JN Fuel your research with LabChart
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

J Neurophysiol 95: 3391-3400, 2006. First published February 1, 2006; doi:10.1152/jn.00024.2006
0022-3077/06 $8.00
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
95/6/3391    most recent
00024.2006v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (10)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Santer, R. D.
Right arrow Articles by Simmons, P. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Santer, R. D.
Right arrow Articles by Simmons, P. J.

Role of an Identified Looming-Sensitive Neuron in Triggering a Flying Locust's Escape

Roger D. Santer, F. Claire Rind, Richard Stafford and Peter J. Simmons

School of Biology and Psychology, University of Newcastle upon Tyne, Newcastle upon Tyne, United Kingdom

Submitted 10 January 2006; accepted in final form 31 January 2006


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 GRANTS
 ACKNOWLEDGMENTS
 REFERENCES
 
Flying locusts perform a characteristic gliding dive in response to predator-sized stimuli looming from one side. These visual looming stimuli trigger trains of spikes in the descending contralateral movement detector (DCMD) neuron that increase in frequency as the stimulus gets nearer. Here we provide evidence that high-frequency (>150 Hz) DCMD spikes are involved in triggering the glide: the DCMD is the only excitatory input to a key gliding motor neuron during a loom; DCMD-mediated EPSPs only summate significantly in this motor neuron when they occur at >150 Hz; when a looming stimulus ceases approach prematurely, high-frequency DCMD spikes are removed from its response and the occurrence of gliding is reduced; and an axon important for glide triggering descends in the nerve cord contralateral to the eye detecting a looming stimulus, as the DCMD does. DCMD recordings from tethered flying locusts showed that glides follow high-frequency spikes in a DCMD, but analyses could not identify a feature of the DCMD response alone that was reliably associated with glides in all trials. This was because, for a glide to be triggered, the high-frequency spikes must be timed appropriately within the wingbeat cycle to coincide with wing elevation. We interpret this as flight-gating of the DCMD response resulting from rhythmic modulation of the flight motor neuron's membrane potential during flight. This means that the locust's escape behavior can vary in response to the same looming stimulus, meaning that a predator cannot exploit predictability in the locust's collision avoidance behavior.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
METHODS
 RESULTS
 DISCUSSION
 GRANTS
 ACKNOWLEDGMENTS
 REFERENCES
 
When predators attack, prey animals must escape quickly. Often, the startle behaviors they use are triggered by large, identified single neurons or small groups of them (Edwards et al. 1999Go; Korn and Faber 2005Go; Levi and Camhi 2000Go; Nolen and Hoy 1984Go). The relative simplicity of these networks reflects the need for a quick escape.

Small avian insectivores are common locust predators—carmine bee-eaters, for example, specialize in capturing locusts in flight (Fry and Fry 1992Go). When such birds attack a flying locust, they appear to it as looming stimuli: their silhouettes expand across its retina (Wheatstone 1852Go). Looming stimuli particularly excite the lobula giant movement detectors (LGMDs) of locusts, an identified bilateral pair of visual neurons that respond to looming stimuli with trains of spikes that increase in frequency as objects approach (Gabbiani et al. 2002Go; Hatsopoulos et al. 1995Go; Rind and Simmons 1992Go; Schlotterer 1977Go). The spikes of each LGMD are transmitted faithfully to a second identified neuron, the descending contralateral movement detector (DCMD), which transmits these spikes to the thoracic motor centers (Burrows and Rowell 1973Go; Killmann and Schürmann 1985Go; Rind 1984Go; Simmons 1980Go).

In response to looming stimuli of predator-like speeds and sizes (Rind and Santer 2004Go; Santer et al. 2005Go), flying locusts briefly interrupt flight with a raised-wing gliding behavior (Robertson and Reye 1992Go; Santer et al. 2005Go). These glides are interpreted as an escape response because they are similar to the dives used by many insect species to evade bats (e.g., Dawson et al. 2004Go; Hoy et al. 1989Go; Roeder 1962Go). A burst of spikes in the second tergosternal flight motor neuron (MN84, an elevator), raises the locust's wings into the gliding posture. DCMD spikes cause short-latency excitatory postsynaptic potentials (EPSPs) in this motor neuron that are larger than those mediated by the DCMD in other flight motor neurons (Simmons 1980Go). Glides are triggered by stimuli that optimally excite the DCMD, during the most vigorous part of its response (Santer et al. 2005Go).

Identified neurons trigger animals' startle behaviors in many ways. Occasionally, single spikes trigger complete behaviors, such as Mauthner neuron–triggered C-starts of teleosts (Korn and Faber 2005Go) and giant interneuron–triggered tail flips of crayfish (Edwards et al. 1999Go). More usually, spike trains in sets of neurons are integrated to trigger and steer escape, e.g., wind-sensitive giant interneurons triggering cockroach escape running (Levi and Camhi 2000Go). Often an emergency behavior must occur during another, ongoing behavior, such as flight, and in these cases the trigger signal must be integrated with the ongoing behavior (e.g., corrective steering by flying locusts; Reichert and Rowell 1986Go).

In this study we investigate the role of the DCMD in triggering escape glides. We find that the DCMD is the sole looming-excited input to MN84 and that high-frequency DCMD spikes >150 Hz cause EPSPs that sum strongly in MN84. Modified looming stimuli that cut short the high-frequency spikes of a DCMD reduce glide occurrence. However, high-frequency DCMD spikes alone cannot trigger a glide—these must occur during the appropriate phase of the wingbeat cycle to be gated into the flight rhythm.


   METHODS
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
RESULTS
 DISCUSSION
 GRANTS
 ACKNOWLEDGMENTS
 REFERENCES
 
Experiments were performed on 40 adult male and female Locusta migratoria L. from a crowded colony maintained at the University of Newcastle upon Tyne, or from an outside supplier (Blades Biological, Edenbridge, UK).

Visual stimuli

Locusts were challenged with looming visual stimuli displayed on a Kikusui COS1611 X-Y monitor controlled by a Cambridge Research Systems (Cambridge, UK) VSG2/3 image synthesizer and RG2 raster-generator system. The monitor screen was aligned with the center of one of a locust's compound eyes, 70 mm from it and parallel with the locust's long axis. Stimuli were computer-generated 80-mm-diameter black discs that loomed straight toward the eye at a constant speed. The discs loomed over a simulated distance of 2 m and a full loom ended when the stimulus was displayed at the monitor screen, 70 mm from the locust's eye and subtending 60° at the eye. The delay between the end of image movement and collision with the simulated disc would have been 70, 35, 23.3, and 14 ms at 1, 2, 3, and 5 m/s, respectively.

