Pigeons (Columba livia) plan future moves on computerized maze tasks

Planning, the internal process of formulating an organized method about one's future behavior, should be advantageous for non-human animals as well as for humans. However, little is known about this process in avian species. We examined planning processes in pigeons (Columba livia) using a computerized maze task. In Experiment 1, we found that the pigeons plan their next one step, and in some cases even correctly adjust their actions after change of goal locations, while performing on a plus-shaped maze. We also showed that the pigeons might even plan two steps on familiar, well-practiced mazes. In Experiment 2, we discovered that the subjects plan the direction they would go first before starting to solve a four-arm shuriken (a Japanese traditional throwing knife)-shaped maze. The birds also corrected their previously planned actions after change of goal locations. Our results from these experiments suggest that planning ahead is within the cognitive capacity of a "bird brain", and that it may be more widespread in the animal kingdom than has been presumed.     

Macaques’ (Macaca mulatta) use of numerical cues in maze trials

We tested the ability of number-trained rhesus monkeys to use Arabic numeral cues to discriminate between different series of maze trials and anticipate the final trial in each series. The monkeys prior experience with numerals also allowed us to investigate spontaneous transfer between series. A total of four monkeys were tested in two experiments. In both experiments, the monkeys were trained on a computerized task consisting of three reinforced maze trials followed by one nonreinforced trial. The goal of the maze was an Arabic numeral 3, which corresponded to the number of reinforced maze trials in the series. In experiment 1 (n=2), the monkeys were given probe trials of the numerals 2 and 4 and in experiment 2 (n=2), they were given probe trials of the numerals 2–8. The monkeys receiving the probe trials 2 and 4 showed some generalization to the new numerals and developed a pattern of performing more slowly on the nonreinforced trial than the reinforced trial before it for most series, indicating the use of the changing numeral cues to anticipate the nonreinforced trial. The monkeys receiving probe trials of the numerals 2–8 did not predict precisely when the nonreinforced trial would occur in each series, but they did incorporate the changing numerals into their strategy for performing the task. This study provides the first evidence that number-trained monkeys can use Arabic numerals to perform a task involving sequential presentations.     

Strategic navigation of two-dimensional alley mazes: comparing capuchin monkeys and chimpanzees

Planning is an important component of cognition that contributes, for example, to efficient movement through space. In the current study we presented novel two-dimensional alley mazes to four chimpanzees and three capuchin monkeys to identify the nature and efficiency of planning in relation to varying task parameters. All the subjects solved more mazes without error than expected by chance, providing compelling evidence that both species planned their choices in some manner. The probability of making a correct choice on mazes designed to be more demanding and presented later in the testing series was higher than on earlier, simpler mazes (chimpanzees), or unchanged (capuchin monkeys), suggesting microdevelopment of strategic choice. Structural properties of the mazes affected both species' choices. Capuchin monkeys were less likely than chimpanzees to take a correct path that initially led away from the goal but that eventually led to the goal. Chimpanzees were more likely to make an error by passing a correct path than by turning onto a wrong path. Chimpanzees and one capuchin made more errors on choices farther in sequence from the goal. Each species corrected errors before running into the end of an alley in approximately 40% of cases. Together, these findings suggest nascent planning abilities in each species, and the prospect for significant development of strategic planning capabilities on tasks presenting multiple simultaneous or sequential spatial relations. The computerized maze paradigm appears well suited to investigate movement planning and spatial perception in human and nonhuman primates alike.     

