Chasing a dummy target: smooth pursuit and velocity control in male blowflies

Male blowflies chase and catch other flies in fast acrobatic flights. To unravel the underlying control system, we presented a black moving sphere instead of a real fly as a pursuit target. By varying the size and speed of the target, we were able to systematically analyse the decisive visual determinants that guide chasing behaviour. Flies pursue targets of a wide range of sizes and velocities. The percentage of pursuits resulting in target capture decreases with increasing target size and speed. Chasing male flies adjust their forward velocity depending on the retinal size of the target, indicating that retinal size is a relevant input variable of the control system. The chasing fly focuses the target with great accuracy in the frontal part of its visual field by means of a smooth pursuit control system using the retinal position of the target to determine the flight direction. We conclude that for a comprehensive understanding of chasing control different time lags in the control systems of angular and forward velocity together with the impact of inertia on fly movements need to be taken into account.     

Characterisation of a blowfly male-specific neuron using behaviourally generated visual stimuli

The pursuit system controlling chasing behaviour in male blowflies has to cope with extremely fast and dynamically changing visual input. An identified male-specific visual neuron called Male Lobula Giant 1 (MLG1) is presumably one major element of this pursuit system. Previous behavioural and modelling analyses have indicated that angular target size, retinal target position and target velocity are relevant input variables of the pursuit system. To investigate whether MLG1 specifically represents any of these visual parameters we obtained in vivo intracellular recordings while replaying optical stimuli that simulate the visual signals received by a male fly during chasing manoeuvres. On the basis of these naturalistic stimuli we find that MLG1 shows distinct direction sensitivity and is depolarised if the target motion contains an upward component. The responses of MLG1 are jointly determined by the retinal position, the speed and direction, and the duration of target motimotion. Coherence analysis reveals that although retinal target size and position are in some way inherent in the responses of MLG1, we find no confirmation of the hypothesis that MLG1 encodes any of these parameters exclusively.     

雄蝇追逐飞行的加速度分析

本文报导了采用高速摄影技术,通过分析雄蝇追逐飞行的加速度对视觉制导问题所进行的研究.我们的结果如下:1.目标蝇的角位置与追逐蝇相应的角加速度分量之间,在追逐蝇的全视场中呈现非线性的关系.追逐蝇的水平角加速度分量与水平误差角在前视场中有较好的线性关系.追逐蝇的俯仰角加速度分量与俯仰误差角之间,当目标蝇位于前上半视场中时,呈现较好的线性关系.2.目标蝇相对追逐蝇的角运动对追逐蝇的相应角加速度分量也有影响,这种影响与目标蝇位置的关系不大.3.对于目标蝇在前后视场中的两种情况,两蝇间的距离对追逐蝇加速度大小影响的规律是不同的:当目标蝇在前视场中时,只经过较短的延迟时间两蝇间的距离与追逐蝇加速度大小出现了正的相关峰,而后视场中的情况不然,它需要较长的延迟时间.两蝇间距离的变化对追逐蝇加速度大小的影响也有类似的现象.4.在追逐过程中雄蝇利用背前区的小眼来追逐带头的目标雌蝇,而组织学研究在雄蝇背前区的小眼中发现了性特化的中心小网膜细胞,与行为研究的结果相呼应.5.文中最后对蝇视觉神经系统中如何获取目标绳的位置和运动参数的问题进行了讨论.     

Tracking and chasing in houseflies (Musca)

The flight trajectories of free flying female and male houseflies have been analyzed in 3 dimensions. Both female and male flies track other flies. The turning velocity α (around the vertical axis) is linearly dependent upon the horizontal angle ψ_F (that is the angle between the trajectory of the tracking fly and the target) for small values of ψ_F in females and for the whole range of ψ_F in males. The 3-dimensional velocity υ _xyz of the chasing fly is linearly dependent upon the distance between leading and chasing fly in males but not in females. Male chasing thus appears to be more efficient than female tracking. It is shown that earlier assumptions on visual control of flight in female flies derived from experiments on fixed flying flies are justified.     

