Distances and directions are computed separately by honeybees in landmark-based search

Honeybees were trained with two landmarks at some angle (e.g., 120°) apart from the target. On crucial unrewarded tests, only a single landmark was present. If distances and directions to landmarks are computed separately (independent averaging), the search distance to the landmark should equal the landmark-target distance found in training. If entire vectors are averaged, the search distance should be much shorter. Three experiments with short target-landmark distances showed results in between the predictions of the two hypotheses. A fourth experiment used longer target-landmark distances and isolated double peaks on single-landmark tests: one predicted by the independent averaging hypothesis, and one very close to the landmark. The near peak is interpreted as arising from approach and exploration of a landmark in a new location, and not from searching.     

Compound conditioning in honeybees: Blocking tests of the independence assumption

Six experiments with color-odor compounds failed to produce convincing evidence of blocking in honeybees even when the possibility of masking by within-compound association could be discounted. The parsimonious assumption that the components of a compound stimulus gain and lose associative strength independently with reinforcement and nonreinforcement of the compounds (which the experiments were designed to challenge) remains tenable for color-odor compounds, although perhaps not for intramodal compounds.     

Blocking optimized SIMD tree search on modern processors

Tree search is a widely used fundamental algorithm. Modern processors provide tremendous computing power by integrating multiple cores, each with a vector processing unit. This paper reviews some studies on exploiting single instruction multiple date (SIMD) capacity of processors to improve the performance of tree search, and proposes several improvement methods on reported SIMD tree search algorithms. Based on blocking tree structure, blocking for memory alignment and dynamic blocking prefetch are proposed to optimize the overhead of memory access. Furthermore, as a way of non-linear loop unrolling, the search branch unwinding shows that the number of branches can exceed the data width of SIMD instructions in the SIMD search algorithm. The experiments suggest that blocking optimized SIMD tree search algorithm can achieve 1.6 times response speed faster than the un-optimized algorithm.     

The determination of direction in landmark-based spatial search in pigeons: A further test of the vector sum model

In four experiments, pigeons were trained to find hidden food at a constant location with respect to one or two arrays of landmarks. On crucial tests, the birds were presented with conflicting cues associated with two different directions, which were 90° apart from the center of the search space at the same radial distance. The direction-averaging model predicts that the radial distance of search should not change on these tests, compared with radial distance of search on control tests without conflicting cues. The vector-averaging (vector sum) model predicts that when pigeons average the two conflicting cues, the radial distance of search should be shorter. Results support the direction-averaging model and suggest that distance and direction are independently computed in landmark-based search. Multiple sources are averaged by pigeons in determining direction.     

Latent inhibition in honeybees

Aversive conditioning was studied in individual honeybees flying back and forth between the hive and the sill of an open laboratory window, where they took sucrose solution from a target so constructed that shock could be delivered while the proboscis was in contact with the solution. During feeding, a conditioned stimulus—substrate vibration or airstream—was paired with brief shock avoidable by interruption of feeding. In Experiment 1, unreinforced preexposure of the conditioned stimulus was found to retard acquisition (latent inhibition). In Experiment 2, which was designed to inquire into the stimulus specificity of the effect, differential conditioning was found to be impaired by unreinforced preexposure of the positive stimulus and facilitated by unreinforced preexposure of the negative stimulus. In Experiment 3, a summation experiment designed to test various alternative explanations of the effect, a preexposed stimulus was found to suppress response to an excitatory conditioned stimulus when the two stimuli were presented together.     

Intramodal blocking in honeybees

Foraging honeybees were trained in a concurrent blocking design with a compound stimulus (AX) reinforced and one of its components (A) either reinforced for a blocking group or nonreinforced for a control group. In Experiment 1, a compound of two colors was used; in Experiment 2, a compound of two odors was used; in Experiment 3, a color-position compound, with position defined in terms of proximity to a distinctive visual landmark, was used; and, in Experiment 4, an odor-position compound was used. In each of the first three experiments, the blocking group responded less than did the control group in a subsequent test with X; in the fourth experiment, the two groups did not differ. The results are in accord with expectations based on those of previous experiments with honeybees in which the independence assumption was found to hold for intermodal compounds but not for intramodal compounds.     

Short-term spatial memory in honeybees

Foraging honeybees were trained individually in two-choice spatial problems. Differentially rewarded for spatial alternation in Experiment 1 (“win-shift” training), they showed instead a clear tendency to perseverate—that is, to prefer on each trial the location of reward on the immediately preceding trial. On the basis of the results of Experiments 2 and 3, in which one location was rewarded over shorter or longer series of consecutive trials, an associative interpretation of the perseveration found in the first experiment was rejected in favor of an interpretation in terms of short-term spatial memory. Experiment 4, in which the animals were rewarded on each trial for choosing either location, also showed perseveration. Honeybees, like rats, seem to remember a rewarded location recently visited, but tend to return to it rather than, like rats, to avoid it.     

