Contrast independent biologically inspired translational optic flow estimation

The visual systems of insects are relatively simple compared to humans. However, they enable navigation through complex environments where insects perform exceptional levels of obstacle avoidance. Biology uses two separable modes of optic flow to achieve this: rapid gaze fixation (rotational motion known as saccades); and the inter-saccadic translational motion. While the fundamental process of insect optic flow has been known since the 1950's, so too has its dependence on contrast. The surrounding visual pathways used to overcome environmental dependencies are less well known. Previous work has shown promise for low-speed rotational motion estimation, but a gap remained in the estimation of translational motion, in particular the estimation of the time to impact. To consistently estimate the time to impact during inter-saccadic translatory motion, the fundamental limitation of contrast dependence must be overcome. By adapting an elaborated rotational velocity estimator from literature to work for translational motion, this paper proposes a novel algorithm for overcoming the contrast dependence of time to impact estimation using nonlinear spatio-temporal feedforward filtering. By applying bioinspired processes, approximately 15 points per decade of statistical discrimination were achieved when estimating the time to impact to a target across 360 background, distance, and velocity combinations: a 17-fold increase over the fundamental process. These results show the contrast dependence of time to impact estimation can be overcome in a biologically plausible manner. This, combined with previous results for low-speed rotational motion estimation, allows for contrast invariant computational models designed on the principles found in the biological visual system, paving the way for future visually guided systems.

Discrimination of edge orientation by bumblebees

Simple feature detectors in the visual system, such as edge-detectors, are likely to underlie even the most complex visual processing, so understanding the limits of these systems is crucial for a fuller understanding of visual processing. We investigated the ability of bumblebees (Bombus terrestris) to discriminate between differently angled edges. In a multiple-choice, “meadow-like” scenario, bumblebees successfully discriminated between angled bars with 7° differences, significantly exceeding the previously reported performance of eastern honeybees (Apis cerana, limit: 15°). Neither the rate at which bees learned, nor their final discrimination performance were affected by the angular orientation of the training bars, indicating a uniform performance across the visual field. Previous work has found that, in dual-choice tests, eastern honeybees cannot reliably discriminate between angles with less than 25° difference, suggesting that performance in discrimination tasks is affected by the training regime, and doesn’t simply reflect the perceptual limitations of the visual system. We used high resolution LCD monitors to investigate bumblebees’ angular resolution in a dual-choice experiment. Bumblebees could still discriminate 7° angle differences under such conditions (exceeding the previously reported limit for Apis mellifera, of 10°, as well as that of A. cerana). Bees eventually reached similar levels of accuracy in the dual-choice experiment as they did under multiple-choice conditions but required longer learning periods. Bumblebees show impressive abilities to discriminate between angled edges, performing better than two previously tested species of honeybee. This high performance may, in turn, support complex visual processing in the bumblebee brain.

Naïve and Experienced Honeybee Foragers Learn Normally Configured Flowers More Easily Than Non-configured or Highly Contrasted Flowers

Angiosperms have evolved to attract and/or deter specific pollinators. Flowers provide signals and cues such as scent, colour, size, pattern, and shape, which allow certain pollinators to more easily find and visit the same type of flower. Over evolutionary time, bees and angiosperms have co-evolved resulting in flowers being more attractive to bee vision and preferences, and allowing bees to recognise specific flower traits to make decisions on where to forage. Here we tested whether bees are instinctively tuned to process flower shape by training both flower-experienced and flower-naïve honeybee foragers to discriminate between pictures of two different flower species when images were either normally configured flowers or flowers which were scrambled in terms of spatial configuration. We also tested whether increasing picture contrast, to make flower features more salient, would improve or impair performance. We used four flower conditions: (i) normally configured greyscale flower pictures, (ii) scrambled flower configurations, (iii) high contrast normally configured flowers, and (iv) asymmetrically scrambled flowers. While all flower pictures contained very similar spatial information, both experienced and naïve bees were better able to learn to discriminate between normally configured flowers than between any of the modified versions. Our results suggest that a specialisation in flower recognition in bees is due to a combination of hard-wired neural circuitry and experience-dependent factors.

