Nerve Cells and Insect Behavior—Studies on Crickets1 This Report on Nerve Cells and Insect Behavior Is Dedicated to the Memory

AMER. ZOOL., 30:609-627 (1990)

Nerve Cells and Insect Behavior—Studies on Crickets1

FRANZ HUBER Max-Planck-Institute fur Verhaltensphysiologie, D 8130 Seewiesen, Federal Republic of Germany

SYNOPSIS. Intraspecific acoustic communication during pair formation in crickets provides excellent material for neuroethological research. It permits analysis of a distinct behavior at its neuronal level. This top-down approach considers first the behavior in quantitative terms, then searches for its computational rules (algorithms), and finally for neuronal implementations. Downloaded from https://academic.oup.com/icb/article/30/3/609/225053 by guest on 30 September 2021 The research described involves high resolution behavioral measurements, extra- and intracellular recordings, and marking and photoinactivation of single nerve cells. The research focuses on sound production in male and phonotactic behavior in female crickets and its underlying neuronal basis. Segmental and plurisegmental organization within the nervous system are examined as well as the validity of the single identified neuron approach. Neuroethological concepts such as central pattern generation, feedback control, command neuron, and in particular, cellular correlates for sign stimuli used in conspecific song recognition and sound source localization are discussed. Crickets are ideal insects for analyzing behavioral plasticity and the contributing nerve cells. This research continues and extends the pioneering studies of the late Kenneth David Roeder on nerve cells and insect behavior by developing new techniques in behavioral and single cell analysis.

INTRODUCTION THE KIND OF APPROACH IN This report on nerve cells and insect NEUROETHOLOGY behavior is dedicated to the memory of the Neuroethology as the study of the neural late Kenneth D. Roeder, a founding father basis of behavior favors what I call the top- of neuroethology, the field of interdisci- down approach: a distinct behavioral strat- plinary research which aims to bridge the egy has to be observed and analyzed first gap between behavioral strategies and the in the field under environmental con- underlying neural substrates and mecha- straints, then quantitatively studied in the nisms. laboratory, and finally explored at its neu- Zoologists have the opportunity to select ronal and, if possible, molecular levels. This among the manifold behaviors formed by top-down approach is chosen because we evolution, and to choose those where they strongly believe that it is the behavior, think an answer is within reach when shaped and adapted by nature's abiotic and applying current technological know-how. biotic forces, which leads us to pose the Their approach is a comparative one and right questions to the nervous system. Thus, should be evolution-oriented. Zoologists neuroethologists should become familiar are interested in individual, population and with the concepts, methods, and data the species solutions, in common principles as study of behavior has to offer, as well as well as in differences. One should never with the whole scenario of modern neu- forget that a cricket differs from a frog, rosciences, including approaches at the sys- and a crayfish differs from a bird in its tem, cellular and molecular levels (Huber, demands. 1988, 1989). Neuroethologists working compara- tively have to consider two equally impor- ACOUSTIC COMMUNICATION IN CRICKETS: tant sets of questions which were formu- A FAVORABLE STRATEGY lated by T. H. Bullock (1984) (Table 1). Orthopteran and homopteran insects were among the first within the animal kingdom to have evolved hearing and 1 From the Symposium on Science as a Way of Know- sound production for intraspecific and ing—Neurobiology and Behavior organized by Edward S. Hodgson and presented at the Centennial Meeting interspecific interactions. For pair forma- of the American Society of Zoologists, 27-30 Decem- tion and reproduction, the main topics ber 1989, at Boston, Massachusetts. here, the information encoded in the send- 609 610 FRANZ HUBER

TABLE 1. Aims of comparative and evolution-orientedbe considered as a releasing stimulus for veuroethology. the subsequent one (for literature see Loher 1. What are the neural correlates and causal rela- and Dambach, 1989). tionships to known behaviors and behavioral differences among animals? 2. How to improve effective calling 2. What are the behavioral correlates and causal re- and sound radiation lationships to known neural differences among an- Mole crickets have developed tactics to imals? make calling songs more efficient. They (Statements made by T. H. Bullock, 1984.) produce them in a burrow which they mod- ify to an exponential horn which amplifies sounds of the correct carrier frequency. Downloaded from https://academic.oup.com/icb/article/30/3/609/225053 by guest on 30 September 2021 er's acoustic signals must be decoded by Flying conspecifics hear such signals already the receiver. The nervous system of the at distances of several hundred meters sender (usually the male) generates sound which guide their orientation (for litera- signals which are species-specific in their ture see Bennet-Clark, 1989). frequency spectra and their temporal orga- Male tree crickets (Oecanthus burmeisteri) nization (Fig. 1) (for literature see Bennet- improve sound intensity and radiation by Clark, 1989). The nervous system of the a baffle. They cut a hole into a leaf, into receiver (usually the female) has to fulfill which they place themselves and sing (Pro- two equally important tasks: it must be able zesky-Schulze et al., 1975). to discriminate conspecific songs from abiotic and biotic noises in order to rec- 3. Satellite behaviors ognize them (song-recognition), and it must In Gryllus integer only some males call localize the sender's position in space (song- and attract female crickets (also females of localization) (for literature see Schildber- the parasitoid flies, Euphasioptery ohracea ger et al., 1989). [Cade, 1975]), whereas other males are In insects, these distinct sender-receiver silent, surround the caller and are named interactions have evolved during the course satellites. If on her way to the singing male of phylogeny; they are formed during the female meets a satellite male, he is able ontogeny and based mainly on genetically to court and to mate with her (Cade, 1980). fixed patterns of behavior. In my report I The physiological conditions responsible will concentrate on crickets and consider for calling or noncalling are still unknown. the topics listed in Table 2. Advantages and disadvantages for callers and noncallers have been discussed, but will SOME BEHAVIORAL STRATEGIES IN not be mentioned here. CRICKETS Although crickets are best known and 4. Prey-predator strategies famous for their songs and acoustically Many cricket species are nocturnally mediated behavior, they have evolved other active. Sound traps broadcasting the con- strategies which should interest us because specific song attract flying males and they point to multisensory and multimodal females from far away (Ulagaraj and conditions demonstrating that the cricket's Walker, 1973; Walker, 1982). During their world is not solely acoustic (Huber, 1988, nocturnal flights these animals can be 1989). Here are a few examples. preyed upon by echolocating and hunting bats. Teleogryllus has evolved avoidance 1. Behaviors involved in pair formation, strategies (Moiseff et al., 1978). The ani- reproduction and aggression mals hear ultrasonic sounds and process During the reproductive season adult them in distinct neurons (Moiseffand Hoy, male and female crickets display sequences 1983) which control their turning away of distinct behavioral patterns which serve from the sound source (for literature see pair formation and mating as well as indi- Pollack and Hoy, 1989). vidual spacing, aggression and territorial Acheta domestica in southern France is defence. Each single behavioral event can known as prey for a parasitic digger wasp NERVE CELLS AND INSECT BEHAVIOR 611

Calling songs Frequency spectra IHHtlHH* Gryllus campestns 2 4 6 8 1012 1416 kHz 2 0

Teleogryllvs oceanicus Downloaded from https://academic.oup.com/icb/article/30/3/609/225053 by guest on 30 September 2021