Intracellular recordings

We made intracellular recordings from the right mesothoracic second tergosternal motor neuron (MN84, nomenclature here and throughout after Snodgrass 1929Go) and DCMD simultaneously to study excitation reaching the motor neuron during a loom. Intracellular recordings were made from processes of the flight motor neuron in the dorsal neuropil of the mesothoracic ganglion and from the cell body of the DCMD in the protocerebrum. In preparation, a locust was secured upright on a block of modeling clay. The thoracic nervous system was exposed and intracellular recordings were made as described in Simmons (1980)Go. The protocerebrum was exposed as described in Rind (1984)Go. The mesothoracic and metathoracic ganglia were stabilized on a platform manipulated through the thoracic–abdominal connective nerves, and the brain on a platform manipulated from the front. Electrodes (DC resistance 40–60 M{Omega}, filled with 2 M potassium acetate) were attached to amplifiers with bridge circuits that allowed current to be injected through the recording electrodes. MN84 was identified by correlating single spikes in it, heard over an audio monitor, with twitches of the posterior part of the mesothoracic tergosternal muscle (M84), viewed under the dissecting microscope.

Extracellular recordings from restrained locusts

We made DCMD recordings from minimally dissected locusts to study the effects of manipulations of the looming stimulus on their DCMD responses. A locust was mounted ventral side up and its head was tilted slightly forward to expose the ventral sclerite of the neck. A small hole was made in the right side of the sclerite, and a 50-µm copper wire, insulted but for its tip, was inserted about 1.5 mm beneath the cuticle. A second copper wire was placed in the abdomen as a reference electrode. In these experiments, consecutive looming stimuli were separated by 150 s, during which a locust was dishabituated by tactile stimulation of the hindleg for 10 s.

Behavioral and extracellular recordings from flying locusts

We recorded behavioral responses of tethered flying locusts to looming stimuli. A locust was tethered to a brass rod by its pronotum and placed in front of a laminarized 3 m/s wind source that evokes typical, strong flight behavior at normal wingbeat frequency and posture (Santer et al. 2005Go). A MotionScope PCI high-speed digital camera (Redlake, San Diego, CA) recorded movements at 125 frames/s or, in one experiment, wing movements were monitored using an infrared beam that was broken by the beating wing (as in Santer et al. 2005Go). In some experiments, we cut one or the other of a locust's ventral nerve cords to abolish DCMD input on that side. To cut a ventral nerve cord, a window was cut in the ventral cuticle of the mesothorax and the nerve cord was sectioned between the pro- and mesothoracic ganglia. The thorax was then resealed with wax and the locust's flight behavior recorded during the approach of simulated objects from the left and right (15 approaches from each side). In other experiments, DCMD spikes and flight muscle activity were recorded during tethered flight. DCMD spikes were recorded as in Santer et al. (2005)Go using an electrode consisting of two 150-µm silver wire hooks implanted through a small window cut in the ventral cuticle of the mesothorax so that the hooks encircled the right pro-mesothoracic ventral nerve cord. Electromyograms (EMGs) were made from flight muscles using pairs of 50-µm copper wires insulated but for their tips. In all cases a 150-s intertrial interval was used, except in DCMD recordings from flying locusts where a shorter interval was sometimes used. In these experiments no habituation was evident in the DCMD.

Data capture and analysis

Electrophysiological data were recorded using Spike2 v.5 (Cambridge Electronic Design, Cambridge, UK). In tethered flying locusts DCMD spikes were obscured by flight muscle activity, which was removed using a high-pass filter with a low cutoff point determined from power spectra of typical DCMD recordings (Santer et al. 2005Go). Recordings in which DCMD spikes were not clear and the largest in the filtered data were discarded.

Statistical analyses were carried out using either SPSS v.11 (SPSS, Chicago, IL) or Minitab v.13.1 (Minitab, State College, PA). To analyze gliding occurrence in response to prematurely ending looming stimuli, a one-way ANOVA of glide occurrence with time removed from end of loom (a fixed factor) was carried out using SPSS. A post hoc Dunnett t-test was used to compare glide occurrence to prematurely ending stimuli with that to the full looming stimulus. To analyze differences in the DCMD responses of locusts during gliding and nongliding trials, spike trains were compared in Minitab using a three-factor repeated-measures ANOVA of "individual locust" (the repeated measure), "behavior" (glide or nonglide), and "time bin nested within gliding and nongliding behaviors." We then compared differences in mean DCMD spike frequencies between gliding and nongliding trials for each locust at each 10-ms time bin, using post hoc Student–Newman–Keuls (SNK) tests. To analyze the timing of high-frequency DCMD spikes within the wingbeat cycles of flying locusts, we plotted instantaneous DCMD spike frequencies for each recorded trial using wingbeat phase, rather than time, as the x-axis. We used forewing depression before a glide, or at an equivalent time where no glide occurred, to synchronize all trials. We then arranged these plots into groups according to the behavior of the locust in that trial, and for each group we plotted the instantaneous DCMD spike frequency by wingbeat phase data on the same axes. This gave us plots of instantaneous DCMD spike frequency against wingbeat phase for each behavior group. We analyzed these by using a MATLAB (The MathWorks, Natick, MA) script to overlay a 30 x 30 grid (with 0.067 wingbeat unit x 26.67-Hz sectors) over these plots and to record the occurrence of individual DCMD spikes within each sector. These spike occurrences were then plotted as contour plots showing the mean number of spikes per trial per sector for gliding and nongliding trials. Throughout, we indicate variability in results using SD.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
DISCUSSION
 GRANTS
 ACKNOWLEDGMENTS
 REFERENCES
 
The looming-elicited gliding behavior of the locust is triggered by the same stimuli that optimally excite the DCMD neuron (Santer et al. 2005Go). A key element of this behavior is contraction of the second tergosternal muscle (M84), caused by a burst of potentials in it, that raises the forewings and holds them elevated above the hindwings in the stereotyped gliding posture (Santer et al. 2005Go). Because DCMD spikes induce larger EPSPs in motor neuron (MN) 84 than in other flight motor neurons (Simmons 1980Go), we focused our investigation on the role of this pathway in the forewing movements integral to the gliding behavior.