Wild non-eusocial bees learn a colour discrimination task in response to simulated predation events

Despite representing the majority of bee species, non-eusocial bees (e.g. solitary, subsocial, semisocial, and quasisocial species) are comparatively understudied in learning, memory, and cognitive-like behaviour compared to eusocial bees, such as honeybees and bumblebees. Ecologically relevant colour discrimination tasks are well-studied in eusocial bees, and research has shown that a few non-eusocial bee species are also capable of colour learning and long-term memory retention. Australia hosts over 2000 native bee species, most of which are non-eusocial, yet evidence of cognitive-like behaviour and learning abilities under controlled testing conditions is lacking. In the current study, I examine the learning ability of a non-eusocial Australian bee, Lasioglossum (Chilalictus) lanarium, using aversive differential conditioning during a colour discrimination task. L. lanarium learnt to discriminate between salient blue- and yellow-coloured stimuli following training with simulated predation events. This study acts as a bridge between cognitive studies on eusocial and non-social bees and introduces a framework for testing non-eusocial wild bees on elemental visual learning tasks using aversive conditioning. Non-eusocial bee species are far more numerous than eusocial species and contribute to agriculture, economics, and ecosystem services in Australia and across the globe. Thus, it is important to study their capacity to learn flower traits allowing for successful foraging and pollination events, thereby permitting us a better understanding of their role in plant-pollinator interactions.

Spatial memory in Vespula germanica wasps: A pilot study using a Y-maze assay

In the present study we analysed spatial learning in Vespula germanica wasps when dealing with a walking Y-maze. We recorded the time taken to leave the maze during two consecutive visits and which of the two short arms was chosen to exit. Two treatments were conducted to evaluate whether wasps learned to leave the Y-maze guided either by spatial or visual cues. In Treatment 1, the colour of both arms remained unchanged between two consecutive visits; and in Treatment 2, the position of the coloured arm was switched after the first trial. Our results demonstrated that the time taken to exit the maze on the second trial was less than half in both treatments and wasps left the maze from the previously chosen arm, irrespective of its colour. This is the first study to demonstrate spatial learning in V. germanica wasps by using a walking Y-maze. Free flying wasps learned to enter the Y-maze on their own volition, walk through it, collect food and find their way out more rapidly after a single foraging experience. The current experimental device is suitable for the evaluation of spatial memory processes and exploratory behaviour in this species.

Different mechanisms underlie implicit visual statistical learning in honey bees and humans

The ability of developing complex internal representations of the environment is considered a crucial antecedent to the emergence of humans' higher cognitive functions. Yet it is an open question whether there is any fundamental difference in how humans and other good visual learner species naturally encode aspects of novel visual scenes. Using the same modified visual statistical learning paradigm and multielement stimuli, we investigated how human adults and honey bees (Apis mellifera) encode spontaneously, without dedicated training, various statistical properties of novel visual scenes. We found that, similarly to humans, honey bees automatically develop a complex internal representation of their visual environment that evolves with accumulation of new evidence even without a targeted reinforcement. In particular, with more experience, they shift from being sensitive to statistics of only elemental features of the scenes to relying on co-occurrence frequencies of elements while losing their sensitivity to elemental frequencies, but they never encode automatically the predictivity of elements. In contrast, humans involuntarily develop an internal representation that includes single-element and co-occurrence statistics, as well as information about the predictivity between elements. Importantly, capturing human visual learning results requires a probabilistic chunk-learning model, whereas a simple fragment-based memory-trace model that counts occurrence summary statistics is sufficient to replicate honey bees' learning behavior. Thus, humans' sophisticated encoding of sensory stimuli that provides intrinsic sensitivity to predictive information might be one of the fundamental prerequisites of developing higher cognitive abilities.

