Predator detection and evasion by flying insects.

Echolocating bats detect prey using ultrasonic pulses, and many nocturnally flying insects effectively detect and evade these predators through sensitive ultrasonic hearing. Many eared insects can use the intensity of the predator-generated ultrasound and the stereotyped progression of bat echolocation pulse rate to assess risk level. Effective responses can vary from gentle turns away from the threat (low risk) to sudden random flight and dives (highest risk). Recent research with eared moths shows that males will balance immediate bat predation risk against reproductive opportunity as judged by the strength and quality of conspecific pheromones present. Ultrasound exposure may, in fact, bias such decisions for up to 24 hours through plasticity in the CNS olfactory system. However, brain processing of ultrasonic stimuli to yield adaptive prey behaviors remains largely unstudied, so possible mechanisms are not known.

Behavioral responses of big brown bats to dives by praying mantises.

Insectivorous echolocating bats face a formidable array of defenses employed by their airborne prey. One such insect defense is the ultrasound-triggered dive, which is a sudden, rapid drop in altitude, sometimes all the way to the ground. Although many previous studies have investigated the dynamics of such dives and their effect on insect survival rate, there has been little work on how bats may adapt to such an insect defense employed in the middle of pursuit. In this study we investigated how big brown bats (Eptesicus fuscus) adjust their pursuit strategy when flying praying mantises (Parasphendale agrionina) execute evasive, ultrasound-triggered dives. Although the mantis dive occasionally forced the bat to completely abort its chase (25% trials), in a number of cases (75% trials) the bat followed the mantis into the dive. In such cases the bat kept its sonar beam locked onto the target and maneuvered to maintain the same time efficient strategy it adopted during level flight pursuit, though it was ultimately defeated by the dive. This study suggests that although the mantis dive can be effective in evading the bat, it does not always deter the bat from continuing pursuit and, given enough altitude, the bat can potentially capture diving prey using the same flight strategy it employs to intercept prey in level flight.

Wind generated by an attacking bat: anemometric measurements and detection by the praying mantis cercal system.

The wind-sensitive cercal system, well-known for mediating terrestrial escape responses, may also mediate insect aerial bat-avoidance responses triggered by wind generated by the approaching bat. One crucial question is whether enough time exists between detection and capture for the insect to perform a successful evasive maneuver. A previous study estimated this time to be 16 ms, based on cockroach behavioral latencies and a prediction for the detection time derived from a simulated predator moving toward a simulated prey. However, the detection time may be underestimated since both the simulated predator and prey lacked certain characteristics present in the natural situation. In the present study, actual detection times are measured by recording from wind-sensitive interneurons of a tethered praying mantis that serves as the target for a flying, attacking bat. Furthermore, using hot-wire anemometry, we describe and quantify the wind generated by an attacking bat. Anemometer measurements revealed that the velocity of the bat-generated wind consistently peaks early with a high acceleration component (an important parameter for triggering wind-mediated terrestrial responses). The physiological recordings determined that the mantis cercal system detected an approaching bat 74 ms before contact, which would provide the insect with 36 ms to perform a maneuver before capture. This should be sufficient time for the mantis to respond. Although it probably would not have time for a full response that completely evades the bat, even a partial response might alter the mantid's trajectory enough to cause the bat to mishandle the insect, allowing it to escape.

Cockroach homologs of praying mantis peripheral auditory system components.

This study identifies the cuticular metathoracic structures in earless cockroaches that are the homologs to the peripheral auditory components in their sister taxon, praying mantids, and defines the nature of the cuticular transition from earless to eared in the Dictyoptera. The single, midline ear of mantids comprises an auditory chamber with complex walls that contain the tympana and chordotonal transduction elements. The corresponding area in cockroaches, between the furcasternum and coxae, has many socketed hairs arranged in discrete fields and the Nerve 7 chordotonal organ, the homolog of the mantis tympanal organ. The Nerve 7 chordotonal organ attaches at the apex of the lateral ventropleurite (LVp), which has the same shape and general structure as an auditory chamber wall. High-speed video shows that when the coxa moves toward the midline, the LVp rotates medially to stimulate socketed hairs, and also moves like a triangular hinge giving the chordotonal organ maximal in-out stimulation. Formation of the mantis auditory chamber from the LVp and adjacent structures would involve only enlargement, a shift toward the midline, and a mild rotation. Almost all proprioceptive function would be lost, which may constitute the major cost of building and maintaining the mantis ear. Isolation from leg movement dictates the position of the mantis ear in the midline and the rigid frame, formed by the cuticular knobs, which protects the chordotonal organs.

Timing of praying mantis evasive responses during simulated bat attack sequences.

Praying mantids perform evasive maneuvers that vary with the level of danger posed by their bat predators. The vocalization pattern of attacking bats provides cues that mantids can potentially use to decide how and when to respond. Using pulse trains simulating bat attack echolocation sequences, this study determines when in the attack sequence the mantis power dive (its response to high-level threat) occurs and predicts the parameters within the echolocation sequence that are important for eliciting the response. For sequences with a rapid transition from low to high pulse repetition rates (PRRs), the evasive response occurred close to the point during the simulated sequence when the bat would have contacted the mantis. However, the evasive response occurred earlier if the transition was gradual. Regardless of the transition type, the prediction data show that sequences trigger the response when PRRs reach 20-40 pulses s(-1). These results suggest that a bat gradually increasing its PRR could 'tip off' the mantis, enabling it to escape. Attack sequences contain gradual transitions when bats engage in strobing behavior, an echolocation phenomenon that may help the bat perceive the auditory scene. Conversely, bat attack sequences that contain rapid increases in PRR close to the point of capture could circumvent the mantid's auditory defense. Based on these findings, mantids as well as other insects could benefit from having a back-up defense response to offset any advantage the bat gains by rapidly switching from low to high PRRs.

Implanted electrode recordings from a praying mantis auditory interneuron during flying bat attacks.

Using an implanted electrode, we recorded the responses from the ultrasound-sensitive mantis interneuron 501-T3 during flying bat attacks in a large flight room where the mantis served as the target. 501-T3 responds to each vocalization emitted with multi-spike bursts when pulse repetition rates (PRRs) are below 55 pulses x s(-1). As PRR increases and pulse durations fall below 3 ms, 501-T3 ceases burst activity. On average, spike bursts cease 272 ms before contact (when the bat is 73 cm away from the preparation). The timing of cessation of activity in 501-T3 is similar to the latency for the diving portion of the response of the mantid (242 ms). Experiments using vocalizing stationary bats confirm that 501-T3 responds more reliably to longer pulse durations (> or =3 ms) when intensities are below 90 dB pe SPL. The cessation of 501-T3 activity is probably due both to the increasing PRR and to the decreasing pulse duration that occur in the terminal buzz phase of a bat attack. 501-T3 may be actively shut off at high PRRs and/or intensities to protect the interneuron from habituation while the mantis performs an escape response. The cessation of 501-T3 activity is consistent with the lack of a very late ultrasound-mediated evasive response by the mantis. However, cessation of 501-T3 activity may allow a true 'last-chance' response to be mediated by other neural systems.