Analysis of choice in honeybees

Free-flying honeybees were trained in a set of four problems to choose between two differently scented targets, one or the other of which contained sucrose solution. The training was simulated quantitatively, always with the same simple linear equations for computing changes produced by reinforcement and nonreinforcement in the strength of association between each target and the sucrose, but with a diverse array of functions for predicting choice on the basis of relative strength. Accuracy of prediction was indexed by the root-mean-square (RMS) deviation of simulated data from real data. The results provide some good approximations of what is tentatively assumed to be the true choice function, setting the stage for further development of the associative features of the model to encompass more complex phenomena of honeybee learning in choice situations.     

The US-preexposure effect in honeybees

Signaled avoidance was studied in individual honeybees that visited the laboratory regularly to take sucrose solution from a target set on the sill of an open window. During feeding, substrate vibration or airstream was used to signal a brief shock that could be avoided by breaking off contact with the food for a few seconds. Aversive conditioning of the context was measured in terms of return time (the time between successive visits). In Experiment 1, experience with unsignaled shock was found to lengthen return time—which experience with signaled shock did not—and to impair performance in subsequent avoidance training with signaled shock (the US-preexposure effect). In Experiment 2, experience with unsignaled shock given after signaled avoidance training lengthened return time but had no effect on response to the signal in a subsequent extinction test. These results closely resemble the results obtained in analogous experiments with vertebrates.     

Working memory for color in honeybees

Honeybees were tested in delayed conditional discrimination procedures (matching-to-sample and nonmatching-to-sample), using color stimuli presented on a video monitor. A small but reliable tendency to choose the color presented as the conditional cue was found, regardless of whether the contingencies reinforced or discouraged this tendency. The perseverative tendency occurred even with a delay of up to 1–2 min between the conditional cue and the choice. The tendency cannot be explained by changes in the associative value of the colors. Explanation of the results requires some form of working memory for color.     

Comparative analysis of learning in honeybees

The performance of free-flying and harnessed honeybees has been studied in a variety of experiments patterned after those in which learning in vertebrates has been studied—among them experiments on amount, quality, and probability of reward; on compound conditioning and discrimination; and on spatial learning and memory. Despite the remoteness of the evolutionary relationship and the vast differences in brain size and structure, the results for honeybees are strikingly similar to those for vertebrates in many respects and different in only a few. The extent to which phenomena of learning common to honeybees and vertebrates can be understood in terms of common functional principles and mechanisms remains to be determined. None of the differences in the results for honeybees and vertebrates points unmistakably to a difference in their learning.     

Analysis of the overlearning-extinction effect in honeybees

In previous experiments with individual honeybees that visited the laboratory regularly to take sucrose solution from a target set on the shelf of an open window, the overlearning-extinction effect was found for high concentrations of sucrose but not for low. The purpose of the present experiment was to examine the possibility that declining resistance to extinction in the course of prolonged training with a high concentration of sucrose could be explained in terms of increasing nutritive level. Three groups of animals were extinguished on a distinctive target, one group after 6 visits to the target, a second after 18 visits to the target, and a third after 6 visits to the target that were interspersed among 12 visits to a different target. More rapid extinction in the second group than in the third, which had fewer training visits to the extinction target than the second but the same total number of training visits, rules out an explanation in terms of nutritive level and points instead to a frustration-like process evidenced also in earlier work on incentive contrast.     

Control of performance by short-term memory in honeybees

Individual honeybees foraging at a laboratory window were trained with a correction method to choose between blue and yellow targets, one of which contained sucrose solution. There were two trials on each visit, with the locus of the sucrose predictable only on the second. Animals differentially rewarded on Trial 2 for choosing the rewarded color of Trial 1, for choosing the alternative color, or for choosing the target in the rewarded position of Trial 1 independently of its color, all showed a small but persistent preference for the rewarded color, with no significant preference for the rewarded position. When the positions of the colored targets were the same on Trial 2 as on Trial 1 (color and position confounded), there was a more substantial but equally persistent preference on Trial 2 for the rewarded color-position of Trial 1, whether the animals were differentially rewarded for perseveration or for alternation. The results provide further evidence of unlearned control of performance by short-term memory in honeybees but no indication of learned control.     

Partial reinforcement and resistance to extinction in honeybees

Previous experiments have shown the partial reinforcement effect in honeybees under conditions which permit an interpretation in terms of sensory carryover. In the five experiments reported here, the effect was sought under conditions which would require an interpretation in terms of associative reinstatement. Since it is not feasible to train honeybees in widely spaced trials, several different interpolated-trials procedures were employed which had in common the feature that nonrewarded response to a stimulus never was followed by rewarded response to the same stimulus. Implications of the negative results for the interpretation of the overlearning-extinction effect and successive negative contrast in honeybees are considered.     