Navigating two-dimensional mazes: chimpanzees (Pan troglodytes) and capuchins (Cebus apella sp.) profit from experience differently

We examined whether navigation is impacted by experience in two species of nonhuman primates. Five chimpanzees (Pan troglodytes) and seven capuchin monkeys (Cebus apella) navigated a cursor, using a joystick, through two-dimensional mazes presented on a computer monitor. Subjects completed 192 mazes, each one time. Each maze contained one to five choices, and in up to three of these choices, the correct path required moving the cursor away from the Euclidean direction toward the goal. Some subjects completed these mazes in a random order (Random group); others in a fixed order by ascending number of choices and ascending number of turns away from goal (Ordered group). Chimpanzees in both groups performed equivalently, demonstrated fewer errors and a higher rate of self-correcting errors with increasing experience at solving the mazes, and made significantly fewer errors than capuchin monkeys. Capuchins were more sensitive to the mode of presentation than chimpanzees; monkeys in the Ordered group made fewer errors than monkeys in the Random group. However, capuchins’ performance across testing changed little, and they remained particularly susceptible to making errors when the correct path required moving away from the goal. Thus, these two species responded differently to the same spatial challenges and same learning contexts. The findings indicate that chimpanzees have a strong advantage in this task compared to capuchins, no matter how the task is presented. We suggest that differences between the species in the dynamic organization of attention and motor processes contribute to their differences in performance on this task, and predict similar differences in other tasks requiring, as this one does, sustained attention to a dynamic visual display and self-produced movements variably towards and away from a goal.     

A model of working memory capacity in the radial-arm maze task

The radial arm maze is one of the most commonly used tests for assessing working memory in laboratory animals. However, to date, there exists no quantitative method of estimating working memory capacity from performance on this task. Here, we present a mathematical model of performance on the radial arm maze from which we can derive estimates of capacity. We derive explicit results for the two most commonly used measures of performance as functions of number of arms in the maze and memory capacity, assuming a uniform random search. We simulate random non-uniform search strategies. Comparing our model to previous experiments, we show that our model predicts a working memory capacity in the range of 3-9 at the level of performance observed in these experiments. This estimate is within the typical estimate of human working memory capacity. Performance of rats on large mazes (e.g. 48 arms) has been used as evidence that the working memory capacity of rats may be significantly larger than that of humans. We report that memory capacity in the range 3-9 is sufficient to explain the performance of rats in very large radial mazes. Furthermore, when we simulate non-uniform random search strategies observed in the experiments, the resulting estimates do not differ significantly from those assuming a uniform random search. We conclude that a list-based model of working memory with modest capacity is more powerful than previously expected.     

Spatial navigation in complex and radial mazes in APP23 animals and neurotrophin signaling as a biological marker of early impairment

Impairment of hippocampal function precedes frontal and parietal cortex impairment in human Alzheimer's disease (AD). Neurotrophins are critical for behavioral performance and neuronal survival in AD. We used complex and radial mazes to assess spatial orientation and learning in wild-type and B6-Tg(ThylAPP)23Sdz (APP23) animals, a transgenic mouse model of AD. We also assessed brain content of nerve growth factor (NGF), brain-derived neurotrophic factor (BDNF), and neurotrophin-3 (NT-3). Performance was alike in wild-type and APP23 animals in the radial maze. In contrast, performance in the complex maze was better in wild-type than APP23 animals. Contrary to the wild-type, hippocampal BDNF levels decreased on training in APP23 animals. Hippocampal and frontal cortex NGF levels in APP23 animals correlated with the time to solve the complex maze, but correlated inversely with escape time in wild-type animals. NT-3 levels were alike in wild-type and APP23 animals and were unchanged even after training. Both types of mazes depend on hippocampal integrity to some extent. However, according to the cognitive mapping theory of spatial learning, the complex maze because of the increased complexity of the environment most likely depends more strongly on preserved hippocampal function than the radial maze in the working memory configuration applied here. Greater impairment in complex maze performance than in radial maze performance thus resembles the predominant affliction of the loss of hippocampal function in human AD. NGF and BDNF levels on maze learning are different in wild-type and transgenic animals, indicating that biological markers of AD may be altered on challenge even though equilibrium levels are alike.     

