Aaron P. Cook, Jeffrey D. Triblehorn, Aaron Lorsong, Marc Lennon
Transcription
Aaron P. Cook, Jeffrey D. Triblehorn, Aaron Lorsong, Marc Lennon
#28 Praying Mantis Evasive Response: Descending Control and Context Gating. Aaron P. Cook, Jeffrey D. Triblehorn, Aaron Lorsong, Marc Lennon, and David D. Yager. Dept. of Psychology and Neuroscience Program, University of Maryland, College Park, 20742. 1 . Mantids hear ultrasound 5 . The head is necessary for the in-flight evasive response Most praying mantids have sensitive hearing via a single ear in the ventral midline of the posterior metathorax. Like vertebrates and some other insects, the ear uses tympana to detect sound. The ear’s sensitivity is highest to ultrasound at 30-120 kHz depending on species. Because the ear is in the midline, mantids are unable to determine the direction a sound is coming from. Approximately 65% of the 2000 mantis species have hearing, but there is a secondary hearing loss in many taxa with shortened wings. We used male Parasphendale agrionina for the experiments described here. The responses are similar for >20 other species that have been tested. The most direct way to test whether or not processing in the brain is necessary for the evasive response is to compare the behavior of normal and headless mantids. In some insects, e.g crickets, the head is necessary for evasion, but it is not necessary for ultrasound-triggered evasive steering in noctuid moths and defensive clicking in arctiid moths. A male decapitated during flight sometimes continued flying and sometimes stopped. Renewed flight could be triggered by tactile stimuli, and, once started, headless males tended to fly longer and more consistently than normal animals. As shown in the photos, headless males performed no obvious response to ultrasound. This was true for all sound frequencies presented at sound pressure levels up to 95 dB SPL (25-30 dB over the normal behavioral threshold). Ear location Without ultrasound Hearing mantids Deafened mantids 79.1% 30.2% Escapes 53 29 Captures 14 67 Total trials 67 96 Escapes The numbers 3 . A complex, multi-component response to ultrasound causes the evasive dives Mantids will fly readily when suspended in an anechoic chamber by a stiff wire attached to their pronotum with dental wax. A fan blows a gentle wind from the front to simulate forward motion. Speakers at the ends of the chamber can produce stimuli at any frequency with any temporal pattern. Mantids in normal flight hold all their legs folded against the body, abdomen level, and the head facing straight ahead. The wingbeat frequency in stable flight is 20-25 cycles/s, and the hindwing stroke leads the forewing stroke by approximately 30°. Within 50-100 ms after the onset of a 300 ms train of 10 ms, 35 kHz pulses, the flying, tethered mantis begins a full extension of the prothoracic legs, a strong head roll, wing beat rate and phase changes, and dorsal flexion of the abdomen. The full behavior takes 200-300 ms to develop, and it continues long after the stimulus has ended. 4 . The neural control system for evasion includes ascending and descending components ganglion connective The central nervous system of mantids consists of a ganglion in each body segment for local sensory and motor processing. Information travels between ganglia via two connectives. The brain (supraesophageal ganglion) contributes to complex behaviors. The tympanal nerve enters the metathoracic ganglion. Within that ganglion there are > six pairs of auditory interneurons, some of which send axons toward the brain. One of these, 501, has a high conduction velocity and has been implicated as a trigger neuron for initiating the evasive response. However, 501 does not appear to control the timing and coordination of the response components. to brain The ultrasound-triggered evasive response utilizes muscles in every body segment acting in a coordinated fashion. Escape behaviors require short latencies, and thoracic control of evasion without the extra time for brain processing (like a reflex) would be fastest. The complexity of the behavior, however, suggests that descending control from the brain may be required. 6 . Component latencies during the evasive behavior suggest multiple descending commands. The role of the brain in ultrasound-triggered evasive behavior could be: a) a single descending signal that sequentially activates the four components of the response b) multiple descending signals that separately control each component c) a descending signal that activates control circuitry in the thorax In the first case, the latencies of the components should increase from head roll to foreleg extension to wing beat changes to abdomen flexion. The other two possibilities predict a sequence dictated, presumably, by aerodynamic demands. 