DCMD input to MN84 during a loom

Our first step was to assess the role of DCMD excitation of MN84 in the gliding behavior. Although MN84 is known to be excited by the DCMD (Simmons 1980Go), the way that this motor neuron responds to visual stimuli has not been described. We wanted to discover whether MN84 was excited by the approach of a looming stimulus and whether the DCMD neuron was the sole source of this excitation. To do this, we made simultaneous intracellular recordings from MN84 and both intra- and extracellular recordings from the DCMD in restrained locusts. By injecting steady depolarizing currents into the DCMD, we evoked spikes that mediated EPSPs one for one in MN84 (Fig. 1A). The maximum frequency of spikes evoked by current injection into the DCMD cell body was 85 Hz, and at this frequency there was little or no summation of the EPSPs, which had a duration of about 10 ms, in MN84 (Fig. 1A, inset). In response to looming stimuli, the EPSPs mediated by a DCMD summed in a frequency-dependent manner for instantaneous frequencies {gtrsim}100 Hz (Fig. 1B). Sometimes, in response to looming stimuli, the EPSPs summed sufficiently to trigger a spike (Fig. 1C).


Figure 1
View larger version (32K):
[in this window]
[in a new window]
  		FIG. 1. Descending contralateral movement detector (DCMD) provides looming-excited input to the second mesothoracic tergosternal motor neuron (MN84). A: DCMD spikes at 85 Hz were elicited by a steady DC current of +15 nA injected into its cell body and were monitored in an extracellular nerve cord recording. Each spike mediated an excitatory postsynaptic potential (EPSP) in MN84, and successive EPSPs showed little summation. Inset: average of 52 EPSPs triggered from DCMD spikes. B: intracellular recordings from the DCMD cell body and from a neuropil process of MN84 during the approach of an object at 1.5 m/s. Bottom trace monitors stimulus image size, and dots show instantaneous DCMD spike frequency. Match between the pattern of EPSPs in MN84 and spikes in the DCMD indicates that the DCMD is the only source of excitation to this motor neuron during a looming visual stimulus. C: at a faster approach speed of 5 m/s, the DCMD generated a series of spikes at >200 Hz and EPSPs in MN84 summed to the threshold for a spike (recordings from MN84 are shown at 2 different gains; after the spike, the trace is not shown for the higher-gain recording). D: injection of steady hyperpolarizing current (–20 nA) into the DCMD cell body made the DCMD respond to a looming stimulus with bursts of spikes rather than its normal response. EPSPs in MN84 summate to follow each DCMD spike burst, and there is no other source of excitation to the motor neuron between the DCMD bursts. Top: comparison of DCMD spike times in separate trials with no current or with –20 nA injected. Main figure shows a further trial with –20 nA injected into the DCMD and simultaneous recording from MN84.

 
To assess whether DCMD spikes were the only excitatory input to MN84 during a looming stimulus, we injected hyperpolarizing current into a DCMD cell body, which caused DCMD excitation to occur in bursts rather than following its normal pattern in response to a looming stimulus (we could not completely prevent spikes in the DCMD in response to the looming stimulus by injecting hyperpolarizing currents into its cell body). A 2 m/s looming stimulus induced a steady increase in DCMD spike frequency when no current was injected (Fig. 1D, top trace); however, the same stimulus induced three distinct bursts of DCMD spikes when –20 nA was injected into the DCMD (Fig. 1D, second trace and main part of the figure). With hyperpolarizing current injected, the pattern of EPSP summation in MN84 closely followed the altered pattern of spikes in the DCMD (Fig. 1D), indicating very strongly that the DCMD is the only source of visually generated excitation to MN84 in response to a looming stimulus.

Manipulation of the DCMD response and effects on behavior

Next, we wanted to investigate whether a burst of high-frequency DCMD spikes, which would cause strongly summing EPSPs in MN84, was necessary for the production of a gliding behavior in a flying locust. If this were the case, abolishing or altering this element of the DCMD response should alter the occurrence of gliding behavior. Because the dissection necessary to expose the dorsally located DCMD axon and then kill it by intracellular injection would have prevented flight, we used two less direct methods to interfere with the normal DCMD response.

In our first manipulation, we altered the response of the DCMD by showing locusts looming stimuli that were modified to stop moving earlier than usual. In response to a complete 5 m/s loom, the DCMD of an aroused, restrained locust follows the expansion of the object and reaches high spike frequencies (>150 Hz) at the approximate time that the stimulus ceases moving (Fig. 2A). By incrementally reducing this stimulus, so that instead of ceasing movement at the monitor screen 70 mm from the locust's eye it stopped moving 10 or 20 ms earlier, we could remove the final high-frequency spikes from the DCMD response without altering the early part of its response before the end of stimulus movement (Fig. 2, B and C). Because our previous results showed the DCMD to be the only looming-excited input to MN84, this manipulation specifically affected this pathway. When we challenged undissected, tethered flying locusts with full and incrementally reduced 5 m/s looming stimuli, we found that gliding behavior was elicited in nearly 70% of trials in response to the full loom and to looms reduced by 5 and 10 ms (Fig. 2D). However, when the final 15 ms or more were omitted, the frequency of occurrence of gliding behavior was significantly reduced (Fig. 2D). Omitting the final 15 ms of a looming stimulus can remove three or more DCMD spikes at instantaneous frequencies ≥200 Hz (Fig. 2C).


Figure 2
View larger version (31K):
[in this window]
[in a new window]
  		FIG. 2. Manipulation of the DCMD response alters the occurrence of looming-elicited gliding behavior. A: a full 5 m/s looming stimulus simulating an approach of an 80-mm-diameter disc over 2 m and ending at the stimulus monitor screen, 70 mm from the locust's eye (subtense shown in bottom left), elicited a normal DCMD response in restrained locusts. B and C: when successive 10-ms intervals were removed from the loom, the DCMD's final high-frequency spike burst was cut short. Example stimulus subtenses for incrementally reduced looming stimuli are plotted in the bottom left and the time that the stimulus stopped moving is indicated by a triangle superimposed on the x-axis of plots A–C. D: when presented to tethered flying locusts, full 5 m/s looms and looms cut short by 5 or 10 ms resulted in similar occurrences of gliding behavior (as recorded by an infrared wingbeat sensor; see METHODS). When looms were cut short by >15 ms, the occurrence of gliding behavior was significantly reduced [**, behavioral occurrences that were significantly different (P < 0.001) from those elicited by the full looming stimulus as determined by one-way ANOVA (showing a significant effect of premature loom ending on glide occurrence; F(12,117) = 21.35, P < 0.001) and post hoc Dunnett t-test]. Labels A, B, and C indicate stimuli used in the production of A–C. E: occurrence of gliding behavior was reduced when stimuli were delivered to the eye contralateral to a cut nerve cord (with only ipsilaterally descending inputs to the thorax intact), but unchanged when stimuli were delivered ipsilateral to it (with only contralaterally descending inputs intact). A–D: 6 stimulus presentations to each of 10 locusts. E: 15 ipsilateral and 15 contralateral looms displayed to each of 4 locusts (2 sectioned each side).