Algorithmic responding on the radial maze in rats does not always imply absence of spatial encoding

Two experiments were conducted to determine whether consistent algorithmic response patterning on 8- and 10-arm versions of the radial maze is independent of spatial encoding. On the 8-arm version well-trained hooded rats were tested in darkness, after maze rotation that rendered room cues ambiguous with respect to arm positions, or with room cues unsystematically relocated. Ambiguous maze rotation was also used with well-trained subjects on the 10-arm version. If algorithmic patterning is a learned, non-spatial strategy, animals using it consistently ought not to have been affected by changes in the spatial layout of the test environment, and the type of pattern used by each subject would have remained constant. On the 8-arm radial maze, responses were most often made to arms 2 or 3 from that just visited. In many animals patterns were interchangeable, switching occurring between preferred angles of turn from day to day. Performance fell when animals were tested in darkness and upon ambiguous maze rotation early (but not later) in training. Testing in darkness increased the angle through which animals turned when responding, perhaps due to the disturbance of intramaze cue use. On the 10-arm maze the “consecutive arm” pattern was used persistently by several animals and appeared to protect their performance from disruption by ambiguous maze rotation. Animals not using rigid patterning were adversely affected. However, on both mazes animals using patterning correctly identified maze arms that had been omitted from otherwise patterned choice sequences. Animals adopted continuous patterning only when spatial encoding had been established. Response patterning appears to serve a mnemonic function and in rats complements rather than replaces the use of a spatial representation of the environment. It was concluded that a complex, flexible relationship exists between spatial functioning and its expression via motor responses.     

Maze navigation by honeybees: learning path regularity

We investigated the ability of honeybees to learn mazes of four types: constant-turn mazes, in which the appropriate turn is always in the same direction in each decision chamber; zig-zag mazes, in which the appropriate turn is alternately left and right in successive decision chambers; irregular mazes, in which there is no readily apparent pattern to the turns; and variable irregular mazes, in which the bees were trained to learn several irregular mazes simultaneously. The bees were able to learn to navigate all four types of maze. Performance was best in the constant-turn mazes, somewhat poorer in the zig-zag mazes, poorer still in the irregular mazes, and poorest in the variable irregular mazes. These results demonstrate that bees do not navigate such mazes simply by memorizing the entire sequence of appropriate turns. Rather, performance in the various configurations depends on the existence of regularity in the structure of the maze and on the ease with which this regularity is recognized and learned.     

Acquisition of navigation by chimpanzees (Pan troglodytes) in an automated fingermaze task

These experiments investigated how chimpanzees learn to navigate visual fingermazes presented on a touch monitor. The aim was to determine whether training the subjects to solve several different mazes would establish a generalized map-reading skill such that they would solve new mazes correctly on the first presentation. In experiment 1, two captive adult female chimpanzees were trained to move a visual object (a ball) with a finger over the monitor surface toward a target through a grid of obstacles that formed a maze. The task was fully automated with storage of movement paths on individual trials. Training progressed from very simple mazes with one obstacle to complex mazes with several obstacles. The subjects learned to move the ball to the target in a curved path so as to avoid obstacles and blind alleys. After training on several mazes, both subjects developed a high level of efficiency in moving the ball to the target in a path that closely approached the ideal shortest path. New mazes were then presented to determine whether the subjects had acquired a more generalized maze-solving performance. The subjects solved 65–100% of the new mazes the first time they were presented by moving the ball around obstacles to the target without making detours into blind alleys. In experiment 2, one of the chimpanzees was trained using mazes with two routes to the target. One of the routes was blocked at one of many possible locations. After training to avoid the blind alley in different mazes, new mazes were presented that also had one route blocked. The subject correctly solved 90.7% of the novel mazes. When the mazes had one short and one long open route to the target the subject preferred the shorter route. When the short route was blocked, the subject solved only 53.3% of the mazes because of the preference for the shorter route even when blocked. The overall results suggest that with the training methods used the subjects learned to solve specific mazes with a trial-and-error method. Although both subjects were able to solve many of the novel mazes they did not fully develop a more general "map-reading" skill. Accepted after revision: 30 July 2001 Electronic Publication     

Effects of neocortical ectopias and environmental enrichment on Hebb-Williams maze learning in BXSB mice