Surpassing the subitizing threshold: appetitive–aversive conditioning improves discrimination of numerosities in honeybees

Animals including humans, fish and honeybees have demonstrated a quantity discrimination threshold at four objects, often known as subitizing elements. Discrimination between numerosities at or above the subitizing range is considered a complex capacity. In the current study, we trained and tested two groups of bees on their ability to differentiate between quantities (4 versus 5 through to 4 versus 8) when trained with different conditioning procedures. Bees trained with appetitive (reward) differential conditioning demonstrated no significant learning of this task, and limited discrimination above the subitizing range. In contrast, bees trained using appetitive–aversive (reward–aversion) differential conditioning demonstrated significant learning and subsequent discrimination of all tested comparisons from 4 versus 5 to 4 versus 8. Our results show conditioning procedure is vital to performance on numerically challenging tasks, and may inform future research on numerical abilities in other animals.

Morphological Complexity as a Floral Signal: From Perception by Insect Pollinators to Co-Evolutionary Implications

Morphologically complex flowers are characterized by bilateral symmetry, tube-like shapes, deep corolla tubes, fused petals, and/or poricidal anthers, all of which constrain the access of insect visitors to floral nectar and pollen rewards. Only a subset of potential pollinators, mainly large bees, learn to successfully forage on such flowers. Thus, complexity may comprise a morphological filter that restricts the range of visitors and thereby increases food intake for successful foragers. Such pollinator specialization, in turn, promotes flower constancy and reduces cross-species pollen transfer, providing fitness benefits to plants with complex flowers. Since visual signals associated with floral morphological complexity are generally honest (i.e., indicate food rewards), pollinators need to perceive and process them. Physiological studies show that bees detect distant flowers through long-wavelength sensitive photoreceptors. Bees effectively perceive complex shapes and learn the positions of contours based on their spatial frequencies. Complex flowers require long handling times by naive visitors, and become highly profitable only for experienced foragers. To explore possible pathways towards the evolution of floral complexity, we discuss cognitive mechanisms that potentially allow insects to persist on complex flowers despite low initial foraging gains, suggest experiments to test these mechanisms, and speculate on their adaptive value.

Sequential learning of relative size by the Neotropical ant Gigantiops destructor

The question of whether insects can perform concept learning or can use the geometry of space as in mammals has been recently addressed in Hymenoptera in an extensive way. We investigate here the ability of the tropical ant Gigantiops destructor to perform sequential learning and to use size relationships during navigation. Ants were trained to solve a dichotomic six-stage linear maze relying on the apparent width of two vertical landmarks. Each individual ant first learnt to associate a given landmark width to the motor decision of turning right or left to avoid dead-ends independently of a motor routine. When confronted for the first time with a new intermediate-sized pattern, for which no supposed snapshot could have been stored, ants made directional choices indicating that bar width judgments were not absolute but rather relative to the familiar visual patterns seen in the previous chambers. This result demonstrates that ants can generalize relationship rules by interpolating the relative width of a novel stimulus according to visual information kept in spatial working memory. In conclusion, ants can perform conditional discriminations reliably not only when stimuli are simultaneous but also when they are sequential.

Functional Significance of Labellum Pattern Variation in a Sexually Deceptive Orchid (Ophrys heldreichii): Evidence of Individual Signature Learning Effects

Mimicking female insects to attract male pollinators is an important strategy in sexually deceptive orchids of the genus Ophrys, and some species possess flowers with conspicuous labellum patterns. The function of the variation of the patterns remains unresolved, with suggestions that these enhance pollinator communication. We investigated the possible function of the labellum pattern in Ophrys heldreichii, an orchid species in which the conspicuous and complex labellum pattern contrasts with a dark background. The orchid is pollinated exclusively by males of the solitary bee, Eucera berlandi. Comparisons of labellum patterns revealed that patterns within inflorescences are more similar than those of other conspecific plants. Field observations showed that the males approach at a great speed and directly land on flowers, but after an unsuccessful copulation attempt, bees hover close and visually scan the labellum pattern for up to a minute. Learning experiments conducted with honeybees as an accessible model of bee vision demonstrated that labellum patterns of different plants can be reliably learnt; in contrast, patterns of flowers from the same inflorescence could not be discriminated. These results support the hypothesis that variable labellum patterns in O. heldreichii are involved in flower-pollinator communication which would likely help these plants to avoid geitonogamy.