92 ms High-speed video (1000 fps) provided latency measurements for each behavioral component for males in tethered flight. The stimuli were 300 ms trains of 10 ms, 35 kHz pulses separated by 10 ms at 85-90 dB SPL, which resembles a bat in mid-phase attack and is highly effective in eliciting evasive behavior. Latencies were measured as the first observable movement of the relevant body part after the stimulus onset. Measured in this way, the latency is a combination of neural and biomechanical delays. Latency measures of wing changes is complicated by the random phase in which the stimulus occurs. We estimated the latency as the time from the stimulus to the first cycle showing a change in wing beat frequency. Extended data: (in ms.) Mean Std. dev. Min. Max. Lucifer yellow stain of interneuron 501 What are the characteristics of the control system? Total trials Animals 71.7 18.0 37.5 120.0 169 17 340.3 57.7 190.0 530.0 167 17 1191.9 387.5 8600.0 123 17 86.9 45.0 30.0 282.5 127 17 Time to Maximum 279.9 129.0 120.0 1347.5 126 17 Duration 748.3 180.8 480.0 1420.0 76 17 Latency 136.8 30.5 67.5 302.5 167 17 Time to Maximum 333.6 69.4 162.5 462.5 166 17 Duration 748.8 155.0 492.5 1675.0 124 17 Forearms Latency Time to Maximum Duration 2132.4 Head Latency tarsus What area(s) of the mantis CNS control the ultrasound-triggered evasive response? Several lines of evidence show that the absent response was not due to loss of sensory input, immediate trauma, or general debilitation. 1) Headless males lived for up to three days after surgery (22-71 hrs; mean = 33.9 ± 14.7 hrs; n=12) and during that time were able to walk, fly, and respond to tactile and wind stimuli. The apparent cause of death was dessication. 2) Headless males tested in flight 24 hours after surgery showed no evasive behavior (n=5). 3) Intact males whose compound eyes and/or ocelli were covered with black paint responded normally (n=10). 4) Immobilizing the antennae in intact males did not alter the response (n=10). 5) Socketed hairs at the neck provide proprioceptive feedback to the CNS about movements like the head roll. Intact animals with the head immobilized and/or the hairs covered continued to respond as before (n=5). 72 ms 87 ms (estimate) Brain Forelegs Wings Abdomen A single command from the brain sequentially activates the individual components of the evasive behavior. Most mantis species perform a deimatic display when threatened. Many, like Parasphendale, fully extend the forelegs. Some, like Mantis religiosa, raise and fold their forelegs to flash an eyespot. The wing and abdomen components are always present. In M. religiosa, repeated abdomen flexions produce a hissing sound. We suggest that the deimatic display was the precursor of the evasive maneuver and that they share the same neural circuits for control. Not supported Photo by Rudolf Bischoff In headless males and females, tactile stimuli can elicit the full defensive display. The forelegs are extended as shown above or are both kept near the midline. The mantis directs the display toward the side of the ‘attack.’ As with intact animals, the females display more readily and vigorously, but the response tends to be more prolonged than normal in both sexes. Thus, the defensive display does not require input from the brain for its complete execution. 9 . Context determines mantids’ response to threat and the effective stimuli for eliciting that response Substrate Hanging Hanging with tarsal contact with no tarsal contact Sound V ision Touch Sound V ision Touch Sound V ision Touch Sound V ision Touch Head on E (E) (E) 0 D D 0 0 D T T 0 Head off 0 ••••• ••••• 0 ••••• D 0 ••••• D 0 ••••• 0 1 The table above presents data from tethered mantids in flight or hanging from the tether without flying, and from mantids standing on a substrate. Tarsal contact for tethered mantids was a ball of paper held by the walking legs. Tarsal contact inhibits flight, so both cannot coincide. Three conclusions: 2) the defensive display requires tarsal contact Key: E = evasive response (E) = weak, partial evasive response D = defensive display T = non-specific twitch 0 = no response ••• = couldn not test •••1 = stopped flying when touched n = >5 for all conditions 3) sound never triggers a defensive display The sequence of foreleg joint extension varied. The bodycoxal joint most often was first. Forelegs first = 68.3% of 167 trials with 17 males Head roll first = 29% Abdomen first = <1% Head roll second = 64% Abdomen second = 35.3% The mean latency for foreleg extension appears shorter than for the more rostral head roll, but the difference is not statistically significant. This is due to the high variability, especially in the head roll, which was absent in 25% of trials. However, even if the neural latencies were identical, the lower mass of the head compared to the forelegs and the more flexible neck joint would predict an earlier head roll. Thus, the shorter foreleg latency takes on more significance. Abdomen The observations above argue that the contribution of the brain is not a single descending signal sequentially activating the components from front to back. The brain controls each component of the evasive behavior separately. The defensive display is controlled by a separate circuit. Both evasive behavior and defensive display are controlled by the same circuit. Evasion/defense can be triggered either by a signal from the brain (requires sound + flight) or from the thorax (requires tarsal contact + touch). Supported Supported and favored The defensive display is controlled by a separate circuit. 