Melanogryftus desertus 2 4 6 8 10 12 14 kHz

Oecanthus pellucens 12 4 6 6 10 12 14 kHz FIG. 1. Calling song patterns and frequency spectra in different species of crickets (modified from Huber, 1990). of the genus Liris. The flying wasp patrols mill while being kept in place by the coun- the cricket area then lands and approaches terrotating treadmill, we recently found a cricket on foot. In the case of no escape, that female Acheta domestica track optical the wasp stings and paralyzes the cricket. targets such as black squares (Atkins et al., The prey is then carried as food to the 1987). When given a choice between a black wasp's nest. But crickets have developed a square and a conspecific calling song (Fig. warning system, consisting of an arrange- 3), the female previously tracking a visual ment of filiform hair sensilla on their cerci. target switches to track the calling song, A flying and fast walking wasp creates air but only if its temporal pattern lies within currents strong enough to stimulate the the attractive range (Stout et al, 1987). filiform hair sensilla and to elicit activity During orientation to the sound source she which is then transmitted to different performs a zig-zag walking course which is ascending interneurons within the ventral characteristic for phonotaxis and expressed nerve cord. Their activity controls quick as a pattern of several steps interrupted by defensive and escape responses, such as short pauses. During orientation to the head stand, kicking with the hindlegs, and visual target she lacks that kind of walking running away (Gnatzy and Heusslein, 1986; mode (Weber et al, 1987). Thus, there for literature see Gnatzy and Hustert, seems to be a different interfacing between 1989). the visual and the acoustical recognition system and the walking generator. The shift 5. Combined visual and acoustical in walking modes indicates the change in cues for orientation the cricket's attention and in the modality Many crickets have well developed com- being processed. pound eyes and ocelli. Some Nemobius species use dark contours as landmarks to find their home territories. If they are pre- TABLE 2. TOPICS. vented from seeing these landmarks, celestical cues suffice for orientation (for A. Some behavioral strategies in crickets. literature see Honegger and Campan, B. The cricket's nervous system. C. Behavioral analysis of song recognition. 1989). D. Cellular correlates for song recognition. By using a closed loop walking compen- E. Behavioral and neuronal aspects of song localiza- sator (Fig. 2), which allows the unre- tion. strained animal to chose direction and F. Song orientation in one-eared crickets. speed of walking on the surface of a tread- 612 FRANZ HUBER

e-vector detection, receptors sensitive to IR-Camera blue light are required arranged in the dorsal rim area of the compound eyes. Ori- entation to e-vector apparently works already at illuminations as low as moon light (Labhart, 1988; Labhart et ai, 1984; Weber et ai, unpublished results). 6. Hormones and phonotactic behavior

Hormonal effects have long been Downloaded from https://academic.oup.com/icb/article/30/3/609/225053 by guest on 30 September 2021 neglected in cricket acoustic behavior. Quite recently it was reported that adult B female Acheta domestica lose phonotaxis and o c n°-.Walking direction (°) mating after removal of the corpora allata, i.e., glands that produce the juvenile hor- —- L1 mone (JH). Both behaviors are restored 180°- after reimplantation of the glands or after application of JH (Stout et al., 1976; Kou- n°- dele et al., 1987). However, subsequent work with female Gryllus bimaculatus 30 60 90 120 showed that allatectomy in the last larval Time (s) instar did not abolish phonotaxis in the adult (Fig. 5), although no JH was present in the hemolymph (Loher et al., unpub- — L2 lished results). 7. Aggression and phonotaxis Male crickets perform phonotaxis (Fig. 6A) (Weber, 1989; Weber and Hissmann, unpublished results). This strategy allows FIG. 2. Design and analysis for studying cricket unsuccessful calling males in the field to phonotaxis on the treadmill under closed-loop conditions. A. Experimental arrangement with the tread- leave their burrows and search for females mill (center), the infrared sensing and detecting device in the neighbourhood of other calling (IR camera), the electronics to control treadmill males. Moreover, males of many cricket movements (Comp. Electr.), and the broadcast of species are famous for their fights (Alex- model calling song (L, Comput.). B. Section of a tracking record of a female to loudspeaker 1 (LI) with a ander, 1961). To our surprise we found switch to loudspeaker 2 (L2). Note the zig-zag course that a male which had won a fight displayed during tracking. C. x, y plots of tracking to LI or L2 phonotaxis (Fig. 6B), whereas in the loser respectively. Each trace represents walking for 20 sec. phonotaxis disappeared for several days (A, B, adapted from Kleindienst, 1987; C, adapted from Huber, 1987.) (Fig. 6C). Thus, aggression and pair formation may be linked by a common neuronal and/or hormonal mechanism which has to be investigated. This finding invites Recently orientation to polarized light neuropharmacological studies to explore (e-vector) has been studied in the genus the physiological basis of winners and los- Gryllus (Fig. 4) (Brunner and Labhart, 1987; ers. Weber et ai, unpublished results). In com- To sum up: The world of crickets is not petition with conspecific calling songs, entirely an acoustical world. This should phonotactic orientation dominates over guide studies concentrating on multisen- e-vector orientation, and again the walking sory and multimodal information process- mode changes when switching from the ing, which is corroborated by the finding e-vector to the sound source occurs. For that many identified brain neurons encode NERVE CELLS AND INSECT BEHAVIOR 613

cm/s 1 O ] WlWWrfcrWllWl^^ 360°

optical target 1 min Downloaded from https://academic.oup.com/icb/article/30/3/609/225053 by guest on 30 September 2021

Square

FIG. 3. Visual and acoustical orientation of female Acheta domestica. A. Arrangement of the treadmill with a horizontal platform leaving an area of ca. 20 cm in diameter of the sphere open (dashed circle) for the cricket's movement. The pulsed-infrared scanning system is shown above this area. The vertical cylindrical curtain provides homogenous illumination by a ring lamp on top of the scanning system. The curtain is acoustically transparent. The direction of acoustical stimulation (L) and the visual target (square) are shown for 90° separation. B. Tracking an optical target (position indicated by arrow head on the left). C. Tracking a model calling in the attractive range (60 ms syllable period SP) with a zig-zag course. The upper traces in B and C show the compensatory sphere velocity, and the different walking modes, i.e., more steps and longer pauses during tracking the square, and a spiky walking with fewer steps and shorter pauses during tracking the calling song. The lower traces show the direction of the sphere motion caused by the female's walking. In C the horizontal line denotes the direction of the loudspeaker (modified from Weber et al., 1987). multimodal stimuli (Schildberger, 1981, and the effects of wing sensory systems on 1984a). adapting wing handedness and tooth impact (for literature see Kutsch and Huber, 1989). THE CRICKET'S NERVOUS SYSTEM In this respect one should never forget Crickets became suitable model systems that large parts of the body are employed for neuroethological research not only in a single behavioral act. When a male because of their clear cut and measurable cricket calls it not only moves the forewings acoustic behavior but also because of the periodically, but also suppresses fast walk- organization of their nervous system which ing, lifts the antennae to a position char- favors analysis of sound production, pair acteristic of calling and raises the body from formation as expressed by phonotaxis, and the ground. In addition, abdominal ven- avoidance behavior down to the single neu- tilation is synchronized with the chirp ron level (for literature see Kutsch and rhythm (Huber, 1960; Koch, 1983). These Huber, 1989; Schildberger et al., 1989; two rhythmically produced motor patterns Pollack and Hoy, 1989). are probably under the control of two sets As shown in Figure 7 the nervous system of alternatively active plurisegmental nerve is divided into discrete ganglia. Moreover, cells which feed information from the sub- each ganglion (and even nerve cells) has a esophageal ganglion down to the respec- bilateral and a mirror image arrangement tive motor generators (Otto and Campan, (Huber, 1989). This facilitates studies of 1978; Otto and Weber, 1982; Otto and segmental motoneuronal (Bentley, 1969; Amon, 1986). Hennig, 1989) and neuromuscular inter- Furthermore, cricket song is another case actions responsible for driving and con- of a growing number of rhythmic behav- trolling the forewings during stridulation. iors organized by a central pattern gen- It enabled us to analyse plurisegmental erator (for definition see Selverston, 1980), interactions, especially the influence of the which is efficiently controlled by sensory brain upon the thoracic song generator, feedback from the wings (see Kutsch and 614 FRANZ HUBER

homogeneous intact (control) light -i- sound Downloaded from https://academic.oup.com/icb/article/30/3/609/225053 by guest on 30 September 2021