 
In our second manipulation, we prevented spikes from one of the DCMDs from reaching the thoracic ganglia by sectioning one of the locust's ventral nerve cords. Under these conditions, locusts flew with reduced wingbeat frequencies (about 14 Hz), and lower than usual wingbeat amplitudes and flight durations. However, with looming stimuli delivered ipsilateral to their cut nerve cord, locusts still produced gliding behaviors in nearly 55% of trials (Fig. 2E), which is slightly less frequently than that in undissected locusts. When looming stimuli were delivered contralateral to the cut nerve cord, the frequency of occurrence of gliding behavior was significantly reduced to around 15% of trials (t-test; df = 6, t = 11.104, P < 0.001; Fig. 2E). These data implicate a contralaterally descending looming-sensitive neuron, such as the DCMD, in the production of the gliding response. They also indicate that the behavior only occurs occasionally in the absence of this input.

Correlating high-frequency spikes and behavior

So far, our data suggest an important role for high-frequency (>150 Hz) DCMD spikes in the production of gliding behavior. If these spikes act as a trigger for gliding, their occurrence should differ between spike trains recorded during trials with glides and spike trains recorded during trials where no glide occurred. In response to the repeated presentation of identical 5 m/s looming stimuli, glides do not always occur (Fig. 2D; in response to a full loom, glides occur in about 70% of presentations). If the DCMD triggers glides, then differences in DCMD responses between trials should reflect whether a glide occurs in a particular trial. Using implanted hook electrodes and simultaneous high-speed video recordings, we were able to obtain clear DCMD and strong flight data from four locusts (from a total of 10 experiments) responding to a total of 104 presentations of a 5 m/s looming stimulus. The low sample number was partly attributable to the low success rates of the experiment and the need to analyze in fine detail the behavior and neuronal data from a low number of samples. As previously noted, gliding behaviors occurred less often when locusts had electrodes implanted (Santer et al. 2005Go). Example DCMD and flight recordings from a tethered flying locust are shown in Fig. 3A.


Figure 3
View larger version (27K):
[in this window]
[in a new window]
  		FIG. 3. DCMD responses in flying locusts differ when they respond to a looming stimulus with a glide and when they produce no apparent response. A: an extracellular recording of DCMD activity in a tethered flying locust responding to a 5 m/s looming stimulus (subtense shown in C). Forewing movements are shown in the top trace (forewing tip position, at maximum elevation and depression only), measured from a recording at 125 frames/s. Locust performs a glide during the most vigorous period of the DCMD response. Bottom trace: DCMD activity. B: mean DCMD responses to 5 m/s looming stimuli (10-ms bins) during trials where locusts did and did not glide. C: looming stimulus subtense at the locusts' eyes. D: number of consecutive >200-Hz DCMD spikes from spike trains recorded during glides and nonglides. Boxes show the 25th and 75th percentiles and bisecting lines indicate the median value. Whiskers indicate the 5th and 95th percentiles and outliers are plotted. A: a single recording from a tethered flying locust. B and D: 104 stimulus presentations to 4 flying locusts divided into 50 glides and 54 no responses using the gliding criterion of Santer et al. (2005) (see RESULTS). In all panels, stimulus is a simulated 80-mm-diameter disc looming over 2 m at 5 m/s.

 
We divided the DCMD responses of the flying locusts into two groups based on whether a gliding behavior was observed. We defined a glide as a pause in flight >1.25 x the mean wingbeat duration of the preceding 10 wingbeat cycles (as in Santer et al. 2005Go). If gliding behavior were triggered solely by the DCMD neuron, we hypothesized that DCMD responses should differ between trials where glides were performed and trials where glides were not performed (Fig. 3B). To test for differences in the DCMD spike frequencies of gliding and nongliding trials, we focused on a 100-ms period between 60 ms before and 40 ms after the end of stimulus movement (from 74 ms before collision until 26 ms after collision) because this was the period in which high-frequency DCMD spikes that could cause summing EPSPs in MN84 occurred. This is also the period during which glides are triggered (Santer et al. 2005Go). The mean frequency of DCMD spikes, averaged over all locusts and all time bins, was slightly higher during gliding than during nongliding trials over the analysis period (242 vs. 226 Hz, respectively). To test for any significant differences in spike frequency between gliding and nongliding trials, and to account for potential interactions between the different behaviors shown during trials (gliding and nongliding), the different locusts and the different time bins, a three-factor repeated-measures ANOVA was performed. "Time bin" was nested within "behavior" (this was done to allow variation between the time bins to be taken into account because we expect the DCMD responses to change over time as a result of the looming stimulus, and still test for differences between DCMD responses in gliding and nongliding trials in spite of this variation). "Locust" was designated a repeated measure because both gliding and nongliding behaviors occurred in trials from the same locust. The ANOVA gave a significant interaction term [locust x time bin (behavior), F(54,1030) = 2.95, P < 0.001]. This meant that the main effect "behavior" was not interpretable because it was qualified by the complex interaction between locust and time bin (behavior) (e.g., Underwood 1997Go). Therefore, to investigate where differences between DCMD responses during gliding and nongliding trials (behaviors) occurred, we needed to use post hoc tests. We conducted post hoc SNK tests comparing differences between DCMD responses recorded during gliding and nongliding trials for each locust at each 10-ms time bin. Most of these multiple comparisons showed no significant differences between the DCMD responses during gliding and nongliding trials. However, the spike frequencies recorded during glides were significantly higher (P < 0.05) than those recorded during nonglides for locust 1, in the 10-ms time bin starting at +16 ms (relative to the time of collision); for locust 2 at the –64-, –54-, –34-, and –24-ms time bins; for locust 3 at the –54-ms time bin; and for locust 4 at the –54-ms time bin. These data demonstrate that there are differences in DCMD spike frequency—at certain times within the final period of the DCMD response—between trials where glides occurred and trials where glides did not occur. The exact timing of these differences in responses varies between locusts, but they could potentially be involved in glide triggering. We have already shown that when this terminal period of a loom is removed, the occurrence of gliding behavior was significantly reduced (Fig. 2).