Phonotactic behaviour and vertical sound source localisation of the parasitoid fly Emblemasoma auditrix (Diptera: Sarcophagidae)

1. Acoustically guided movement in a three‐dimensional space is a complex behavioural task performed notably by birds, bats, and some insect species. The precision of acoustic orientation depends on the directionality of the hearing system as well as on auditory behaviour. 2. The fly Emblemasoma auditrix Diptera (Sarcophagidae) is a parasitoid of the cicada Okanagana rimosa Auchenorrhyncha (Cicadidae) and locates its host in the complex habitat of a forest. The phonotactic behaviour of the fly was analysed experimentally with emphasis on the vertical domain in the field. Different experimental setups allowed discriminating subsequent steps in the phonotactic behaviour of E. auditrix. 3. During the phonotactic flight, flies first landed on landmarks, which were used to re‐adjust to the elevation of the sound source. Acoustic targets were located from these resting positions. The sound source elevation was detected at the start of the flight as the longitudinal body axis was adjusted to the inclination of the target sound source. 4. Flies usually did not land directly upon the sound source, but landed nearby, and most often above the target. Within the target area, types of movement for the final approach differed in respect to target position; flies walked predominantly if the final target was located above or below, but for horizontally located targets much of the distance was covered by flight. 5. In conclusion, E. auditrix can locate the acoustic target in complex habitats and uses a flexible multi‐step approach for short‐range phonotaxis.     

雄蝇追逐行为的分析

本文报告了在自由飞行条件下雄蝇追逐的行为实验及其分析的初步结果.其结果如下:1.追逐雄蝇水平方向偏转的角速度dF_1线性地依赖于目标蝇水平方位误差角T_1的大小.当目标在前视场中,即空间误差角|G|<π/4时,线性回归直线的斜率约为37**;而当空间误差角|G|>π/4时,线性回归直线的斜率约为6.7.2.追逐雄蝇俯仰方向偏转角速度dF_2在(-(π/2),π/2)的范围内线性依赖于俯仰误差角T_2的大小,其回归直线的斜率约为14.3.雄蝇追逐行为中,水平方位误差角频数分布的直方图呈现为峰值在零点的对称型分布;而俯仰误差角T_2频数分布的直方图是非对称型的,即仰角出现的频数大大超过俯角出现的频数.4.雄蝇主要利用了两蝇间距离变化dD的信息以及目标误差角来控制向前飞行的速度V.当误差角小时(即目标在前视场中),dD一般为负值,说明两蝇间的距离减小,而雄蝇追逐飞行的加速度A却与dD呈现正的线性关系.当误差角大时(即目标位于后视场中),dD一般为正值,说明两蝇间的距离增加.     

Virtual-reality techniques resolve the visual cues used by fruit flies to evaluate object distances

Insects can estimate distance or time-to-contact of surrounding objects from locomotion-induced changes in their retinal position and/or size. Freely walking fruit flies (Drosophila melanogaster) use the received mixture of different distance cues to select the nearest objects for subsequent visits. Conventional methods of behavioral analysis fail to elucidate the underlying data extraction. Here we demonstrate first comprehensive solutions of this problem by substituting virtual for real objects; a tracker-controlled 360 degrees panorama converts a fruit fly's changing coordinates into object illusions that require the perception of specific cues to appear at preselected distances up to infinity. An application reveals the following: (1) en-route sampling of retinal-image changes accounts for distance discrimination within a surprising range of at least 8-80 body lengths (20-200 mm). Stereopsis and peering are not involved. (2) Distance from image translation in the expected direction (motion parallax) outweighs distance from image expansion, which accounts for impact-avoiding flight reactions to looming objects. (3) The ability to discriminate distances is robust to artificially delayed updating of image translation. Fruit flies appear to interrelate self-motion and its visual feedback within a surprisingly long time window of about 2 s. The comparative distance inspection practiced in the small fruit fly deserves utilization in self-moving robots.     

Haltere-mediated equilibrium reflexes of the fruit fly, Drosophila melanogaster.

Flies display a sophisticated suite of aerial behaviours that require rapid sensory-motor processing. Like all insects, flight control in flies is mediated in part by motion-sensitive visual interneurons that project to steering motor circuitry within the thorax. Flies, however, possess a unique flight control equilibrium sense that is encoded by mechanoreceptors at the base of the halteres, small dumb-bell-shaped organs derived through evolutionary transformation of the hind wings. To study the input of the haltere system onto the flight control system, I constructed a mechanically oscillating flight arena consisting of a cylindrical array of light-emitting diodes that generated the moving image of a 30 degrees vertical stripe. The arena provided closed-loop visual feedback to elicit fixation behaviour, an orientation response in which flies maintain the position of the stripe in the front portion of their visual field by actively adjusting their wing kinematics. While flies orientate towards the stripe, the entire arena was swung back and forth while an optoelectronic device recorded the compensatory changes in wing stroke amplitude and frequency. In order to reduce the background changes in stroke kinematics resulting from the animal's closed-loop visual fixation behaviour, the responses to eight identical mechanical rotations were averaged in each trial. The results indicate that flies possess a robust equilibrium reflex in which angular rotations of the body elicit compensatory changes in both the amplitude and stroke frequency of the wings. The results of uni- and bilateral ablation experiments demonstrate that the halteres are required for these stability reflexes. The results also confirm that halteres encode angular velocity of the body by detecting the Coriolis forces that result from the linear motion of the haltere within the rotating frame of reference of the fly's thorax. By rotating the flight arena at different orientations, it was possible to construct a complete directional tuning map of the haltere-mediated reflexes. The directional tuning of the reflex is quite linear such that the kinematic responses vary as simple trigonometric functions of stimulus orientation. The reflexes function primarily to stabilize pitch and yaw within the horizontal plane.     