The forest or the trees: preference for global over local image processing is reversed by prior experience in honeybees

Traditional models of insect vision have assumed that insects are only capable of low-level analysis of local cues and are incapable of global, holistic perception. However, recent studies on honeybee (Apis mellifera) vision have refuted this view by showing that this insect also processes complex visual information by using spatial configurations or relational rules. In the light of these findings, we asked whether bees prioritize global configurations or local cues by setting these two levels of image analysis in competition. We trained individual free-flying honeybees to discriminate hierarchical visual stimuli within a Y-maze and tested bees with novel stimuli in which local and/or global cues were manipulated. We demonstrate that even when local information is accessible, bees prefer global information, thus relying mainly on the object's spatial configuration rather than on elemental, local information. This preference can be reversed if bees are pre-trained to discriminate isolated local cues. In this case, bees prefer the hierarchical stimuli with the local elements previously primed even if they build an incorrect global configuration. Pre-training with local cues induces a generic attentional bias towards any local elements as local information is prioritized in the test, even if the local cues used in the test are different from the pre-trained ones. Our results thus underline the plasticity of visual processing in insects and provide new insights for the comparative analysis of visual recognition in humans and animals.

Insect vision models under scrutiny: what bumblebees (Bombus terrestris terrestris L.) can still tell us

Three contending models address the ability of bees to detect and discriminate colours: the colour opponent coding (COC) model, the colour hexagon (CH) model and the receptor noise-limited (RN) model, but few studies attempt to determine which model fits experimental data best. To assess whether the models provide an accurate description of bumblebee colour space, we trained bees to discriminate four colour pairs. The perceptual distance between the colours of each pair was similar according to the CH model but varied widely according to the COC and RN models. The time that bees required to select a flower and the proportion of correct choices differed between groups: decision times decreased as achromatic contrast increased, and the proportion of correct choices increased with achromatic contrast and perceptual distance, as predicted by the COC and RN models. These results suggest that both chromatic and achromatic contrasts affected the discriminability of colour pairs. Since flower colour affects the foraging choices of bees and foraging choices affect the reproductive success of plants, a better understanding of which model is more accurate under each circumstance is required to predict bee behaviour and the ecological implications of flower choice and colour.

Conceptualization of relative size by honeybees

The ability to process visual information using relational rules allows for decisions independent of the specific physical attributes of individual stimuli. Until recently, the manipulation of relational concepts was considered as a prerogative of large mammalian brains. Here we show that individual free flying honeybees can learn to use size relationship rules to choose either the larger or smaller stimulus as the correct solution in a given context, and subsequently apply the learnt rule to novel colors and shapes providing that there is sufficient input to the long wavelength (green) photoreceptor channel. Our results add a novel, size-based conceptual rule to the set of relational concepts that honeybees have been shown to master and underline the value of bees as an animal model for studying the emergence of conceptualization abilities.

Blue colour preference in honeybees distracts visual attention for learning closed shapes

Spatial vision is an important cue for how honeybees (Apis mellifera) find flowers, and previous work has suggested that spatial learning in free-flying bees is exclusively mediated by achromatic input to the green photoreceptor channel. However, some data suggested that bees may be able to use alternative channels for shape processing, and recent work shows conditioning type and training length can significantly influence bee learning and cue use. We thus tested the honeybees' ability to discriminate between two closed shapes considering either absolute or differential conditioning, and using eight stimuli differing in their spectral characteristics. Consistent with previous work, green contrast enabled reliable shape learning for both types of conditioning, but surprisingly, we found that bees trained with appetitive-aversive differential conditioning could additionally use colour and/or UV contrast to enable shape discrimination. Interestingly, we found that a high blue contrast initially interferes with bee shape learning, probably due to the bees innate preference for blue colours, but with increasing experience bees can learn a variety of spectral and/or colour cues to facilitate spatial learning. Thus, the relationship between bee pollinators and the spatial and spectral cues that they use to find rewarding flowers appears to be a more rich visual environment than previously thought.