11, SO d8 L2. 50 dB 4 f=T 1 =fO lU^' 50 dB 90° 180° L 1 270° L 2

allatectomized B polarized light + sound

/////////////////? M/////////////!!'/////////J FIG. 5. Phonotaxis of intact female Gryllus bimacu- ////// latus (control) and of adult females after allatectomy in the penultimate larval instar (allatectomized). For explanation see Figure 4 (courtesy of Loher et al., unpublished).

Huber, 1989) and the cerci (see Dambach, 1989).

BEHAVIORAL ANALYSIS OF SONG RECOGNITION In the past, my laboratory has concen- trated mainly on high-resolution behavioral experiments developed to elucidate the acoustical constraints of female phonotaxis, a behavior, which expresses both song recognition and localization of the caller /////////f/l///!_ (for literature see Weber and Thorson, lm 1989). That behavior was combined with ///////////////Fiirrp a search for single nerve cell correlates and causal relationships (for literature see FIG////////////////. 4. Relativ7////777Te frequenc7y of tracking angles of female Gryllus bimaculatus to model calling songs and Schildberger et al, 1989). homogenous light (A), to model calling songs and polarized light of different e-vector orientations (B), and to polarized light of different e-vector orientations alone (C). The position of the sound source is nor polarized light at different e-vector orientations marked by LI and L2. The experiments were carried abolished phonotaxis (A, B). In C orientation to polar- out with changing loudspeaker positions from LI to ized light is demonstrated by changing the e-vector L2 (A, B), and by increasing sound intensit) (from 50 in steps of 45°. 1 m gi\ es a calibration for the tracking to 80 dB SPL). Note that neither homogenous light angles (courtesy of Weber et al., unpublished). NERVE CELLS AND INSECT BEHAVIOR 615 Downloaded from https://academic.oup.com/icb/article/30/3/609/225053 by guest on 30 September 2021

/////////////////////////////////// //////////'(/)itiff/fffffffflfffffiff //////////////////////////////////// 1 m /////////////

loser, 20 mm after combat FIG. 7. Aspects of the cricket's nervous system. A. FIG. 6. Male Gryllus campestris phonotaxis (A) per- CNS of Gryllus campestris with the ganglia, connectives sists in the winner of the combat (B) but is lost in loser and lateral nerves. B, brain; SEG, subesophageal gan- (C). For explanation of the details compare Figures glion; Tl-3, thoracic ganglia; A3-7, free abdominal 2 and 3 (A) and Figure 4 (A, B). (Courtesy of Weber, ganglia. In crickets Al and A2 are fused with T3. B. 1989; Weber et al, unpublished.) Prothoracic ganglion with the bilateral arrangement of the auditory nerve bundles (dotted areas) within the leg nerve (LN) and the left and right auditory neuropiles (LAN, RAN). C. Structure of the mirror image Omega cells (ONI L—black) (ONI R— 1. Phonotaxis in the field and in a hatched). Arrows indicate the excitatory auditory input closed-loop arrangement to the left and right cell respectively. D. Scheme of a transversal section through the mesothoracic gan- Crickets perform phonotaxis by means glion to demonstrate the bilateral arrangement of of walking or flying in the field (Klopf- main neuropile areas (hatched vertically and horizon- fleisch, 1973; Walker, 1982). In order to tally) and several of the mirror image fiber tracts evaluate the important calling song param- (hatched densely and oblique). DIT, dorsal interme- dia! tract; DLT, dorsal lateral tract; DMT, dorsal eters for phonotaxis, a closed-loop setup medial tract; LVT, lateral ventral tract; VIT, ventral was developed, as already described (Fig. intermedial tract; VT, ventral tract. (Modified from 2). Using this experimental tool we found Huber, 1990.) a threshold for phonotaxis to model calling songs usually around 50 dB SPL. By changing loudspeaker positions the female changed her orientation often within sec- alous phonotaxis. The female tracked the onds and she performed a zig-zag course sound source with an erroneous angle and while tracking the sound source by mean- this angle was carrier frequency dependent dering ca. 40° around the loudspeaker (Thorson et al., 1982). She behaved as if direction (Wendler et al., 1980; Weber et the sound source had changed in space. al., 1981). This phenomenon will be treated The mechanism for anomalous phonotaxis later. is not completely understood. However, anomalous phonotaxis clearly demon- 2. Carrier frequency and walking angle strates that songs with wrong carrier fre- Calling songs of the correct temporal quencies but correct patterns do not abol- pattern but with a wrong and higher than ish recognition but influence sound natural carrier frequency did elicit anom- localization. 616 FRANZ HUBER

song recognition, as discussed in the next section.

CELLULAR CORRELATES FOR SONG RECOGNITION 1. Functional properties of cricket ears and prothoracic auditory interneurons The auditory pathway in crickets begins S P [ms] 23 28 35 43 53 67 81 100