Vision in flying insects

Vision guides flight behaviour in numerous insects. Despite their small brain, insects easily outperform current man-made autonomous vehicles in many respects. Examples are the virtuosic chasing manoeuvres male flies perform as part of their mating behaviour and the ability of bees to assess, on the basis of visual motion cues, the distance travelled in a novel environment. Analyses at both the behavioural and neuronal levels are beginning to unveil reasons for such extraordinary capabilities of insects. One recipe for their success is the adaptation of visual information processing to the specific requirements of the behavioural tasks and to the specific spatiotemporal properties of the natural input.     

Detection of object motion by a fly neuron during simulated flight

Object detection on the basis of relative motion was investigated in the fly at the neuronal level. A representative of the figure detection cells (FD-cells), the FD1b-cell, was characterized with respect to its responses to optic flow which simulated the presence of an object during translatory flight. The figure detection cells reside in the fly's third visual neuropil and are believed to play a central role in mediating object-directed turning behaviour. The dynamical response properties as well as the mean response amplitudes of the FD1b-cell depend on the temporal frequency of object motion and on the presence or absence of background motion. The responses of the FD1b-cell to object motion during simulated translatory flight were compared to behavioural responses of the fly as obtained with identical stimuli in a previous study. The behavioural responses could only partly be explained on the basis of the FD1b-cell's responses. Further processing between the third visual neuropil and the final motor output has to be assumed which involves (1) facilitation of the object-induced responses during translatory background motion at moderate temporal frequencies, and (2) inhibition of the object-induced turning responses during translatory background motion at high temporal frequencies. Accepted: 9 October 1999     

The neuro-ecology of resource localization in Drosophila: Behavioral components of perception and search

From the moment an adult fruit fly ecloses, its primary objective in life is to disperse and locate the source of an attractive food odor upon which to feed and reproduce. The evolution of flight has greatly enhanced the success of fruit flies specifically and insects more generally.1 Control of flight by Drosophila melanogaster is unequivocally visual. Strong optomotor reflexes towards translatory and rotational visual flow stabilize forward flight trajectory, altitude, and speed. 2, 3 The steering responses to translatory and rotational flow in particular are mediated by computationally separate neural circuits in the fly’s visual system,4 and gaze-stabilizing body saccades are elicited by threshold integration of expanding visual flow .5 However, visual information is not alone sufficient to enable a fruit fly to recognize and locate an appropriately smelly object due in part to the relatively poor resolution of its compound eyes. Rather, the animal uses an acute sense of smell to actively track odors during flight. Without a finely adapted olfactory system, the fly’s remarkable visual capabilities are for naught. The relative importance of vision is apparent in the cross-modal fusion of the two modalities for stable active odor tracking.6, 7 Olfactory processing in Drosophila is shaped by ecological and functional forces which are inextricably linked. Thus physiologists seeking the functional determinants of olfactory coding as well as ecologists seeking to understand the mechanisms of speciation do well to consider each others’ point of view. Here we synthesize a broad perspective that integrates across ultimate and proximate mechanisms of odor tracking in Drosophila.     

Crossmodal visual input for odor tracking during fly flight

Flies generate robust and high-performance olfactory and visual behaviors. Adult fruit flies can distinguish small differences in odor concentration across antennae separated by less than 1 mm [1], and a single olfactory sensory neuron is sufficient for near-normal gradient tracking in larvae [2]. During flight a male housefly chasing a female executes a corrective turn within 40 ms after a course deviation by its target [3]. The challenges imposed by flying apparently benefit from the tight integration of unimodal sensory cues. Crossmodal interactions reduce the discrimination threshold for unimodal memory retrieval by enhancing stimulus salience [4], and dynamic crossmodal processing is required for odor search during free flight because animals fail to locate an odor source in the absence of rich visual feedback [5]. The visual requirements for odor localization are unknown. We tethered a hungry fly in a magnetic field, allowing it to yaw freely, presented odor plumes, and examined how visual cues influence odor tracking. We show that flies are unable to use a small-field object or landmark to assist plume tracking, whereas odor activates wide-field optomotor course control to enable accurate orientation toward an attractive food odor.     

