Evidence of predictive selective attention in fiddler crabs during escape in the natural environment

Contrast sensitivity, visual acuity and the effect of behavioural state on optokinetic gain in fiddler crabs

Most animals rely on visual information for a variety of everyday tasks. The information available to a visual system depends in part on its spatial resolving power and contrast sensitivity. Because of their competing demands for physical space within an eye, these traits cannot simultaneously be improved without increasing overall eye size. The contrast sensitivity function is an integrated measure of visual performance that measures both resolution and contrast sensitivity. Its measurement helps us identify how different species have made a trade-off between contrast sensitivity and spatial resolution. It further allows us to identify the evolutionary drivers of sensory processing and visually mediated behaviour. Here, we measured the contrast sensitivity function of the fiddler crab Gelasimus dampieri using its optokinetic responses to wide-field moving sinusoidal intensity gratings of different orientations, spatial frequencies, contrasts and speeds. We further tested whether the behavioural state of the crabs (i.e. whether crabs are actively walking or not) affects their optokinetic gain and contrast sensitivity. Our results from a group of five crabs suggest a minimum perceived contrast of 6% and a horizontal and vertical visual acuity of 0.4 cyc deg-1 and 0.28 cyc deg-1, respectively, in the crabs' region of maximum optomotor sensitivity. Optokinetic gain increased in moving crabs compared with restrained crabs, adding another example of the importance of naturalistic approaches when studying the performance of animals.

Attention and Prediction-Guided Motion Detection for Low-Contrast Small Moving Targets

Small target motion detection within complex natural environments is an extremely challenging task for autonomous robots. Surprisingly, the visual systems of insects have evolved to be highly efficient in detecting mates and tracking prey, even though targets occupy as small as a few degrees of their visual fields. The excellent sensitivity to small target motion relies on a class of specialized neurons, called small target motion detectors (STMDs). However, existing STMD-based models are heavily dependent on visual contrast and perform poorly in complex natural environments, where small targets generally exhibit extremely low contrast against neighboring backgrounds. In this article, we develop an attention-and-prediction-guided visual system to overcome this limitation. The developed visual system comprises three main subsystems, namely: 1) an attention module; 2) an STMD-based neural network; and 3) a prediction module. The attention module searches for potential small targets in the predicted areas of the input image and enhances their contrast against a complex background. The STMD-based neural network receives the contrast-enhanced image and discriminates small moving targets from background false positives. The prediction module foresees future positions of the detected targets and generates a prediction map for the attention module. The three subsystems are connected in a recurrent architecture, allowing information to be processed sequentially to activate specific areas for small target detection. Extensive experiments on synthetic and real-world datasets demonstrate the effectiveness and superiority of the proposed visual system for detecting small, low-contrast moving targets against complex natural environments.

Fiddler crabs are unique in timing their escape responses based on speed-dependent visual cues

Predation risk imposes strong selection pressures on visual systems to quickly and accurately identify the position and movement of potential predators.^1,2 Many invertebrates and other small animals, however, have limited capacity for distance perception due to their low spatial resolution and closely situated eyes.^3,4 Consequently, they often rely on simplified decision criteria, essentially heuristics or "rules of thumb", to make decisions. The visual cues animals use to make escape decisions are surprisingly consistent, especially among arthropods, with the timing of escape commonly triggered by size-dependent visual cues such as angular size or angular size increment.^5,6,7,8,9,10 Angular size, however, confuses predator size and distance and provides no information about the speed of the attack. Here, we show that fiddler crabs (Gelasimus dampieri) are unique among the arthropods studied to date as they timed their escape response based on the speed of an object's angular expansion. The crabs responded reliably by running away from visual stimuli that expanded at approximately 1.7 degrees/s, irrespective of stimulus size, speed, or its initial distance from the crabs. Though the threshold expansion speed was consistent across different stimulus conditions, we found that the escape timing was modulated by the elevation at which the stimulus approached, suggesting that other risk factors can bias the expansion speed threshold. The results suggest that the visual escape cues used by arthropods are less conserved than previously thought and that lifestyle and environment are significant drivers determining the escape cues used by different species.

Behavioural and neural responses of crabs show evidence for selective attention in predator avoidance

Selective attention, the ability to focus on a specific stimulus and suppress distractions, plays a fundamental role for animals in many contexts, such as mating, feeding, and predation. Within natural environments, animals are often confronted with multiple stimuli of potential importance. Such a situation significantly complicates the decision-making process and imposes conflicting information on neural systems. In the context of predation, selectively attending to one of multiple threats is one possible solution. However, how animals make such escape decisions is rarely studied. A previous field study on the fiddler crab, Gelasimus dampieri, provided evidence of selective attention in the context of escape decisions. To identify the underlying mechanisms that guide their escape decisions, we measured the crabs' behavioural and neural responses to either a single, or two simultaneously approaching looming stimuli. The two stimuli were either identical or differed in contrast to represent different levels of threat certainty. Although our behavioural data provides some evidence that crabs perceive signals from both stimuli, we show that both the crabs and their looming-sensitive neurons almost exclusively respond to only one of two simultaneous threats. The crabs' body orientation played an important role in their decision about which stimulus to run away from. When faced with two stimuli of differing contrasts, both neurons and crabs were much more likely to respond to the stimulus with the higher contrast. Our data provides evidence that the crabs' looming-sensitive neurons play an important part in the mechanism that drives their selective attention in the context of predation. Our results support previous suggestions that the crabs' escape direction is calculated downstream of their looming-sensitive neurons by means of a population vector of the looming sensitive neuronal ensemble.

A new, fluorescence-based method for visualizing the pseudopupil and assessing optical acuity in the dark compound eyes of honeybees and other insects

Recent interest in applying novel imaging techniques to infer optical resolution in compound eyes underscores the difficulty of obtaining direct measures of acuity. A widely used technique exploits the principal pseudopupil, a dark spot on the eye surface representing the ommatidial gaze direction and the number of detector units (ommatidia) viewing that gaze direction. However, dark-pigmented eyes, like those of honeybees, lack a visible pseudopupil. Attempts over almost a century to estimate optical acuity in this species are still debated. Here, we developed a method to visualize a stable, reliable pseudopupil by staining the photoreceptors with fluorescent dyes. We validated this method in several species and found it to outperform the dark pseudopupil for this purpose, even in pale eyes, allowing more precise location of the gaze centre. We then applied this method to estimate the sampling resolution in the frontal part of the eye of the honeybee forager. We found a broad frontal acute zone with interommatidial angles below 2° and a minimum interommatidial angle of 1.3°, a broader, sharper frontal acute zone than previously reported. Our study provides a new method to directly measure the sampling resolution in most compound eyes of living animals.