SRR [Hz] 43 36 29 23 19 15 12 10 with the ears (tympanal organs) located Downloaded from https://academic.oup.com/icb/article/30/3/609/225053 by guest on 30 September 2021 within the proximal parts of the foretibiae. FIG. 8. Behavioral tuning (bandpass-property) to a specific range of syllable repetition rates (SRR) during Each auditory organ consists of about 50- phonotaxis of female Cryllus campestris presented in 60 auditory sense cells arranged in rows a "to and fro" sequence. The dotted areas indicate and attached to the upper wall of the inner the degree of variation in the response of all females trachea (Eibl, 1978). This arrangement of tested with a preference near 30 Hz, and the black dots denote the extreme values. (Modified from auditory sense cells reflects tonotopicity, Thorson etai, 1982.) i.e., auditory receptors according to their location are tuned to different sound frequencies: proximal receptors respond 3. Bandpass-property for song recognition preferably to lower, distal receptors to By changing one of several temporal higher frequencies (Oldfield et al, 1986). parameters of the calling song, we found Thus, the ear analyzes frequencies, an abil- one parameter especially important for ity required, for instance, to encode calling song recognition: the syllable repetition rate and courtship songs having different car- (SRR) (Fig. 8) (Thorson et al., 1982), rier frequencies as well as ultrasonic sounds whereas even considerable changes in other (for literature see Bennet-Clark, 1989; Pol- parameters were much less critical. This lack and Hoy, 1989). indicates that at least Gryllus campestris and Sound intensity is encoded in the spike Gryllus bimaculatus have developed a win- frequency of the auditory sense cell. More- dow for attractive SRRs in phonotaxis, a over, auditory receptors at their best fre- bandpass, ranging from about 19-43 Hz, quency copy the temporal structure of the and peaking around 25-35 Hz. The prob- song, but, with an important restriction: lem of a trade-off strategy in song pattern they are not specifically tuned to the timing recognition, i.e., the evaluation and of the conspecific pattern (Esch et al., 1980). weighting of several parameters, is still This indicates that the ears of crickets cover unsolved and will not be discussed here a much wider range of sound frequencies (Stouts al., 1983, Doherty, 19856, c). and copy a broader range of patterns than used for intraspecific communication. From 4. Temperature: Sound production and a biological point of view, this is not at all phonotaxis surprising because the ears have also Crickets are poikilothermic animals. evolved as sensory devices for predator They have to match acoustic communica- avoidance where different sound frequen- tion patterns with changes in ambient tem- cies and temporal patterns are used (Huber, peratures. "Hot" males produce faster 1989). SRRs than "colder" males, and equally Auditory nerve fibers project to the pro- acclimatized females are tuned to them, thoracic ganglion and terminate within a i.e., they shift their bandpass, respectively part of the ring tract, an area called the (Doherty, 1985a). acoustic neuropile. Each ear is represented To sum up: Our finding that the SRR is by fiber terminals only in the correspond- the most important recognition parameter ing hemiganglion (for literature see Schild- encouraged us to search for neuronal cor- berger et al., 1989). Within the prothoracic relates and to propose a mechanism for ganglion auditory information is transmit- NERVE CELLS AND INSECT BEHAVIOR 617 ted to a family of neurons which exist as mirror image pairs, and some of them have been identified by intracellular recording and staining in several genera of crickets. We can distinguish local prothoracic interneurons such as the Omega neurons (ON) with arborizations restricted to both halves of the ganglion, ascending (AN) and descending (DN) plurisegmental neurons, projecting to the brain or to lower parts of Downloaded from https://academic.oup.com/icb/article/30/3/609/225053 by guest on 30 September 2021 the ventral nerve cord, and neurons with T-shaped structures (TN). Only for Teleogryllus commodus has monosynaptic transmission between auditory afferents and AN1 and AN2 neurons been substantiated (Fig. 9) (Hennig, 1988). But there is no indication of specific tuning in any of these prothoracic interneurons to SRRs necessary for the female to exhibit phonotaxis. This led us to search for cellular correlates of song recognition in the next station of the auditory pathway, the brain (Schildberger, 19846).

2. Song recognition by local brain neurons FIG. 9. Indications for monosynaptic connections in Teleogryllus oceankus between auditory afferent fibers Based on behavioral studies, conspecific (HNF) and two ascending auditory interneurons (AN 1, song recognition requires neurons sensi- AN2) located within the prothoracic ganglion, stud- tive to phonotactically effective SRRs. ied with electrical stimulation of the auditory nerve. A. Prothoracic ganglion with the structure of AN1 Auditory information is conducted to the (left) and AN2 (right). Arrows indicate the position brain via ascending interneurons (AN1, of the intracellular electrodes (for the auditory affer- AN2 and probably others) and processed ent fibers only shown left). B. Superimposed traces of there by at least two classes of local brain spike potentials at expanded scale to demonstrate the time relationships between afferent auditory fiber neurons (BNC1, BNC2). According to ana- spikes and postsynaptic responses in AN1 and AN2. tomical arrangements and latency mea- Top to bottom: HNF, auditory afferent fiber; AN1, surements, Schildberger (19846) proposed received excitatory input mediated by a low frequency a cascade of events in each brain hemi- (4 kHz) receptor fiber; AN2, received excitatory input from a high frequency receptor fiber. In each trace sphere: Ascending neurons feed song the initial deflection indicates the stimulus artefact. information into members of BNC 1 cells, AN1 and AN2 exhibit short and constant latencies and they and BNC2 cells are targets for to the stimulus of 3.5 ms, and latencies to the onset further information processing. Within the of the receptor spike of ca. 1 ms (AN1) and 0.6 ms classes of BNC 1 and BNC2 cells three func- (AN2). (Modified from Henning, 1988.) tional types were discovered (Fig. 10A), (i) Neurons acting as highpass filters (HP-F) by responding to faster SRRs and (ii) neu- taxis. But these cells showed no pattern rons acting as lowpass filters (LP-F) by copying of the syllable rhythm and they responding to slower SRRs, both covering lost encoding of sound intensity at mod- a range inside and outside the phonotactic erate and higher SPLs. attractiveness, (iii) Within the class of BNC2 Schildberger (see Huber and Thorson, cells, a subclass was identified with a band- 1985) proposed a model (Fig. 10B) about pass-property (BP-F), i.e., cells that a cellular and network mechanism for song responded only to those SRRs to which the recognition in the brain, based on the AND- female on the treadmill exhibited phono- gate property. However, the remaining gap 618 FRANZ HUBER

75- 75 _ 3?

O x as O 0) c

CD o Downloaded from https://academic.oup.com/icb/article/30/3/609/225053 by guest on 30 September 2021

Q. CO 25 25

26 50 74 98 Syllable interval [ms]

B Thorax Brain

BP-F Chirp Auditory AND- • ••• Pathway Gate -Phonotaxis SI LP-F

FIG. 10. Neuronal correlates for song recognition in the brain of Gryllus bimaculatus. A. Correlation of the phonotactic response (hatched area indicating a band-pass property, peaking around 30 Hz SSR) with the activity of bandpass neurons BP-F (marked by open circles for different females and by closed circles connected by lines in one female). Other local brain neurons of the same classes exhibit either highpass properties (HP- F, triangles) or lowpass properties (LP-F, squares) responding preferably to faster or slower syllable repetition rates, respectively. Note that HP-F and LP-F form the boundaries of BP-F. B. Model to explain how the bandpass property for attractive SRRs could arise from AND-gating highpass and lowpass neurons. (A, modified from Schildberger, 19846; B, modified from Huber and Thorson, 1985.)

involves our ignorance of the detailed neu- 1981). It allows sound to travel to the tym- ronal implementation, especially with panum from outside and via the acoustic respect to synaptic mechanisms and con- trachea from inside. Thus, vibrations of nectivities among the cells. the tympanum, necessary for hearing (Kleindienst etal, 1983), are based on pres- BEHAVIORAL AND NEURONAL ASPECTS sure and phase differences of the sound OF SONG LOCALIZATION waves impinging on the tympana. 1. Cricket ears as pressure gradient Pressure gradient receivers have cardoid receivers directional characteristics, i.e., the sense The cricket ear is a pressure gradient cells are excited with different strengths receiver (Fig. 11 A) (for literature see Lar- depending on the angle of sound incidence sen et at, 1989). The tracheal system to (Fig. 11B) (Boyd and Lewis, 1983). Both which the tympana are connected acts as ears exhibit nearly mirror image direc- an internal sound conducting pathway tional characteristics with a frontal and (Kleindienst, 1980, 1987; KleindienstWaZ., caudal intersection point. During phono- NERVE CELLS AND INSECT BEHAVIOR 619

cavity 2

step stimulus attenuator - generator Downloaded from https://academic.oup.com/icb/article/30/3/609/225053 by guest on 30 September 2021