Visual stabilization dynamics are enhanced by standing flight velocity

A flying insect must travel to find food, mates and sites for oviposition, but for a small animal in a turbulent world this means dealing with frequent unplanned deviations from course. We measured a fly''s sensory-motor impulse response to perturbations in optic flow. After an abrupt change in its apparent visual position, a fly generates a compensatory dynamical steering response in the opposite direction. The response dynamics, however, may be influenced by superimposed background velocity generated by the animal''s flight direction. Here we show that constant forward velocity has no effect on the steering responses to orthogonal sideslip perturbations, whereas constant parallel sideslip substantially shortens the lags and relaxation times of the linear dynamical responses. This implies that for flies stabilizing in sideslip, the control effort is strongly affected by the direction of background motion.     

Search strategies of fruit flies in steady and shifting winds in the absence of food odours

Abstract. Flight behaviour by females of two species of fruit flies, Drosophila funebris and Drosophila immigrans , was videorecorded in a wind tunnel in still air and in wind with a constant or shifting direction. Flies which were deprived of food overnight took flight in the absence of food odours. Both species responded to the presented winds in agreement with two models that predict the shortest distance to an odour plume. According to these models, the shortest distance to an odour plume is travelled when insects fly at right angles to the wind with a steady direction. In winds shifting over more than 60, the shortest distance to an odour plume is achieved when insects fly parallel to the time-averaged wind direction. We propose a behavioural mechanism which accomplishes the observed flight directions taken by the two species of fruit flies in response to the tested wind regimes.     

Proboscis extension response (PER) assay in Drosophila

Proboscis extension response (PER) is a taste behavior assay that has been used in flies as well as in honeybees.On the surface of the fly's mouth (labellum), there are hair-like structures called sensilla which houses taste neurons. When an attractive substance makes contact to the labellum, the fly extends its proboscis to consume the material. Proboscis Extension Response (PER) assay measures this taste behavior response, and it is a useful method to learn about food preferences in a single fly. Solutions of various sugars, such as sucrose, glucose and fructose, are very attractive to the fly. The effect of aversive substances can also be tested as reduction of PER when mixed in a sweet solution.Despite the simplicity of the basic procedure, there are many things that can prevent it from working. One of the factors that requires attention is the fly's responsive state. The required starvation time to bring the fly to the proper responsive state varies drastically from 36 to 72 hours. We established a series of controls to evaluate the fly's state and which allows screening out of non-responsive or hyper-responsive individual animals. Another important factor is the impact level and the position of the contact to the labellum, which would be difficult to describe by words. This video presentation demonstrates all these together with several other improvements that would increase the reproducibility of this method.     

Three‐dimensional flight tracking shows how a visual target alters tsetse fly responses to human breath in a wind tunnel

Tsetse flies Glossina spp. (Diptera; Glossinidae) are blood‐feeding vectors of disease that are attracted to vertebrate hosts by odours and visual cues. Studies on how tsetse flies approach visual devices are of fundamental interest because they can help in the development of more efficient control tools. The responses of a forest tsetse fly species Glossina brevipalpis (Newstead) to human breath are tested in a wind tunnel in the presence or absence of a blue sphere as a visual target. The flight responses are video recorded with two motion‐sensitive cameras and characterized in three dimensions. Although flies make meandering upwind flights predominantly in the horizontal plane in the plume of breath alone, upwind flights are highly directed at the visual target presented in the plume of breath. Flies responding to the visual target fly from take‐off within stricter flight limits at lower ground speeds and with a significantly lower variance in flight trajectories in the horizontal plane. Once at the target, flies fly in loops principally in the horizontal plane within 40 cm of the blue sphere before descending in spirals beneath it. Successful field traps designed for G. brevipalpis take into account the strong horizontal component in local search behaviour by this species at objects. The results suggest that trapping devices should also take into account the propensity of G. brevipalpis to descend to the lower parts of visual targets.     