If a criterion number of high-frequency DCMD spikes acts as a glide trigger, this DCMD response feature should be found in all spike trains recorded from trials in which a locust glides, but not in the spike trains from trials where it does not glide. Despite the differences in the mean DCMD responses, we found no consistent differences on a trial-by-trial basis for any DCMD response features when we tested numbers of consecutive spikes >100, >150, >200, or >250 Hz (all spike train features that should cause EPSPs to summate strongly in MN84). For example, although there were never fewer than six consecutive >200-Hz DCMD spikes in gliding trials, both gliding and nongliding groups could contain similar larger numbers of consecutive spikes >200 Hz (Fig. 3D). This may indicate that triggering of the gliding behavior is a stochastic process or that the criterion DCMD response for triggering the complete behavior is more complex than a combination of spike frequency and number.

Flight-gating of the DCMD response

A likely reason that a high DCMD spike rate does not always trigger a glide is that, during flight, the membrane potential of flight motor neurons is modulated in a cyclical fashion by inputs from the flight central pattern generator (Hedwig and Becher 1998Go; Robertson and Pearson 1982Go, 1985Go). As a result, the effects of EPSPs from a DCMD will depend on the wingbeat phase in which they arrive (Reichert and Rowell 1985Go, 1986Go; Reichert et al. 1985Go). Consequently, a glide could be gated appropriately into flight, preventing elevator activity during wing depression and thus potential musculoskeletal damage.

Glides usually follow a normal and complete wingbeat cycle and, before a glide, a locust's wingbeat frequency declines during the approach of the looming stimulus (Santer et al. 2005Go). We analyzed the forewing movements of gliding locusts and synchronized them using the time of forewing elevation into the glide as a reference (Fig. 4). In all trials where a glide occurred, it followed a complete, although very occasionally reduced amplitude, wingbeat (Fig. 4A); wingbeat adjustments were evident during the approach of a looming stimulus (Fig. 4B). These data suggest that DCMD input must be gated into an appropriate phase of the wingbeat cycle for gliding to be triggered because glides can only occur at one point in the wingbeat cycle. This may also explain why the timing of differences in the DCMD responses from gliding and nongliding trials differed between locusts.


Figure 4
View larger version (26K):
[in this window]
[in a new window]
  		FIG. 4. Gliding behaviors occur at a set point in the flight rhythm. A: forewing elevation (marked only during maximum wing elevation and depression) during the approach of a looming stimulus for trials in which a locust performed glides. Trials are aligned so that wing depression before the glide is at t = 0. Glides always followed a complete and apparently normal wingbeat cycle. B: wingbeat frequency declines during the approach of a looming stimulus. Plot shows mean instantaneous wingbeat frequency for the 6 wingbeats preceding a glide for the trials plotted in A.

 
To investigate whether the DCMD response is flight-gated, we examined further the flight data of the four locusts used in the previous analysis. If high-frequency DCMD spikes were gated into the flight rhythm, we expected to find their distribution—within the wingbeat cycle and relative to normal MN84 activity—to be different between trials where glides were performed and trials where glides were not performed.

We plotted DCMD responses relative to wingbeat phase for all trials in which glides occurred and for all trials in which no glides occurred. One wingbeat phase unit described the period from one wingbeat depression to the next, and phase 0 was the phase at which the forewing was last fully depressed before a glide, or the equivalent relative to stimulus movement for trials in which there was no glide (in these trials the forewing was often elevated more than usual at the time a glide would normally occur). For each trial we plotted the instantaneous DCMD spike frequencies for each spike in a response against forewing phase for that trial (indicated schematically in Fig. 5A, left). For each behavior category (glides and no glides), we then combined data from all trials into a plot of instantaneous DCMD spike frequencies against wingbeat phase. This plot included a total of 1,343 individual DCMD spikes recorded during nongliding trials and 1,518 recorded during gliding trials over the two wingbeat analysis period. As described in METHODS, we overlaid each plot with a 30 x 30-sector grid (indicated schematically in Fig. 5A). Within each grid box, we counted the number of spike occurrences. These numbers were divided by the number of trials constituting the group and transferred into shading density on contour plots of DCMD instantaneous spike frequency against wingbeat phase (Fig. 5, C and D). The most densely shaded area on each of these plots (indicating the most commonly occurring spike frequencies and timings) occurred at a spike frequency >200 Hz, but at a wingbeat phase that differed between behaviors. In trials that ended with glides (Fig. 5C), the area with densest shading occurred during and after phase 0, when the wing was elevating into a glide and an extended burst of spikes in MN84 was occurring (Fig. 5, B and C). In trials where a glide was not performed, high-frequency DCMD spikes still occurred but were distributed during the downstroke, before normal MN84 spikes and rarely extending past phase 0 (Fig. 5, B and D). Therefore the timing of the DCMD's high-frequency spikes, relative to the wingbeat cycle and normal MN84 activity was a good predictor of gliding behavior occurrence. The necessity for these spikes to occur at (and past) the time that muscle 84 is normally active indicates that they must be gated into the flight rhythm by modulations in MN84 membrane potential resulting from flight central pattern generator activity.


Figure 5
View larger version (30K):
[in this window]
[in a new window]
  		FIG. 5. Effect of DCMD spikes on M84 activity depends on the wingbeat phase at which they occur. A: schematic illustration of data preparation. DCMD spike trains of gliding and nongliding locusts were compared by plotting the instantaneous frequencies of each DCMD spike during a trial against wingbeat phase (1 unit = 1 complete wingbeat cycle). Left: trials were then plotted together on the same axes for each behavior group (gliding and nongliding). Right: grid was overlayed over the plots and spike occurrence within each wingbeat phase/frequency bin was counted. In analysis, we used a 30 x 30 grid rather than the 3 x 3 illustrated here. Spike occurrence was standardized between the 2 groups by dividing spike counts per bin by number of trials. B: during a normal wingbeat cycle, elevator muscles 83 and 84 contract just before wing elevation. During a glide, muscle 84 continues to contract for a sustained period, holding the forewing elevated. C: during glides, high-frequency DCMD spikes >200 Hz occurred at the same time as the muscle 84 burst that defines a glide and are thus able to sum with normal rhythmic excitation of MN84 by the flight central pattern generator. D: when locusts did not glide, high-frequency DCMD spikes occurred before the elevator muscle burst and not during wing elevation (out of phase with normal MN84 activity). Data are as plotted in Fig. 3.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
GRANTS
 ACKNOWLEDGMENTS
 REFERENCES
 
We have investigated the role of the DCMD neuron in looming-elicited emergency gliding behavior of flying locusts, concentrating on the DCMD–MN84 pathway and forewing movements. We found that DCMD spikes are the only source of visually generated excitation to MN84 during a loom, and that EPSPs resulting from these spikes sum strongly—because of their duration—only when the DCMD spikes at >150 Hz. When these high-frequency spikes are removed from the DCMD response, the frequency of occurrence of gliding is reduced. Although the mean DCMD spike frequencies across a 100-ms period at the end of a loom are higher when locusts glide than when they do not, the times at which significant differences occur between gliding and nongliding spike trains, relative to the loom, are variable between locusts. The point in the wingbeat cycle where these high-frequency DCMD spikes occur is crucial to their effectiveness, probably arising from the large rhythmic input to MN84 during normal flight (Robertson and Pearson 1982Go, 1985Go). Thus we propose that high-frequency DCMD spikes, gated into the flight rhythm, trigger MN84 to raise a locust's forewings into the gliding posture.