inhibitory stimulus (dB) 65 75 85 95

FIG. 11. Cricket ears as pressure gradient receivers. A. Prothoracic segment opened and seen from a fron- E 4 tal view with the two ear bearing forelegs. ATY, anterior tympanum; PTY, posterior tympanum; AT, acoustic trachea connecting both ears; PTG, prothoracic ganglion; SR, SL, right and left stigma (lat- 1 ' ' r—constant stimulus 65 dB *\ ' 0 eral opening for sound entrance). B. Cardoid and 45 55 65 excitatory stimulus (dB) mirror image directional characteristics of Gryllus FIG. 12. Quantitative analysis of contralateral inhi- campestris ears, obtained by recordings from whole bition in the Omega cell type 1 in response to 5 kHz auditory nerves of left (L) and right (R) ear. Polar sound signals. A. Closed sound field arrangement (leg plots show evoked responses to single sound pulses of phones = miniature sound chambers) for external and 20 ms duration and 5 ms rise/fall time, averaged over internal isolation of excitatory and inhibitory inputs 128 presentations and delivered at constant sound to the ONI (fi). Ml, M2 microphones 1 and 2 acting intensities (70 dB SPL) and of 4.8 kHz from different as miniature loudspeakers. B. Response characteristic angles. The two directional curves cross frontal and caudal, and exhibit greatest left-right differences to (•—•) and latency (O O) of the Omega cell for lateral stimulation (adapted and modified from Boyd various excitatory and inhibitory stimulus settings. and Lewis, 1983). C. Diagram to explain binaural Symbols represent means of 30 consecutive sound directional hearing based on the algorithm "turn to presentations with standard deviations. Sound pulse the side more strongly stimulated." According to the duration: 50 ms, rise and fall times: 2 ms (adapted directional responses of the left and the right ear their from Kleindienst el al., 1981). information is fed into left and right central prothoracic ascending neurons (NL and NR), respectively. Their activity is compared by a central comparator 2. Processing of monaural and binaural (possibly located within the brain) which evaluates auditory input by prothoracic left/right excitation differences (AIR/L) for correcting the course, indicated by zig-zagging of the female interneurons, and network analysis during phonotaxis (adapted from Huber, 1987). With the invention of the legphones (Fig. 12A) {i.e., miniature sound chambers around the ear) (Kleindienst et al., 1981), each ear could be stimulated separately tactic tracking crickets follow the algo- after severing the acoustic trachea con- rithm "turn to the side more strongly stimu- necting both ears while recording simul- lated." Their zig-zag course allows them to taneously from different types of protho- pursue the frontal crossover point where racic interneurons. Thus, binaural and the left and right intensity and excitation monaural inputs to these neurons and some differences are minimized (Fig. 11C) (for network properties could be studied (Woh- literature see Huber, 1987). lersand Huber, 1982). 620 FRANZ HUBER

pattern copying in these cells (for literature seeSchildberger^a/., 1989; Huber, 1989). 3. A direct approach to song localization by hyperpolarizing single ascending interneurons Network studies of prothoracic neurons and effects of cell killing (see also Atkins et al., 1984) led to the assumption that some of these neurons are involved in song localization. To test this hypothesis, an open- Downloaded from https://academic.oup.com/icb/article/30/3/609/225053 by guest on 30 September 2021 loop arrangement was used to study phonotactic behavior and single cell

60 0 100 200 300 0 100 200 300 responses simultaneously in female Gryllus Time (s) Tims (ms) Time (mi) bimaculatus (Fig. 13) (Schildberger and FIG. 13. Correlation and causal relationship between Horner, 1988). phonotactic course and the activity of an ascending To obtain a causal relationship between auditory interneuron in the cricket, Gryllus bimaculatus. A. Open-loop experimental arrangement for the phonotactic course and neuronal activ- intracellular recordings from walking animals. The ities one needs reversible manipulation of animal is fixed on a holder and can only walk straight single neurons during the behavioral per- forward, but the legs can turn an airsupported sty- formance which cell killing does not offer. rofoam ball. A special camera (IR) senses two com- The female cricket was mounted as shown ponents of the ball's (animal's) movement—rotation and translation—and thus its turning tendency. Call- in Figure 13A; it could only walk straight ing song is delivered by one of two loudspeakers (L, forward but turned an airsuspended ball R) 50° on either side of the longitudinal axis of the with its legs, and ball rotation was mea- restrained animal. The computer controls model call- sured. One loudspeaker positioned 50° to ing song broadcast, and later evaluates tape recorded the left or to the right in azimuth to the movements of the animal and neuronal activities. B. Graph with rotation of the animal by sound delivered body axis of the female broadcast the call- from the left speaker (a), from the right speaker (b), ing song while either AN1, ONI, or AN2 and to the right (c) after the left AN1 neuron had neurons were recorded intracellularly and been hyperpolarized, despite sound stimulation from later on manipulated electrically. With a the left. C. Outlines of the prothoracic ganglion with the mirror image AN1 cells (LAN], dendritic field model calling song presented via the left and axon on the left and RANI, dendritic field and loudspeaker, the left AN1 neuron (with the axon on the right). In both neurons the axon courses dendritic field on the left side) (Fig. 13C), to the brain. D. Response patterns of AN1 cells to a was more strongly excited and copied the four syllabic calling chirp (bottom trace). Further pattern (a, in Fig. 13D). The animal fol- explanations are given in the text (adapted from Huber, 1989). lowed the algorithm "turn to side more strongly excited" and turned to the left sound source (a, in Fig. 13B). By consid- Between the mirror image ONI type ering the directional characteristic of the neurons reciprocal inhibitory connections right ear, under this condition the right were discovered which serve to increase AN1 neuron must have been less excited the binaural contrast (Fig. 12B). This was (a', in Fig. 13D). When the calling song was broadcast via the right loudspeaker, the further established by selective cell killing animal turned to the right (b, in Fig. 13B). with photoinactivation (Selverston et al., While still recording from the left AN1 1985), which removed inhibition from one neuron it was now less excited than before ON 1 neuron upon the other. When testing (b, in Fig. 13D). On the other hand, the the animal after killing of one ON 1 neuron mirror-image partner cell on the contra- on the treadmill, it tracked the sound lateral side was now more strongly excited source with a slightly erroneous angle. (b\ in Fig. 13D). Morever, ONI-neurons inhibit contralateral AN1 and AN2 neurons, which may The crucial experiment for establishing assist to enhance directionality as well as the hypothesis of binaural comparison was NERVE CELLS AND INSECT BEHAVIOR 621 to hyperpolarize the left AN 1, i.e., to reduce and the activation levels of these two neu- its activity even below the level of its right rons corresponded only if pattern copying mirror image cell (compare c with c' in Fig. of the neurons was considered. With call- 13D) and to broadcast the sound from the ing song from above and continuous tone left, its excitatory side. An animal with a from ipsilateral to the neurons recorded, nonhyperpolarized left AN 1 cell turned to their pattern copying was masked, whereas the left (a, in Fig. 13B). As predicted by in the contralateral neurons pattern copy- the rule, the animal changed its course to ing was still maintained. the right when the neuron was hyperpo- These results demonstrate that the over- larized (c, in Fig. 13B). all activity of the two mirror image ascend- Downloaded from https://academic.oup.com/icb/article/30/3/609/225053 by guest on 30 September 2021 This indicates that the system responsi- ing neurons does not directly guide ori- ble for orientation must have received entation. It first has to pass a filter tuned "wrong directional information" due to the to the temporal structure of syllables and reduced activity of this single hyperpolar- chirps. Thus, recognition of the conspe- ized cell. The effects of hyperpolarizing cific song and localization of the sound single AN2 and ONI neurons caused less source are not independent events. dramatic changes in walking courses (Schildberger and Horner, 1988). SONG ORIENTATION IN 3. Pattern dependence of sound ONE-EARED CRICKETS localization 1. Walking courses in one-eared crickets When Pollack et al. (1984) studied sound The algorithm based on binaural com- frequency effects during tethered flight and parison predicts that animals with only one compensated walking in Teleogryllus oceani- ear ought to circle to the side of the cus, they found a larger shift in angular remaining ear. However, Huber et al. error of tracking at the same high fre- (1984), Schmitz et al. (1988) and Schmitz quency (15 kHz) after the sound pattern (1989) found that more than 30% of mon- had changed from a four-pulsed calling aural females of Gryllus campestris and G. song to a sound containing a single pulse bimaculatus exhibited phonotaxis even with of the chirp rhythm (2/sec). one ear. These females recognized the con- Stabel (1988) and Stabel et al. (1989) ana- specific song by input from one ear, and, lyzed phonotaxis of female Gryllus bimacu- even more surprising, they succeeded in latus with an open-loop device. They tracking the sound source. recorded the turning tendency of a teth- In adult female crickets, one foreleg ered female Gryllus bimaculatus on paired bearing an ear in the tibia was amputated tread wheels in a complex acoustic stimulus between coxa and femur. When the ampu- paradigm. When calling song at the correct tation occurred before the 6th instar (4-5 carrier frequency was broadcast horizon- instars before imaginal molt) the leg regen- tally the female turned to its side as erated often to its full length and size, but expected. However, as soon as the same without an auditory organ. Several exter- calling pattern was broadcast from above nal sensory systems reappeared and leg {i.e., without directional cues), and a con- muscle innervation was completed, as indi- tinuous tone of a slightly different carrier cated by the coordinated walking of the frequency was presented horizontally, the regenerated leg (Huber, 1987; Schildber- female changed its direction and turned ger and Huber, 1988). away from the horizontal sound source. When a calling song was broadcast from This sign reversal in turning was then stud- a horizontal direction, some females kept ied at the level of two ascending auditory course, although with an error angle usu- interneurons, AN1 and AN2, recorded ally below 90°, which in nature would guide extracellularly from the cervical connec- them through an arc to the singing male tives while the animal walked. Both the (Fig. 14A, C). Phonotactic tracking was characteristic curves in turning tendency most commonly elicited in a smaller sound (toward the sound source or away from it) intensity range, preferably at lower inten- 622 FRANZ HUBER