8 . Headless mantids perform the defensive display Flight Evasion/defense circuit Neck 1) evasive behavior requires flight 137 ms Hypotheses for the control of the ultrasound-triggered evasive behavior: Mantids faced with a sudden visual or tactile threat perform a defensive ‘deimatic’ display. The extended foreleg and wings make them look bigger, and the often colorful underside of the forewings provide a ‘flash’ component. Eyespots are frequently prominent on the hindwings. Females, without the options of flight or speedy retreat, display more readily and more vigorously. With ultrasound The ultrasonic stimuli may have elicited changes in wing beat rate or phase compared to intact mantids that would be difficult to observe directly. The wingbeat frequency in stable flight for headless males was lower than that for intact males (20.3 ± 1.3 Hz vs. 19.3 ± 0.9 Hz; 107 trials from 11 animals). The phase lag of the forewings was the same before and after decapitation (118.6° ± 44.6°; 80 trials from 8 animals). The graphs show the patterns of change in these parameters following a stimulus for intact and headless males, and confirm that there are clear changes in intact animals, but no response to ultrasound by headless animals. (Asterisks indicate a statistically significant difference from the pre-stimulus mean.) 1 1 . Summary resembles the ultrasound-triggered evasive behavior The defensive display has two characteristics not shared with the evasive behavior: a gaping mouth and directionality. Touch and vision—but not sound—are directional cues for mantids When a flying mantis hears the echolocation cries of a hunting bat (or bat-like 35 kHz pulses in these experiments), it performs an evasive maneuver within 150-250 ms. At lower stimulus intensities (right photo), it may be a dive and turn. At higher intensities, i.e. a closer bat (below), the mantis goes into a steep spiralling power dive that may exceed twice the normal flight velocity of 1.5-2.0 m/s. The direction of the turn or dive is random. (The strobe flash interval in the photos was 20 ms. The white arrow indicates the stimulus onset.) Is the evasive maneuver effective? We released mantids into a large flight room where a bat was hunting. Some mantids were normal, and others had been deafened by filling the ear with Vaseline. The trials were run doubleblind with 2-3 observers. The results show an advantage of almost 50% for hearing mantids. 7 . The defensive ‘deimatic’ display to diurnal predators The primary components of the defensive display are the same as the components of the airborne evasive behavior triggered by ultrasound: foreleg extension, head turn, wing changes, and abdomen dorsiflexion. 2 . Hearing helps mantids evade capture by bats ddyager@umd.edu The twitch in response to ultrasound or a flash of light was a slight, brief flexion of most of the legs. 1 0 . What CNS area(s) control the evasive response? Our results are consistent with the hypothesis that the brain controls evasive behavior when the mantis is flying and thoracic circuitry controls defensive display when the tarsi are in contact with the ground. Hearing is essential in the former case; vision or touch are needed in the latter case. The necessary CNS circuitry would be reduced, however, if evasion and defense are expressions of the same basic behavior. In that case, all of the timing and coordinating circuitry for the behavior would be a ‘module’ located in the thorax. The fundamental behavior is the diurnal defensive display triggered by vision and/or touch. The brain serves simply as an ‘AND gate’ for context and modality: if flight and ultrasound coincide, a descending signal turns on the thoracic display/evasion module. The module model makes sense in an evolutionary perspective as well. The appearance of mantids 170-150 million years ago predates the appearance of echolocating bats by at least 60 million years. During the pre-bat period, a diurnal defensive display would have served the mantids well, and aerial evasive behavior, especially triggered by sound, would not have been necessary. Bat predation provided impetus for the evolution of ultrasonic hearing and of an effective way to avoid capture by echolocating predators. Rather than evolving new escape circuitry, it would be more efficient to tap into an existing behavioral module. The only major change required would be the addition of a switching mechanism to link the new sensory modality and context to the existing defensive circuitry. According to our data, this switch resides in the brain.