Approximately 40-60% of BXSB mice have neocortical ectopias, a developmental anomaly characterized by migration of neurons into the neuron-sparse layer I of cortex. Previous studies have shown that ectopic BXSB mice have superior reference, but inferior working, memory on spatial tasks. Female BXSB mice were housed either in an enriched environment or in standard cages at weaning. Subsequently, these animals were tested on four of the Hebb-Williams mazes in a water-based version of this maze. Theoretically, two of the maze configurations placed greater emphasis on reference memory to find the goal, whereas the other two favored working memory. Ectopics reared in standard housing conditions were better than nonectopics on mazes that favored the use of reference memory, but poorer on mazes that favored working memory. In contrast, subjects raised in the enriched environment showed no ectopia differences. A comparison of enriched and standard housing conditions found that the enriched animals had better reference memory but poorer working memory. The latter effect may be because the enriched environment, although more stimulating, did not change in time or space; and other researchers have shown that daily replacement of stimuli in complex environments is correlated with better working memory.     

Perceptual completion in rhesus monkeys (Macaca mulatta) and pigeons (Columbia livia)

In a two-dimensional drawing, when the narrow edge of a bar appears to touch the edge of a large rectangle, humans overestimate the length of the bar (Kanizsa, 1979). Kanizsa has suggested that this illusion occurs because humans perceive the bar as continuing behind the rectangle and complete the "occluded" portion of the bar. Rhesus monkeys and pigeons were trained to classify black target bars with a variety of lengths as "long" or "short." In training, the bar was always located at the same distance from a gray box. After learning this discrimination, the subjects were tested on novel stimuli, in which the bar was located at three new locations. Monkeys showed a consistent response bias for "long" when the bar touched the box, but pigeons did not. Monkeys appear to have completed the "occluded" part like humans, whereas pigeons failed to do so. Because this procedure does not require animals to complete the "occluded" part with any particular form, their failure suggests that pigeons do not even perceive the target bar as continuing behind the "occluding" figure. The failure of pigeons may be due to difficulty in perceiving depth from two-dimensional drawings.     

Perceptual completion in rhesus monkeys (Macaca mulatta) and pigeons (Columba livia)

In a two-dimensional drawing, when the narrow edge of a bar appears to touch the edge of a large rectangle, humans overestimate the length of the bar (Kanizsa, 1979). Kanizsa has suggested that this illusion occurs because humans perceive the bar as continuing behind the rectangle and complete the “occluded” portion of the bar. Rhesus monkeys and pigeons were trained to classify black target bars with a variety of lengths as “long” or “short.” In training, the bar was always located at the same distance from a gray box. After learning this discrimination, the subjects were tested on novel stimuli, in which the bar was located at three new locations. Monkeys showed a consistent response bias for “long” when the bar touched the box, but pigeons did not. Monkeys appear to have completed the “occluded” part like humans, whereas pigeons failed to do so. Because this procedure does not require animals to complete the “occluded” part with any particular form, their failure suggests that pigeons do not even perceive the target bar as continuing behind the “occluding” figure. The failure of pigeons may be due to difficulty in perceiving depth from two-dimensional drawings.     

Rats use a sense of direction to alternate on T-mazes located in adjacent rooms

Lister hooded rats were trained on a forced-sample T-maze alternation task in an environment lacking spatial landmarks. An early study of spontaneous alternation on the T-maze had shown that rats use a "spatial sense" to select alternate maze arms across mazes. As this phenomenon may provide a useful tool for studying the neural substrates of a directional sense, we wished to confirm this finding on a different version of the T-maze task, with well-trained animals. We found that rats successfully selected the appropriate maze arm when the choice phase of the task was presented on a second maze, oriented in the same direction, and located in an adjacent room. However, choice performance fell to chance level when the second maze was oriented 90° relative to the first. This result suggests that the rats do not simply alternate turns across the two environments, but rather that they rely on a sense of direction that is carried across environments. Electronic Publication     

Gerbils in space: performance on the 17-arm radial maze.