with the central neurons of the deprived side. Or, crickets can switch from the mechanism of binaural comparison to one with consecutive measurements of monaural input, that is, to a scanning mechanism, since auditory excitation varies according to the directional characteristics of the remaining ear. Here I only discuss results that favor the first alternative; for the second, refer to Schildberger and Kleindienst (1989). Downloaded from https://academic.oup.com/icb/article/30/3/609/225053 by guest on 30 September 2021 2. Bilateral central comparison due to -1.2 changes in structure and function of central prothoracic auditory neurons In adults which had lost one foreleg in a larval instar and regenerated it, or in adults with enough time elapsed between FIG. 14. Phonotactic tracking in one-eared Gryllus amputation, bilateral comparison could bimaculatus females to model calling song. A. x, y plots result from two different mechanisms: (i) of sound source dependent walking courses of an intact primary auditory fibers, known to termi- female. C. Change in walking angle of the same female after amputation of the right foreleg. Note the devia- nate preferably within the ipsilateral audi- tion from the sound source by about 70° toward the tory neuropile, could grow processes across left ear. B. Precise phonotactic tracking of an adult the ganglionic midline to meet neurons female which lost the right foreleg in an earlier larval with dendritic fields in the contralateral instar and regenerated it to its normal length, but hemiganglion. Many cobalt backfills of without a functional auditory organ. D. No change in walking courses in the same female after amputa- auditory nerves in monaural crickets tion of the previously regenerated right foreleg. L 1, revealed that only very few primary audi- 2, loudspeakers 1 and 2, respectively, positioned 135° tory fibers crossed the ganglion midline, apart. Sound intensity was 70 dB SPL. Each trace and perhaps not more than already seen represents a 20 sec walking (adapted from Huber, 1987). in binaural animals (Schmitz, 1989). It is still unknown whether these fibers carry auditory information from the remaining sides, and it could be observed in some ear to the contralateral central neurons adults already within 24 hr after ear loss. deprived from their previous auditory Course accuracy in some adults improved inputs. when the foreleg had been amputated in A second alternative is that central audi- earlier larval instars (Fig. 14B, D), or in tory neurons, the target cells of primary adults, after a week or two had elapsed auditory fibers, change their morphology between amputation and test (Schmitz et and function after monaural deprivation. ai, 1988; Schmitz, 1989). Some females This was found in several of the identified tracked the sound source as accurately as local and ascending interneurons in differ- binaural animals. This led to the question, ent cricket species (for literature see what kind of localization mechanism is Schildberger et ah, 1989). The previously involved in monoaural crickets. deafferented neurons grew dendrites from There are, in principle, two mechanisms their former input area which crossed the which could account for tracking the sound ganglion midline to invade the auditory source with only one ear. Either the cricket neuropile of the intact side (cf, Fig. 15 A, reorganizes the auditory pathway so that B). These cross-grown dendrites made the remaining ear drives a central bilateral functional connections, very probably with system which would allow central compar- primary auditory fiber terminations, which ison. Such a mechanism would imply that were manifested by their responses to stim- the existing ear makes new connections ulation of the remaining intact ear. Thus, NERVE CELLS AND INSECT BEHAVIOR 623

they were rewired with the "wrong ear." According to their threshold curves (Fig. 15C) and intensity-response functions (Fig. 15D), the rewiring created functional properties very similar to intact cells, including the cell specific pattern copying (inset in Fig. 15C). Thus, one can state that the cross-grown dendrites must have "found" terminals of auditory sense cells

