximal cue itself during the first four trials reduced the effectiveness of the DR rats to learn about the location of the platform and its relationship to distal cues. It should be evident that the deficits in the performance we report here cannot (yet) be characterized simply as reflective of simple impairments in strategy selection or (spatial) reference memory. It is clear that any further progress in understanding the nature of the role of visual experience in the ontogeny of spatial competence and memory will require more detailed analysis of the movements of differentially reared rats both during swimming and on the platform itself (e.g., Whishaw & Mittleman, 1986) and the use of selective neurochemical interventions and cortical lesions to test hypotheses about which processes are affected by early stimulation history (Whishaw & Petrie, 1988). Finally, our complex reared subjects were significantly better than LR and DR rats on the initial test and on the first switch (Trial 7) of the platform’s location (with the proximal cue). However, overall this test of spatial competence did not appear as sensitive as other (land-based) behavioral tests (Juraska, 1990) of the effects of complex rearing and we are engaged in further work establishing more sensitive and more difficult water-maze-based assays of the impact of CR on spatial navigation.

Notes This investigation was supported by the Natural Sciences and Engineering Council of Canada Grant AP0179 to Richard C. Tees and an NSERC Predoctoral Fellowship to K. Buhrmann. The assistance of those testing animals, Mike Dunbar, Karen Steichele, Mike Laycock, Rafael Daudet, and Lucille Hoover is gratefully acknowledged.

References Buhrmann, K., Tees, R. C., & Symons, L. A. (1989). The effect of early environmental enrichment on dewhiskered rats. Society for Neuroscience Ahsiract, 15, 1050. Chamizo, V. D., & Mackintosh, N. J . (1989). Latent learning and latent inhibition in maze discriminations. The Quarierly Journul of Experimentul Psychology, 4 / B , 2 1-3 1. Church. R. M., & Meck, W.H. (1983). Acquisition and cross-modal transfer of classification rules for temporal intervals. In M. L. Commons, R. J. Herrnstein, & A. R. Wagner (Eds.). Qucrntittriivc. unrr1y.si.s qf behavior: Discrimination processes (Vol. 4, pp. 75-97). Cambridge, MA: Ballinger. Forgus. R. H. (1954). The effect of early perceptual learning on the behavioral organization of adult rats. Journal q f Comp. Physiologicul Psychology 47, 33 1-336. Gage, F. H . , Kelly, P. A. T., & Bjrklund, A. (1984). Regional changes in brain glucose metabolism reflect cognitive impairments in aged rats. Journal of Neuro.science, 1 1 , 2856-2865. Gamboni, W. R . (1964). Visual deprivation and hypothesis behavior in rats. Perceptuul und Moior Skills, 19, 501-502. Gibson. E. J . (1969). Principles uf percepiiiul learning and development. New York: Appleton. Goodale. M. A . , & Carey, D. P. (1990). The role of cerebral cortex in visuomotor control. In B. Kolb & R. C . Tees (Eds.), Cerebral coriex ufthe raf (pp. 309-340). Cambridge: MIT Press.