On the invariance of visual stimulus efficacy with respect to variable spatial positions

Summary In the fly,Calliphora erythrocephala, visual stimuli presented in an asymmetrical position with respect to the fly elicit roll or tilt movements of the head by which its dorsal part is moved towards the light areas of the surroundings (Figs. 4–7). The influence of passive body roll and tilt (gravitational stimulus) on the amplitude of these active head movements was investigated for two types of visual stimuli: (1) a dark hollow hemisphere presented in different parts of the fly's visual field, and (2) a moving striped pattern stimulating the lateral parts of one eye only.The response characteristics of the flies in the bimodal situation in which the gravitational stimulus was paired with stimulation by the dark hollow hemisphere can be completely described by the addition of the response characteristics for both unimodal situations, i.e. by the gravity-induced and visually induced characteristics (Figs. 8, 9). Therefore, the stimulus efficacy of the dark hollow hemisphere is independent of (=invariant with respect to) the flies' spatial position. The advantage of this type of interaction between gravity and visual stimulation for the control of body posture near the horizontal is discussed.In contrast, the efficacy of moving patterns depends on (=non-invariant with respect to) the spatial position of the walking fly. Regressive pattern movements exhibit their stronger efficacy with respect to progressive ones only when the gravity receptor system of the legs is stimulated. The stronger efficacy of downward vs upward movements can only be demonstrated when the flies are walking horizontally, independently of whether the leg gravity receptor system is stimulated by gravity or not (Fig. 10).The results are discussed with respect (1) to the invariance and non-invariance of the efficacy of visual stimuli with respect to the direction of the field of gravity, (2) to the formation of reference lines by the gravitational field which are used by the walking fly to determine the orientation of visual patterns, and (3) to the possible location of the underlying convergence between gravitationally and visually evoked excitation. As all types of head responses occur only in walking flies, we also discussed the possible influences of some physiological processes like arousal, proprioceptive feedback during walking and various peripheral sensory inputs on the performance of behavioural responses in the fly (Fig. 11).     

Responses of tsetse to ox sebum: a video study in the field

The behaviour of tsetse (mainly Glossina pallidipes Austen) around odour-baited targets, with or without a coating of ox sebum, was recorded in the field using video. The addition of sebum increased the total time a fly was in contact with the target, as well as the time spent flying around and landing on it. When carbon dioxide was released as part of the attractant odour plume, the presence of sebum on the target increased the number of landings made by each fly, but did not significantly affect the duration of each contact. When carbon dioxide was absent from the odour plume, sebum did not affect the number of landings made by flies but the duration of each contact with the target did increase. Evidence for an interactive effect of sebum and carbon dioxide was obtained. In addition, the presence of sebum on the target increased the percentage of landed flies which walked on its surface; such behaviour may represent an 'inspection' of the artificial host. The potential tsetse control application of the current findings are discussed.     

Dynamic properties of large-field and small-field optomotor flight responses in <Emphasis Type="Italic">Drosophila</Emphasis>

Optomotor flight control in houseflies shows bandwidth fractionation such that steering responses to an oscillating large-field rotating panorama peak at low frequency, whereas responses to small-field objects peak at high frequency. In fruit flies, steady-state large-field translation generates steering responses that are three times larger than large-field rotation. Here, we examine the optomotor steering reactions to dynamically oscillating visual stimuli consisting of large-field rotation, large-field expansion, and small-field motion. The results show that, like in larger flies, large-field optomotor steering responses peak at low frequency, whereas small-field responses persist under high frequency conditions. However, in fruit flies large-field expansion elicits higher magnitude and tighter phase-locked optomotor responses than rotation throughout the frequency spectrum, which may suggest a further segregation within the large-field pathway. An analysis of wing beat frequency and amplitude reveals that mechanical power output during flight varies according to the spatial organization and motion dynamics of the visual scene. These results suggest that, like in larger flies, the optomotor control system is organized into parallel large-field and small-field pathways, and extends previous analyses to quantify expansion-sensitivity for steering reflexes and flight power output across the frequency spectrum.     

Flying mate detection and chasing by tsetse flies (Glossina)

Abstract Male tsetse flies, probably Glossina morsitans morsitans Westw., were video-recorded in the field as they took off and chased other tsetse flies. Chasers responded (took off) to a target fly at a maximum distance of c. 55 cm, when it subtended c. 1.6^o to their eye (–1 foveal ommatidial subtense). Chased targets were always within this range (mean subtense at take-off = 3.2^o) and approaching the chaser. The most significant difference between chased and non-chased targets was in the rate of approach of the target fly in terms of the increase in its image size immediately before the chaser took off ( 21^o s⁻¹), especially as its relative increase (690% s^-1 P< 0.005). No feature of the target's translational velocity, nor any relationship between that and the image size approached this level of significance. Chasers seemed to 'slipstream' their target at c. 20 cm directly behind it, perhaps suggesting target identification by speed matching. Chases were apparently abandoned when the target image shrank from covering at least two of the chaser's foveal ommatidia to covering only one. Parallax-free measurements of flight speeds indicated a preferred, stable mean groundspeed of 4.8±0.1 m s^_1 (SE), at a mean wing-beat frequency of 209±3 Hz.