of the wrong ear comparable to those with Downloaded from https://academic.oup.com/icb/article/30/3/609/225053 by guest on 30 September 2021 which these neurons were previously connected on their intact side, and even the synaptic organization must have been restored. This plasticity was unexpected in crickets and calls for future research in FIG. 15. Structural and functional changes in pro- developmental and postembryonic neuro- thoracic auditory neurons (Omega cells of type 1) in biology. one-eared Gryllus bimaculatus, after the loss of the left auditory input in an earlier larval instar. A. Lucifer 3. Constraints for phonotaxis with one ear yellow fill of a left Omega neuron in the intact animal (horizontal view). B. Left Omega neuron with den- and experimental proofs drites grown across the ganglion midline from the Despite this time dependent and unex- former input area and arborizations within the con- pected reorganization within the protho- tralateral neuropile (which is normally the output area of the cell). C. Comparison of auditory thresholds in racic auditory pathway, the mechanism normal and deafferented ON 1 cells shows only slight underlying monaural tracking is not yet differences within the tested frequency range. Inset: explained. If localization in monaural ani- Pattern copying of the deprived ONI cell now wired mals is the result of a central bilateral com- with the wrong, intact ear. D. Intensity-response parison, then one should propose different curves of intact and deprived ONI cells listed in C. The slopes are very similar except for a decrease in thresholds, based on synaptic efficacies response in the deprived ONI at higher stimulus and/or combinations of excitation and intensities (adapted from Huber, 1989). inhibition. Furthermore different slopes of the intensity-response functions should be expected, resulting in different directional in the corresponding neuronal pair cor- characteristics between neurons connected related within this intensity range. Non- to the remaining ear, and those rewired to orienting females lacked such an intersec- the wrong ear. For tracking, monaural ani- tion point (cf Fig. 16B, lower left with lower mals need an intersection point of the inten- right). sity-response functions of a neuronal pair Thus, it seems that the structural reor- (similar to the frontal intersection point of ganization of the central auditory pathway binaural animals), which would enable them in monaural crickets is accompanied by to set an equilibrium between the excita- changes in some functional properties of tion of both auditory pathways necessary the participating neurons. Both together to perform phonotaxis, by following the allow a bilateral central comparison used rule similar to that used by intact binaural for tracking. However, one should not for- crickets (Fig. 16A). get that only some females tracked the Such intersection points have recently sound source within a restricted intensity been discovered in the intensity-response range while others, with probably a similar functions of the pair of AN2 neurons by structural reorganisation but a missing Schildberger and Kleindienst (1989) (Fig. functional repair did not (Schmitz et ai, 16B). Only those females which exhibited 1988). phonotaxis on the treadmill (closed-loop) This and the fact that some females or showed a reverse in turning tendency tracked the sound source already several in the open-loop condition at a distinct hours to one day after loss of one ear, where sound intensity, had an intersection point the described morphological changes have 624 FRANZ HUBER

not been observed (Schmitz, 1989), pose questions regarding normal and regener- ative development, functional plasticity in single neurons, and they even point strongly to a second sound orientation mechanism which has recently been found: a monaural scanning device (Schildberger and Klein- dienst, 1989).

CONCLUDING REMARKS Downloaded from https://academic.oup.com/icb/article/30/3/609/225053 by guest on 30 September 2021 The top-down approach applied to song recognition and song localization in crickets was successful because results of high- resolution studies of phonotactic behavior guided the questions addressed to single nerve cells. We first learned that the world of crickets is not solely an acoustic world which invites studies of multisensory and multi- FIG. 16. Evidences for a central bilateral comparison modal inputs and their processing within in one-eared crickets with loss of the right ear in an the CNS in the context of behavior. earlier larval instar and after subsequent regeneration of the right foreleg without its ear. A. Left, diagram We have identified a cellular correlate of the situation: The intact left ear did not change its in the brain with a bandpass-property for directional characteristic. Its input reaches both left SRRs similar to the one obligatory for (NL) and right (NR) prothoracic ascending neurons. phonotaxis in the behaving animal. With- They conduct song specific directional information to a central comparator (Com), which evaluates left/ out, however, knowing the neuronal hard- right activity differences (AIL/R) and minimizes them ware in detail, we even present a mecha- to establish the course. A. Middle, hypothetical and nism which could be responsible. Students rather parallel intensity—response functions of a neu- of animal behavior would perhaps call the ronal pair with the NR newly connected. Turning bandpass-cells in the brain an element of tendencies are shown, if the animal follows the algorithm "turn to the side more strongly stimulated." the innate releasing mechanism (IRM). A. Right, hypothetical intensity-response functions For song localization the binaural algo- of the same neuronal pair with different slopes now rithm "turn to the side most strongly stim- intersecting. The intersection point would be a stable point for tracking. Turning tendencies are indicated ulated" could be established at the level of below and above the intersection point. B. Experi- a single ascending neuronal pair. However, mental proof for A. Upper left, examples are shown the comparison and transfer of directional for one animal (Animal A) which tracked the sound information (possibly within the brain) is source under closed loop conditions at 60 dB SPL, still a matter of speculation. and for a second animal (Animal B) with the same treatment but without phonotaxis. Upper right, Ani- Other studies have shown that localiza- mal B followed the algorithm "turn to the side more tion of a calling song source needs pat- strongly stimulated" by consistently turning to the terned information from the ventral nerve left under open-loop conditions, whereas Animal A cord, which clearly suggests that recogni- showed a reversal in turning in the range between 60 to 70 dB SPL. At low sound intensities it turned to the side of the deprived ear; at higher sound intensities it changed to the side of the intact left ear. Lower left, intensity—response curves of the mirror image AN2 neurons recorded extracellularly in Ani- curves (IR-function) in the two AN2 cells, within the mal B with no phonotaxis indicate that the intact and corresponding sound intensity range (compare upper normally wired left A\2-cell is more sensitive to sound right with lower right). At lower intensities the within the whole intensity range tested. Thus the left deprived right AN2 cell is more strongly excited and AX2 seems to determine turning toward its side. determines turning to its side: at higher intensities Lower right, in Animal A, howeser, which exhibits the intact AN2 cell is more strongly excited which phonotaxis under closed-loop conditions and shows a results in turning to the left as expected by following change in turning under open-loop conditions, an the algorithm (modified and adapted from Schild- intersection point is seen in the intensity-response berger and Kleindienst, 1989). NERVE CELLS AND INSECT BEHAVIOR 625 tion and localization are not based on two Cade, W. H. 1980. Alternative male reproductive completely independent neuronal circuits. behaviors. Fla. Entomol. 63:30-45. Finally, one-eared crickets with their Dambach, M. 1989. Vibrational responses. In F. Huber, T. E. Moore, and W. Loher (eds.), Cricket ability to track a sound source have eluci- behavior and neurobiology, pp. 178—197. Cornell dated an unexpected plasticity in the University Press, Ithaca, New York. behavior related to structural and func- Doherty, J. A. 1985a. Temperature coupling and tional changes of central neurons. These trade-off phenomena in the acoustic communi- findings will certainly facilitate further cation system of the cricket, Gryllus bimaculatus DeGeer (Gryllidae). J. Exp. Biol. 114:17-35. studies on the development and the regen- Doherty, J. A. 19856. Trade-off phenomena in call- eration power of the auditory pathway, ing song recognition and phonotaxis in the cricket, Downloaded from https://academic.oup.com/icb/article/30/3/609/225053 by guest on 30 September 2021 including sensory deprivation or overstim- Gryllus bimaculatus (Orthoptera, Gryllidae). J. ulation in single cells. Thus, the nervous Comp. Physiol. A 156:787-801. system of the cricket is not completely Doherty,}. A. 1985c Phonotaxis in the cricket, Gryl- lus bimaculatus DeGeer: Comparisons of choice hardwired. Plasticity within nerve cells and no-choice paradigms. J. Comp. Physiol. A allows adjustment of synaptic efficacies and 157:279-289. formation of new connections with chang- Eibl, E. 1978. Morphology of the sense organs in the ing conditions within the body and in the proximal parts of the tibiae of Gryllus campestris environment. L., and Gryllus bimaculatus DeGeer (Insecta, Ensi- fera). Zoomorphologie 89:185-205. Esch, H., F. Huber, and D. W. Wohlers. 1980. Pri- ACKNOWLEDGMENTS mary auditory neurons in crickets: Physiology and central projections. J. Comp. Physiol. 137:27-38. I am most grateful to my coworkers and Gnatzy, W. and R. Heusslein. 1986. Digger wasp guests for their motivation, experimental against crickets. Naturwissenschaften 73:212- skills and ongoing interest in the work 215. which is presented here. Gnatzy, W. and R. Hustert. 1989. Mechanoreceptors in behavior. In F. Huber, T. E. Moore, and W. Loher (eds.), Cricket behavior and neurobiology, pp. REFERENCES 198-226. Cornell University Press, Ithaca, New Alexander, R. D. 1961. Aggressiveness, territorial- York. ity, and sexual behavior in field crickets (Orthop- Hennig, R. M. 1988. Ascending auditory interneu- tera: Gryllidae). Behaviour 17:130-223. rons in the cricket Teleogryllus commodus (Walker): Atkins, C, D. Atkins, D. Schoun, and J. F. Stout. Comparative physiology and direct connections 1987. Scototaxis and shape discrimination in the with afferents.J. Comp. Physiol. A 163:135-143. female cricket Acheta domesticus in an arena and Hennig, R. M. 1989. Neuromuscular activity during on a compensatory treadmill. Physiol. Entomol. stridulation in the cricket Teleogryllus commodus. 12:125-133. J. Comp. Physiol. A 165:837-846. Atkins, G., F. Ligman, F. Burghardt, and J. F. Stout. Honegger, H.-W. and R. Campan. 1989. Vision and 1984. Changes in phonotaxis by the female cricket visually guided behavior. In F. Huber, T. E. Acheta domesticus after killing identified acoustic Moore, and W. Loher (eds.), Cricket behavior and interneurons. J. Comp. Physiol. A 154:795—804. neurobiology, pp. 147-177. Cornell University Bennet-Clark, H. 1989. Cricket songs and the physics Press, Ithaca, New York. of sound production. In F. Huber, T. E. Moore, Huber, F. 1960. Experimentelle Untersuchungen zur and W. Loher (eds.), Cricket behavior and neuro- nervosen Atmungsregulation der Orlhopteren biology, pp. 227-261. Cornell University Press, (Saltatoria, Gryllidae). Z. Vergl. Physiol. 43:359- Ithaca, New York. 391. Bentley, D. R. 1969. Intracellular activity in cricket Huber, F. 1987. Plasticity in the auditory system of neurons during generation of song patterns. Z. crickets: Phonotaxis with one ear and neuronal Vergl. Physiol. 62:267-283. reorganization within the auditory pathway. J. Boyd, P. and B. Lewis. 1983. Peripheral auditory Comp. Physiol. A 161:583-604. directionality in the cricket (Gryllus campestris L., Huber, F. 1988. Invertebrate neuroethology: Guid- Teleogryllus oceanicus Le Guillon). J. Comp. Phys- ing principles. In J. M. Camhi (ed.), Invertebrate iol. 153:523-532. neuroethology. Experentia 44:428-431. Brunner, D. and T. Labhart. 1987. Behavioral evi- Huber, F. 1989. Cricket neuroethology: Neuronal dence for polarization vision in crickets. Physiol. basis of intraspecific acoustic communication. Entomol. 12:1-10. Adv. Study Behav. 19. Academic Press, San Diego. Bullock, T. H. 1984. Comparative neuroscience holds (In press) promise for quiet revolution. Science 225:473- Huber, F. 1990. Aus der Welt der Grillen: Verhalten 478. und Nervenzellen. Abh. Mainzer Akad. Wiss. Lit. Cade, W. H. 1975. Acoustically orienting parasit- (In press) oids: Fly phonotaxis to cricket song. Science 190: Huber, F., H.-U. Kleindienst, T. Weber, andj. Thor- 1312-1313. son. 1984. Auditorybehaviorofthecricket.III. 626 FRANZ HUBER