High-frequency DCMD spikes and glide triggering

Looming-elicited glides are characterized by a burst of potentials in elevator M84 that cause it to contract and raise the locust's forewings into its gliding posture (Santer et al. 2005Go). The DCMD is the only source of excitation to MN84 during a loom and high-frequency DCMD spikes are sufficient to cause MN84 to spike in a restrained locust. In tethered flying locusts, shortening a 5 m/s loom by 15 ms caused a reduction in glide occurrence from about 70 to 30% of trials. This procedure selectively removes the final high-frequency spikes from the DCMD response. Also, the DCMD spike trains of the same flying locusts contained higher spike frequencies at variable times during the terminal period of a loom in trials where they performed a glide than in trials where they did not, providing further evidence for the need for a burst of high-frequency DCMD spikes during flight for a glide to occur.

The requirement for several high-frequency DCMD spikes contrasts with some well-characterized escape behaviors that are triggered by single spikes in identified neurons (Edwards et al. 1999Go; Korn and Faber 2005Go). It is similar to the bat-cry evasion behavior of field crickets that is triggered by sustained high-frequency spikes in an identified auditory interneuron (Nolen and Hoy 1984Go). Because high-frequency DCMD spikes occur late in its response to a looming stimulus, the requirement for them may reflect the need to trigger a glide when a predator is at close quarters, permitting the glide to be used as a last-chance escape behavior. Similar last-chance evasive behaviors are used by many nocturnally flying insects in response to bat cries with high pulse repetition rates, indicating that the attacking bat is close (Triblehorn and Yager 2005Go).

Sectioning all descending inputs to the thorax on the side contralateral to a looming stimulus greatly reduced the occurrence of gliding behavior, whereas sectioning ipsilateral inputs did not. This evidence strongly implicates a contralaterally descending looming detector and is consistent with our proposal that the DCMD triggers a glide. However, that glides could still occasionally be performed when contralaterally descending neurons were sectioned indicates that either these glides were chance occurrences, because flight in these animals was weaker and more erratic than normal, or that other pathways, perhaps involving the DIMD—an ipsilaterally descending movement detector (Burrows and Rowell 1973Go)—may trigger glides under some circumstances. This arrangement would be similar to that in teleosts, where the Mauthner neuron is the first to trigger a C-start, but is reinforced by additional, slower pathways that can trigger the behavior when the Mauthner neuron is ablated (Di Domenico et al. 1988Go; Eaton et al. 1982Go; Korn and Faber 2005Go).

Flight-gating of high-frequency DCMD spikes

The DCMD provides the only visually generated input to MN84 during a loom and high-frequency DCMD spikes can cause this motor neuron to spike, although high-frequency DCMD spikes can often be seen in trials where a locust does not glide. This is probably because the gliding behavior occurs only at the end of a normal, complete wingbeat cycle, so DCMD input must be gated into the appropriate phase of the flight cycle (Santer et al. 2005Go). During glides, the high-frequency DCMD spikes that cause summing EPSPs in MN84 occur coincidentally with the onset of the MN84 burst that elevates the forewings into the gliding posture and continue throughout it. They are thus summed with rhythmic modulations of the motor neuron's membrane potential during normal flight (Hedwig and Becher 1998Go; Robertson and Pearson 1982Go, 1985Go) to trigger powerful and sustained elevator muscle contraction. In contrast, where the locust does not glide, high-frequency DCMD spikes usually arrive before MN84 activity during the time when it is not strongly excited. In these cases DCMD input is insufficient to cause prolonged excitation of MN84, but an increased amplitude forewing elevation is often seen in these trials that might result from subtle modulation of MN84 activity by preceding DCMD spikes. DCMD spikes trigger glides on most occasions (approximately 70%) but where they cannot, potential musculoskeletal damage may be preclusive and forewing modulations an alternative tactic. In a similar way, it has been proposed that spike trains from the locust's deviation detector neurons are gated into the flight rhythm to evoke course correction during flight (Reichert and Rowell 1985Go, 1986Go; Reichert et al. 1985Go). DCMD activity could also directly affect the flight central pattern generator during a glide.

Behavioral significance

Locusts respond to an approaching object with steering behaviors and, if these fail, a last-chance emergency glide (Gray et al. 2001Go; Robertson and Johnson 1993Go; Robertson and Reye 1992Go; Santer et al. 2005Go). Low DCMD firing rates 200 ms before collision might trigger collision avoidance steering (Gray 2005Go; Matheson et al. 2004Go), whereas higher rates later in the response trigger a glide. Although tethered flying locusts most often perform glides in response to a loom, occasionally they do not. In these trials the forewing was often elevated more than usual. On a trial-by-trial basis, the behavior of the locust cannot be predicted from its DCMD response alone (or thus from the type of stimulus the locust encounters) because of the variation in the effectiveness of these spikes at different phases of the flight rhythm. In this way, the locust exhibits a form of protean behavior (Edut and Eilam 2004Go; Humphries and Driver 1967Go, 1971Go) that might prevent predictability in its escape responses being exploited by a predator (e.g., Jablonski and Strausfeld 2001Go). This range of behaviors, from steering to diving, also indicates that locusts' evasive behaviors vary with the level of threat posed by a looming stimulus. The evasive responses used by some nocturnal insects to evade bats also vary with the perceived danger posed by the stimulus (Miller and Olesen 1979Go; Roeder 1967Go; Yager et al. 1990Go).

Looming-elicited gliding by flying locusts appears to be an emergency response suited to the evasion of fast aerial predators such as the carmine bee-eater Merops nubicus, the lanner falcon Falco biarmicus, and the black kite Milvus migrans, which are all reported to capture locusts in flight (Fry and Fry 1992Go; Nickerson 1958Go; Smith and Popov 1953Go). When these birds attack locusts, they appear as looming stimuli and thus the DCMD neuron is ideally suited to their detection (e.g., Rind and Simmons 1992Go). Furthermore, the DCMD is most sensitive to objects moving in its caudal rather than frontal field of view (Krapp and Gabbiani 2005Go), suiting it to the detection of pursuing predators.