EARLY EXPERIENCE AND SPATIAL LEARNING

439

Gottlieb, G. (1971). Ontogenesis of sensory function in birds and mammals. In E. Tobach, Z. R. Aronson, & E . Shaw (Eds.), The biopsychology of development (pp. 67-128). New York: Academic Press. Greenough, W. T., & Green, E. J. (1981). Experience and the changing brain. In J. L. McGaugh & S. B. Kiesler (Eds.), Aging: Biology and behavior (pp. 159-200). New York: Academic Press. Greenough, W. T., Wood, W. E., & Madden, T. C. (1972). Possible memory storage differences among mice reared in environments varying in complexity. Behavioral Biology, 7, 717-722. Jacobs, W. J., Zaborowski, J., & Whishaw, I. Q. (1989). Failure to find latent spatial learning in the Morris Water Task: Retraction of Jacobs, Zaborowski, and Whishaw. Journal ofExperimental Psychology: Animal Behavior Processes, 15, 286. Juraska, J. (1990). The structure of the rat cerebral cortex: Effects of gender and the environment. In B. Kolb & R. C. Tees (Eds), Cerebral cortex o f t h e rat (pp. 483-506). Cambridge: MIT Press. Keith, J. R. (1989). Does latent learning produce instantaneous transfer of place navigation? Psychobiob ogy, 17, 210-211. Keith, J. R., & McVety, K. M. (1988). Latent place learning in a novel environment and the influences of prior training in rats. Psychobiology, 16, 146-151. Morris, R. G. M. (1981). Spatial location does not require the presence of local cues. Learning and Motivation. 12, 239-260. Rosenzweig, M. R., Love, W., & Bennett, E. L. (1968). Effects of a few hours a day of enriched experience on brain chemistry and brain weights. Physiology and Behavior, 3, 819-825. Rudy, J. W., Stadler-Morris, S., & Albert, P. (1987). Ontogeny of spatial navigation behaviors in the rat: Dissociation of “proximal” and “distal”-cue-based behaviors. Behavioral Neuroscience. 101, 62-73. Spigelman, M. N., & Bryden, M. P. (1967). Effects of early and late blindness on auditory spatial learning in the rat. Neuropsychologia, 5 , 267-274. Sutherland, R. J., & Dyck, R. H. (1984). Place navigation by rats in a swimming pool. Canadian Journal of Psychology, 38, 322-347. Sutherland, R. J., & Linggard, R. (1982). Behavioral and Neural Biology, 36, 103-107. Sutherland, R. J., Chew, G. L., Baker, J. C.. & Linggard, R. C. (1987). Some limitations on the use of distal cues in place navigation by rats. Psychobiology, 15, 4448-4457. Sutherland, R. J., Kolb, B . , & Whishaw, I. Q. (1982). Spatial mapping: Definitive disruption by hippocampal or medial frontal cortical damage in the rat. Neuroscience Letters, 31, 271-276. Tees, R. C. (1968). Effect of early restriction on later form discrimination in the rat. Canadian Journal of Psycholog>),22, 294-298. Tees, R. C. (1984). Visual experience, unilateral cortical lesions, and lateralization of function in rats. Behavioral Neuroscience, 98, 969-978. Tees, R. C. (1986). Experience and visual development: Behavioral evidence. In W. T . Greenough & J . M. Juraska (Eds.), Deuelopmental NeuroPsychobiology (pp. 317-36). New York: Academic Press. Tees, R. C. (1990). Experience, perceptual competences and rat cortex. In B. Kolb & R. C. Tees (Eds.), The cerebral cortex of the rat (pp. 507-536). Cambridge: MIT Press. Tees, R. C., & Buhrmann, K. (1989). Parallel perceptuaUcognitive functions in humans and rats: Space and time. Canadian Journal of Psychology. 43, 266-285. Tees, R. C., & Symons, L. A. (1987). Intersensory coordination and the effects of early sensory deprivation. Developmental Psychobiology, 20, 497-507. Tees, R. C . , Midgley, G., & Nesbit, J. C. (1981). The effect of early visual experience on spatial maze learning in rats. Developmental Psychobiology, 14, 425-438. Turkewitz, G . , & Kenny, P. A. (1982). Limitations on input as a basis for neural organization and perceptual development: A preliminary theoretical statement. Developmenral Psychobiology, 15, 357-358. Whishaw, I . Q., & Petrie, B. F. (1988). Cholinergic blockade in the rat impairs strategy selection but not learning and retention of nonspatial visual discrimination problems in a swimming pool. Behavioral Neuroscience, 102, 662-677. Whishaw, I. Q . , & Mittleman, G . (1986). Visits to starts, routes and places by rats (Rattus norvegicus) in swimming pool navigation tasks. Journal qf Comparative Psychology, 100. 422-43 1.