Blocking between landmarks during 2-D (touchscreen) and 3-D (ARENA) search tasks with pigeons

Many studies investigating cue competition have focused on the blocking effect. We investigated the blocking effect with pigeons using a landmark-based spatial search task in both a touchscreen preparation (Exp. 1a) and an automated remote environmental navigation apparatus (Exp. 1b). In Phase 1, two landmarks (LMs: A and Z) appeared on separate trials as colored circles among a row of eight (touchscreen) or six (ARENA) identical response units. Subjects were rewarded for pecking at a target response unit to the right of LM A and to the left of an extraneous LM, Z. During the blocking trials in Phase 2, LM X was presented in compound with a second LM (A) that had been previously trained. On control trials, LM Y was presented in compound with LM B and a target in the same manner as in the trials of AX, except that neither landmark had previously been trained with the target. All subjects were then tested with separate trials of A, X, B, and Y. Testing revealed poor spatial control by X relative to A and Y. We report the first evidence for a spatial-blocking effect in pigeons and additional support for associative effects (e.g., blocking) occurring under similar conditions (e.g., training sessions, spatial relationships, etc.) in 3-D and 2-D search tasks.     

Performance of honeybees in analogues of the rodent radial maze

The performance of individual honeybees pretrained to forage at a laboratory window was studied in three rudimentary analogues of the radial maze designed for the study of short-term spatial memory in rats. A linear arrangement of three targets was used in Experiment 1, a triangular arrangement of three targets in Experiment 2, and a rectangular arrangement of four targets in Experiment 3, with reward only for the first response to each of the targets presented on any given trial. Several systematic patterns of responding were observed, with no indication that the choices made by the animals were influenced by memory of targets recently visited.     

Blocking despite changes in reinforcer identity

Rats were trained on a discriminated operant barpressing task according to a standard blocking design. In some conditions, the reinforcer was changed between the pretraining and compound conditioning phases; for other conditions, the reinforcer remained the same across phases. In three separate experiments using both between- and within-subject designs, strong blocking effects occurred regardless of the change in the reinforcer. In a fourth experiment, a multiple schedule of reinforcement was used in which response-independent reinforcers were superimposed on the schedule of response-contingent reinforcers. The degree of response suppression caused by the free reinforcer was greater when the free reinforcers were the same as the response-contingent reinforcers than when they were different. The role played by the reinforcer identity in contingency experiments thus appears to be different from the role it plays in blocking experiments.     

Choice in honeybees as a function of the probability of reward

Honeybees foraging for sucrose at a laboratory window were trained in a series of ten 100-trial problems to choose between two targets differing in odor, one of them providing 10 µl of a 50% sucrose solution and the other 10 µl of water. In 9 of the problems, two odors were used, and the reward ratio was varied systematically over a wide range. In the 10th problem, three odors were used in an ambiguous-cue (A+/B?, B+/C?) design. The results were predicted quantitatively, and with substantial accuracy, from a simple theory of learning and choice developed in previous work on simultaneous discrimination in honeybees.     

The existence and extent of spatial working memory ability in honeybees

Honeybees foraged from six locations, each of which was baited with sugar solution prior to each experimental trial. Under a variety of conditions, bees exhibited a small but reliable tendency to avoid revisits to locations that they had visited earlier during the experimental trial. These results replicate those of Brown and Demas (1994), who concluded that bees use working memory to discriminate previously visited locations from those not yet visited. The present experiments included procedures that allowed alternatives to this explanation to be more completely ruled out. The extent of spatial working memory performance exhibited by honeybees in these experiments appears to be limited by a process other than working memory capacity, perhaps the ability of bees to discriminate among several locations in close proximity to one another.     

Reward and learning in honeybees: analysis of an overshadowing effect

Previous experiments have shown that honeybees trained with colored targets baited with 5- versus 20-µl drops of sucrose solution fail to develop a preference for the 20-µl color when the location of the drop on each target is marked by a white dot (dot-color overshadowing) but that discrimination is not impaired by dots when the targets differ in odor rather than in color. In Experiments 1–3, dot-color overshadowing failed to appear with differences in concentration rather than amount of sucrose (50% vs. 20% or 0%), but it did appear in Experiments 4 and 5 with a difference in probability of reward (consistent vs. partial). Experiment 6 showed no dot-odor overshadowing with a difference in probability of reward. The results are not generally predictable from the Rescorla-Wagner principle of shared associative strength, but point instead (in conjunction with those of earlier experiments) to competition for visual attention.