The relatively small diameter looming stimuli used in our study are close to the pectoral widths of typical avian locust predators and result in a DCMD response that increases in frequency until after stimulus movement has ceased (e.g., Money et al. 2005Go; Rind and Santer 2004Go; Santer et al. 2005Go). However, the DCMD's response to larger and/or slower stimuli is curtailed earlier in the loom as a result of feedforward inhibition to the LGMD (Gabbiani et al. 2005Go; Rind and Santer 2004Go). The important role for high-frequency spikes at the end of a looming stimulus in triggering an emergency glide raises the interesting question of whether the variation in DCMD response with object speed and size suits it to the triggering of different emergency behaviors according to the perceived nature of a looming threat. Because large but not small head-on looming objects trigger flight steering behaviors (Gray et al. 2001Go), it has been proposed that different avoidance reactions may be triggered by predators and conspecifics (Gray 2005Go; Matheson et al. 2004Go). However, rather than being very large stimuli, during flight the most specialized of these predators are small, fast birds with small pectoral diameters (about 30–45 mm) and thin profile wings that may be best evaded by gliding rather than steering (Santer et al. 2005Go).


   GRANTS
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 GRANTS
ACKNOWLEDGMENTS
 REFERENCES
 
This work was supported by the Biotechnology and Biological Sciences Research Council and European Union Grant LOCUST IST-2001-38097.


   ACKNOWLEDGMENTS
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 GRANTS
 ACKNOWLEDGMENTS
REFERENCES
 
The authors thank E. W. Childs, on whose preliminary observations elements of this work are based, and G. A. Wright for comments on an earlier version of this manuscript.


   FOOTNOTES
 
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Address for reprint requests and other correspondence: R. D. Santer, School of Biology and Psychology, Ridley Building, University of Newcastle upon Tyne, Newcastle upon Tyne, NE1 7RU, UK (E-mail: r.d.santer{at}0040ncl.ac.uk).


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 GRANTS
 ACKNOWLEDGMENTS
 REFERENCES
 
Burrows M and Rowell CHF. Connections between descending visual interneurons and metathoracic motoneurons in the locust. J Comp Physiol A Sens Neural Behav Physiol 85: 221–234, 1973.[CrossRef]

Dawson JW, Kutsch W, and Robertson RM. Auditory-evoked evasive manoeuvres in free-flying locusts and moths. J Comp Physiol A Sens Neural Behav Physiol 190: 69–84, 2004.[CrossRef][ISI][Medline]

Di Domenico R, Nissanov J, and Eaton RC. Lateralization and adaptation of a continuously variable behaviour following lesions of a reticulospinal command neuron. Brain Res 473: 15–28, 1988.[CrossRef][ISI][Medline]

Eaton RC, Lavender WA, and Wieland CM. Alternative neural pathways initiate fast-start responses following lesions of the Mauthner neuron in goldfish. J Comp Physiol A Sens Neural Behav Physiol 145: 485–496, 1982.[CrossRef]

Edut S and Eilam D. Protean behavior under barn-owl attack: voles alternate between freezing and fleeing and spiny mice flee in alternating patterns. Behav Brain Res 155: 207–216, 2004.[CrossRef][ISI][Medline]

Edwards DH, Heitler WJ, and Krasne FB. Fifty years of a command neuron: the neurobiology of escape behavior in the crayfish. Trends Neurosci 22: 153–161, 1999.[CrossRef][ISI][Medline]

Fry CH and Fry K. Kingfishers, Bee-Eaters and Rollers. London: A&C Black, 1992.

Gabbiani F, Cohen I, and Laurent G. Time-dependent activation of feed-forward inhibition in a looming-sensitive neuron. J Neurophysiol 94: 2150–2161, 2005.[Abstract/Free Full Text]

Gabbiani F, Krapp HG, Koch C, and Laurent G. Multiplicative computation in a visual neuron sensitive to looming. Nature 420: 320–324, 2002.[CrossRef][Medline]

Gray JR. Habituated visual neurons in locusts remain sensitive to novel looming objects. J Exp Biol 208: 2515–2532, 2005.[Abstract/Free Full Text]

Gray JR, Lee JK, and Robertson RM. Activity of descending contralateral movement detector neurons and collision avoidance behaviour in response to head-on visual stimuli in locusts. J Comp Physiol A Sens Neural Behav Physiol 187: 115–129, 2001.[CrossRef][Medline]

Hatsopoulos N, Gabbiani F, and Laurent G. Elementary computation of object approach by a wide-field visual neuron. Science 270: 1000–1003, 1995.[Abstract/Free Full Text]

Hedwig B and Becher G. Forewing movements and intracellular motoneurone stimulation in tethered flying locusts. J Exp Biol 201: 731–744, 1998.[Abstract/Free Full Text]

Hoy RR, Nolen T, and Brodfuehrer P. The neuroethology of acoustic startle and escape in flying insects. J Exp Biol 146: 287–306, 1989.[Abstract/Free Full Text]

Humphries DA and Driver PM. Erratic display as a device against predators. Science 156: 1767–1768, 1967.[Abstract/Free Full Text]

Humphries DA and Driver PM. Protean defence by prey animals. Oecologia 5: 285–302, 1971.

Jablonski PG and Strausfeld NJ. Exploitation of an ancient escape circuit by an avian predator: relationships between taxon-specific prey escape circuits and the sensitivity to visual cues from the predator. Brain Behav Evol 58: 218–240, 2001.[CrossRef][ISI][Medline]

Killmann F and Schürmann FW. Both electrical and chemical transmission between the "lobula giant movement detector" and the "descending contralateral movement detector " neurons of locusts are supported by electron microscopy. J Neurocytol 14: 637–652, 1985.[CrossRef][ISI][Medline]

Korn H and Faber DS. The Mauthner cell half a century later: a neurobiological model for decision-making? Neuron 47: 13–28, 2005.[CrossRef][ISI][Medline]

Krapp HG and Gabbiani F. Spatial distribution of inputs and local receptive field properties of a wide-field looming sensitive neuron. J Neurophysiol 93: 2240–2253, 2005.[Abstract/Free Full Text]

Levi R and Camhi JM. Wind direction coding in the cockroach escape response: winner does not take all. Neuroscience 20: 3814–3821, 2000.[Abstract/Free Full Text]

Matheson T, Rogers SM, and Krapp HG. Plasticity in the visual system is correlated with a change in lifestyle of solitarious and gregarious locusts. J Neurophysiol 91: 1–12, 2004.[Abstract/Free Full Text]

Miller LA and Olesen J. Avoidance behavior in green lacewings. I. Behavior of free flying green lacewings to hunting bats and ultrasound. J Comp Physiol 131: 113–120, 1979.[CrossRef]

Money TGA, Anstey ML, and Robertson RM. Heat stress-mediated plasticity in a locust looming-sensitive visual interneuron. J Neurophysiol 93: 1908–1919, 2005.[Abstract/Free Full Text]

Nickerson B. Some observations of the carmine bee-eater Merops nubicus Gmelin in the French Sudan. Ibis 100: 454–457, 1958.