In Experiment 1 six hungry gerbils received six trials per day on a 17-arm radial maze. During each trial the subjects were allowed to choose freely among the arms, each of which contained a food pellet, until each arm had been visited once or until eight minutes had elapsed. An error was recorded when the subject entered a previously visited arm. The gerbils quickly learned not to re-enter previously visited arms and generally made errors on fewer than 15% of entries, performance comparable to that of the rat and superior to that of other species tested in the radial arm maze. The intertrial-interval duration did not affect accuracy of arm choices during acquisition but did influence asymptotic accuracy. Accuracy did not change systematically over the six trials. A high proportion of arm entries were to nearby arms. Errors occurred most often towards the end of a trial. Odor cues were not important. When the number of trials per day was reduced from six to one, accuracy deteriorated slightly. In Experiment 2 neither the transposition of extramaze cues nor the placement of the maze in a different room had large disruptive effects on accuracy. In Experiment 3 the addition of three explicit intramaze brightness cues aided accuracy, perhaps by permitting the subjects to decompose the large maze into three smaller mazes, although there was no direct evidence that this was the case. Implications of a number of these results for models of spatial maze performance were discussed.     

Caudate nucleus-dependent response strategies in a virtual navigation task are associated with lower basal cortisol and impaired episodic memory

The present research examined the relationship between endogenous glucocorticoids, navigational strategies in a virtual navigation task, and performance on standard neuropsychological assessments of memory. Healthy young adult participants (N = 66, mean age: 21.7) were tested on the 4 on 8 virtual maze (4/8 VM) and standard neuropsychological tests such as the Rey-Osterrieth Complex Figure (RO) and the Rey Auditory Verbal Learning Task (RAVLT), which measure episodic memory. The 4/8 VM differentiates between navigational strategies, where participants either use a hippocampal-dependent spatial strategy by building relationships between landmarks, or a caudate nucleus-dependent stimulus–response strategy by automatizing a pattern of open and closed arms to learn the location of objects within the maze. Degree of stress was assessed by administering the Perceived Stress Scale (PSS) questionnaire. Cortisol samples were taken on two consecutive days upon waking, 30 min after waking, at 11 am, 4 pm, and 9 pm. There was a significant difference in basal levels of cortisol between spatial and response learners. Interestingly, response learners had significantly lower cortisol levels throughout the day. The two groups did not differ in terms of perceived stress as measured with the PSS questionnaire. Moreover, there was no significant correlation between PSS scores and salivary cortisol levels, indicating that the higher cortisol levels in the spatial group were not associated with greater perceived stress. In addition, participants who spontaneously used a spatial strategy performed significantly better on the RAVLT and RO. These data indicate that the cortisol levels in the spatial group may be optimal in terms of episodic memory performance whereas the cortisol levels in the response group may be associated with poorer memory. These results are suggestive of an inverted U-shaped curve describing the effects of basal levels of circulating cortisol on memory in young adults.     

Spatial vision meets spatial cognition: examining the effect of visual blur on human visually guided route learning

Visual navigation is a task that involves processing two-dimensional light patterns on the retinas to obtain knowledge of how to move through a three-dimensional environment. Therefore, modifying the basic characteristics of the two-dimensional information provided to navigators should have important and informative effects on how they navigate. Despite this, few basic research studies have examined the effects of systematically modifying the available levels of spatial visual detail on navigation performance. In this study, we tested the effects of a range of visual blur levels--approximately equivalent to various degrees of low-pass spatial frequency filtering--on participants' visually guided route-learning performance using desktop virtual renderings of the Hebb-Williams mazes. Our findings show that the function of blur and time to finish the mazes follows a sigmoidal pattern, with the inflection point around +2 D of experienced defocus. This suggests that visually guided route learning is fairly robust to blur, with the threshold level being just above the limit for legal blindness. These findings have implications for models of route learning, as well as for practical situations in which humans must navigate under conditions of blur.     