The effect of flower-like and non-flower-like visual properties on choice of unrewarding patterns by bumblebees

How do distinct visual stimuli help bumblebees discover flowers before they have experienced any reward outside of their nest? Two visual floral properties, type of a pattern (concentric vs radial) and its position on unrewarding artificial flowers (central vs peripheral on corolla), were manipulated in two experiments. Both visual properties showed significant effects on floral choice. When pitted against each other, pattern was more important than position. Experiment 1 shows a significant effect of concentric pattern position, and experiment 2 shows a significant preference towards radial patterns regardless of their position. These results show that the presence of markings at the center of a flower are not so important as the presence of markings that will direct bees there.

Numerical cognition in bees and other insects

The ability to perceive the number of objects has been known to exist in vertebrates for a few decades, but recent behavioral investigations have demonstrated that several invertebrate species can also be placed on the continuum of numerical abilities shared with birds, mammals, and reptiles. In this review article, we present the main experimental studies that have examined the ability of insects to use numerical information. These studies have made use of a wide range of methodologies, and for this reason it is striking that a common finding is the inability of the tested animals to discriminate numerical quantities greater than four. Furthermore, the finding that bees can not only transfer learnt numerical discrimination to novel objects, but also to novel numerosities, is strongly suggestive of a true, albeit limited, ability to count. Later in the review, we evaluate the available evidence to narrow down the possible mechanisms that the animals might be using to solve the number-based experimental tasks presented to them. We conclude by suggesting avenues of further research that take into account variables such as the animals' age and experience, as well as complementary cognitive systems such as attention and the time sense.

Honeybees can discriminate between Monet and Picasso paintings

Honeybees (Apis mellifera) have remarkable visual learning and discrimination abilities that extend beyond learning simple colours, shapes or patterns. They can discriminate landscape scenes, types of flowers, and even human faces. This suggests that in spite of their small brain, honeybees have a highly developed capacity for processing complex visual information, comparable in many respects to vertebrates. Here, we investigated whether this capacity extends to complex images that humans distinguish on the basis of artistic style: Impressionist paintings by Monet and Cubist paintings by Picasso. We show that honeybees learned to simultaneously discriminate between five different Monet and Picasso paintings, and that they do not rely on luminance, colour, or spatial frequency information for discrimination. When presented with novel paintings of the same style, the bees even demonstrated some ability to generalize. This suggests that honeybees are able to discriminate Monet paintings from Picasso ones by extracting and learning the characteristic visual information inherent in each painting style. Our study further suggests that discrimination of artistic styles is not a higher cognitive function that is unique to humans, but simply due to the capacity of animals-from insects to humans-to extract and categorize the visual characteristics of complex images.

The mysterious cognitive abilities of bees: why models of visual processing need to consider experience and individual differences in animal performance

Vision is one of the most important modalities for the remote perception of biologically important stimuli. Insects like honeybees and bumblebees use their colour and spatial vision to solve tasks, such as navigation, or to recognise rewarding flowers during foraging. Bee vision is one of the most intensively studied animal visual systems, and several models have been developed to describe its function. These models have largely assumed that bee vision is determined by mechanistic hard-wired circuits, with little or no consideration for behavioural plasticity or cognitive factors. However, recent work on both bee colour vision and spatial vision suggests that cognitive factors are indeed a very significant factor in determining what a bee sees. Individual bumblebees trade-off speed for accuracy, and will decide on which criteria to prioritise depending upon contextual information. With continued visual experience, honeybees can learn to use non-elemental processing, including configural mechanisms and rule learning, and can access top-down information to enhance learning of sophisticated, novel visual tasks. Honeybees can learn delayed-matching-to-sample tasks and the rules governing this decision making, and even transfer learned rules between different sensory modalities. Finally, bees can learn complex categorisation tasks and display numerical processing abilities for numbers up to and including four. Taken together, this evidence suggests that bees do have a capacity for sophisticated visual behaviours that fit a definition for cognition, and thus simple elemental models of bee vision need to take account of how a variety of factors may influence the type of results one may gain from animal behaviour experiments.