Nolen TG and Hoy RR. Initiation of behaviour by single neurons: the role of behavioural context. Science 226: 992–994, 1984.[Abstract/Free Full Text]

Reichert H and Rowell CHF. Integration of nonphaselocked exteroceptive information in the control of rhythmic flight in the locust. J Neurophysiol 53: 1201–1218, 1985.[Abstract/Free Full Text]

Reichert H and Rowell CHF. Neuronal circuits controlling flight in the locust: how sensory information is processed for motor control. Trends Neurosci 9: 281–283, 1986.[CrossRef][ISI]

Reichert H, Rowell CHF, and Griss C. Course correction circuitry translates feature detection into behavioural actions in locusts. Nature 315: 142–144, 1985.[CrossRef][ISI]

Rind FC. A chemical synapse between two motion detecting neurons in the locust brain. J Exp Biol 110: 143–167, 1984.[Abstract/Free Full Text]

Rind FC and Santer RD. Collision avoidance and a looming sensitive neuron: size matters but biggest is not necessarily best. Proc R Soc Lond B Biol Sci 271: S27–S29, 2004.[CrossRef]

Rind FC and Simmons PJ. Orthopteran DCMD neuron: a reevaluation of responses to moving objects. I. Selective responses to approaching objects. J Neurophysiol 68: 1654–1666, 1992.[Abstract/Free Full Text]

Robertson RM and Johnson AG. Collision avoidance of flying locusts: steering torques and behaviour. J Exp Biol 183: 35–60, 1993.[Abstract]

Robertson RM and Pearson KG. A preparation for the intracellular analysis of neuronal activity during flight in the locust. J Comp Physiol A Sens Neural Behav Physiol 146: 311–320, 1982.[CrossRef]

Robertson RM and Pearson KG. Neural circuits in the flight system of the locust. J Neurophysiol 53: 110–128, 1985.[Abstract/Free Full Text]

Robertson RM and Reye DN. Wing movements associated with collision-avoidance manoeuvres during flight in the locust Locusta migratoria. J Exp Biol 163: 231–258, 1992.[Abstract/Free Full Text]

Roeder KD. The behaviour of free flying moths in the presence of artificial ultrasonic pulses. Anim Behav 10: 300–304, 1962.[CrossRef]

Roeder KD. Nerve Cells and Insect Behaviour. Cambridge, MA: Harvard Univ. Press, 1967.

Santer RD, Simmons PJ, and Rind FC. Gliding behaviour elicited by lateral looming stimuli in flying locusts. J Comp Physiol A Sens Neural Behav Physiol 191: 61–73, 2005.[CrossRef][ISI][Medline]

Schlotterer GR. Response of the locust descending movement detector neuron to rapidly approaching and withdrawing visual stimuli. Can J Zool 55: 1372–1376, 1977.

Simmons PJ. Connexions between a movement-detecting visual interneurone and flight motoneurones of a locust. J Exp Biol 86: 87–97, 1980.[Abstract/Free Full Text]

Smith KD and Popov GB. On birds attacking desert locust swarms in Eritrea. Entomologist 86: 3–7, 1953.

Snodgrass RE. The thoracic mechanism of a grasshopper, and its antecedents. Smithson Misc Collns 82: 1–111, 1929.

Triblehorn JD and Yager DD. Timing of praying mantis evasive responses during simulated bat attack sequences. J Exp Biol 208: 1867–1876, 2005.[Abstract/Free Full Text]

Underwood AJ. Experiments in Ecology: Their Logical Design and Interpretation Using Analysis of Variance. Cambridge, UK: Cambridge Univ. Press, 1997.

Wheatstone C. Contributions to the physiology of vision. II. Philos Trans R Soc Lond B Biol Sci 142: 1–18, 1852.

Yager DD, May ML, and Fenton MB. Ultrasound-triggered, flight-gated evasive maneuvers in the praying mantis Parasphendale agrionina I. Free flight. J Exp Biol 152: 17–39, 1990.[Abstract/Free Full Text]




This article has been cited by other articles:


Home page
 J. Neurophysiol.Home page
V. Medan, D. Oliva, and D. Tomsic
Characterization of Lobula Giant Neurons Responsive to Visual Stimuli That Elicit Escape Behaviors in the Crab Chasmagnathus
J Neurophysiol, October 1, 2007; 98(4): 2414 - 2428.
[Abstract] [Full Text] [PDF]

Home page
 J. Neurosci.Home page
H. Fotowat and F. Gabbiani
Relationship between the Phases of Sensory and Motor Activity during a Looming-Evoked Multistage Escape Behavior
J. Neurosci., September 12, 2007; 27(37): 10047 - 10059.
[Abstract] [Full Text] [PDF]

Home page
 J. Neurosci.Home page
S. M. Rogers, H. G. Krapp, M. Burrows, and T. Matheson
Compensatory Plasticity at an Identified Synapse Tunes a Visuomotor Pathway
J. Neurosci., April 25, 2007; 27(17): 4621 - 4633.
[Abstract] [Full Text] [PDF]

Home page
 J. Exp. Biol.Home page
D. Oliva, V. Medan, and D. Tomsic
Escape behavior and neuronal responses to looming stimuli in the crab Chasmagnathus granulatus (Decapoda: Grapsidae)
J. Exp. Biol., March 1, 2007; 210(5): 865 - 880.
[Abstract] [Full Text] [PDF]

Home page
 J. Neurophysiol.Home page
S. P. Peron, H. G. Krapp, and F. Gabbiani
Influence of Electrotonic Structure and Synaptic Mapping on the Receptive Field Properties of a Collision-Detecting Neuron
J Neurophysiol, January 1, 2007; 97(1): 159 - 177.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
95/6/3391    most recent