Disentangling component learning and executive processes in hidden pathway maze learning in children: A process-based approach

Identification of cognitive processes that affect children’s ability to manage complex information is critical to understanding the development of executive functions. However, characterization of these processes is hampered by a lack of appropriate tasks and reliance on single outcome measures that are unsuitable for studying complex aspects of executive function. The current study aimed to circumvent these limitations by employing a hidden maze learning paradigm (Groton Maze Learning Test; independence of component cognitive GMLT) to evaluate the processes—spatial memory and rule use—that underlie hidden pathway maze learning in children. Specifically, we investigated the impact of withholding rule instructions (Study 1) and nonrepeating pathways on each trial (Study 2) on the ability to use rules and to locate pathways in 10 × 10 mazes in a sample of 8-year-old children. Results of these studies suggested that manipulations of task rules did not affect spatial memory and that manipulations of the maze pathway did not affect rule use. These findings demonstrate the independence of spatial memory and rule use on the GMLT and provide evidence of a “double dissociation” of cognitive processes that underlie hidden maze learning in children. Implications for understanding the coordination of component cognitive processes that underlie executive function in childhood are discussed.     

Anxiolytic effects of repeated victory in male Wistar rats

Recurring evidence suggests that social stress has anxiogenic‐like effects in laboratory rodents. However, despite the fact that competitive situations are stressful, success in competitive situations reduces anxiety in humans. The aim of the present study was to investigate whether repeated experience of winning in aggressive encounters affects anxiety measures in laboratory rodents. Male rats were housed together with a female for 2 weeks. Cohabitation with females was necessary to induce high levels of aggressiveness in these animals. During the second week, half of the male rats were exposed daily for 30 min to an intruder of smaller size, and the other half remained undisturbed in their home cage. Group‐housed male rats were also used as controls. Residents attacked and defeated intruders, who did not retaliate. After the fifth encounter, all animals were tested for anxiety on the elevated plus‐maze. Repeated victory lowered anxiety measures considerably, despite the fact that aggressive encounters are stressful even for the victor. It is concluded that repeated victory has an anxiolytic action in male rats. The similarities with human data suggest that repeated winning in rats can be used as a laboratory model for success‐induced changes in human anxiety. Aggr. Behav. 26:257–261, 2000. © 2000 Wiley‐Liss, Inc.     

Analyzing the path of responding in maze-solving and other tasks

Response time and accuracy are sensitive measures of overall performance but may mask underlying response strategies. For example, analysis of latency and accuracy measures produced in a computerized-maze task does not reveal whether rhesus monkeys really “solve a maze” or simply move as much as is possible toward the target, negotiating barriers through trial and error as they encounter them. Regression procedures are described for analyzing response path against several hypothetical response curves, and analyses of response path for rhesus monkeys’ performance on the computerized MAZE task are presented as an illustration. The data suggest that rhesus monkeys do invoke a response strategy of solving the maze, because the observed response topography is significantly associated with the optimal path of responding. Many experimental paradigms should similarly benefit from analysis of the response paths that subjects exhibit.     

Within-session reversal learning in rhesus macaques (<Emphasis Type="Italic">Macaca mulatta</Emphasis>)

In a midsession reversal (MSR) task, animals are typically presented with a simple, simultaneous discrimination (S1+, S2?) where contingencies are reversed (S1?, S2+) half-way through each session. This paradigm creates multiple, relevant cues that can aid in maximizing overall reinforcement. Recent research has shown that pigeons show systematic anticipatory and perseverative errors across the session, which increase as a function of proximity to the reversal trial. This behavior has been theorized to indicate primary control by temporal cues across the session, instead of the cues provided by recent reinforcement history that appear to control behavior shown by humans. Rats, however, appear to be guided by recent reinforcement history when tested in an operant context, thereby demonstrating behavior that parallels that seen in humans, but they appear to be guided by temporal cues when tested in an open-field apparatus, showing behavior more akin to that seen in pigeons. We tested rhesus macaques (Macaca mulatta) on the MSR with a computerized simultaneous visual discrimination to assess whether they would show errors indicative of control by time or by recent reinforcement history. When a single reversal point occurred midsession, rhesus macaques showed no anticipation of the reversal and a similar level of perseveration to rats tested in an operant setting. Nearly identical results also were observed when the monkeys were trained with a single, variable reversal point or with multiple, variable reversal points within a session. These results indicate that temporal cues are not guiding response flexibility in rhesus macaque visual discrimination.