Tracking of male calling song by surgically and Oldfield, B. P., H.-U. Kleindienst, and F. Huber. developmentally one-eared females, and the curi- 1986. Physiology and tonotopic organization of ous role of the anterior tympanum. J. Comp. auditory receptors in the cricket Gryllus bnnacu- Physiol. A 155:725-738. latus DeGeer. J. Comp. Physiol. A 159:457-464. Huber, F., T. E. Moore, and W. Loner. 1989. Cricket Otto, D. and T. Amon. 1986. Interneuronen des behavior and neurobiology. Cornell University Press, Unterschlundganglions der Grille, die den Atem- Ithaca, New York. rhythmus iibertragen. In N. Eisner and W. Rath- Huber, F. and J. Thorson. 1985. Cricket auditory mayer (eds.), Sensomotonk identifizierle Neurone, p. communication. Sci. Am. 253:60-68. 48, Beitrage zur 14. Gottinger Neurobiologen- Kleindienst, H.-U. 1980. Biophysikalische Unter- tagung, G. Thieme, Stuttgart, New York. suchungen am Gehorsystem von Feldgrillen. Otto, D. and R. Campan. 1978. Descending inter-

Ph.D. Thesis, Universitat Bonn. neurons from the cricket subesophageal gan- Downloaded from https://academic.oup.com/icb/article/30/3/609/225053 by guest on 30 September 2021 Kleindienst, H.-U. 1987. AkustischeOrtungbei Gril- glion. Naturwissenschaften 65:491. len. Dtsch. Arbeitsgemeinschaft Akustik (DAGA) Otto, D. and T. Weber. 1982. Interneurons descend- 87:25-46. ing from the cricket cephalic ganglia that dis- Kleindienst, H.-U., T. U. Koch, and D. W. Wohlers. charge in the pattern of two motor rhythms. J. 1981. Analysis of the cricket auditory system by Comp. Physiol. 148:209-219. acoustic stimulation using a closed sound field. J. Pollack, G. S. and R. R. Hoy. 1989. Evasive acoustic Comp. Physiol. 141:283-296. behavior and neurobioiogical basis. In F. Huber, Kleindienst, H.-U., D. W. Wohlers, and O. N. Larsen. T. E. Moore, and W. Loher (eds.), Cricket behavior 1983. Tympanal membrane motion is necessary and neurobiology, pp. 340-363. Cornell University for hearing in crickets. J. Comp. Physiol. 151: Press, Ithaca, New York. 397-400. Pollack, G. S., F. Huber, and T. Weber. 1984. Fre- Klopffleisch, K. D. 1973. Ethologische Untersuch- quency and temporal pattern-dependent phono- ungen an der Feldgrille Cryllus campestris L. an taxis of crickets (Teleogryllus oceanicus) during natiirlichen Standorten. Diploma-Thesis, Uni- tethered flight and compensated walking. J. versitat Koln. Comp. Physiol. A 154:13-26. Koch, U. T. 1983. A method for recording respi- Prozesky-Schulze, L., O. P. M. Prozesky, F. Ander- ratory movements in an unrestrained insect. In son, and G. J. J. van der Merwe. 1975. Use of a W. Nachtigall (ed.), Insect flight, I-Biona report, self made sound baffle by a tree cricket. Nature pp. 41-50. G. Fischer, Stuttgart, New York. 255:142-143. Koudele, K., J. F. Stout, and D. Reichert. 1987. Fac- Schildberger, K. 1981. Some physiological features tors which influence female crickets' (Acheta of mushroom-body linked fibers in the house domesticus) phonotactic and sexual responsiveness cricket brain. Naturwissenschaften 67:623. to males. Physiol. Entomol. 12:67-80. Schildberger, K. 1984a. Multimodal interneurons in Kutsch, W. and F. Huber. 1989. Neuronal basis of the cricket brain: Properties of identified mush- song production. In F. Huber, T. E. Moore, and room body cells. J. Comp. Physiol. A 154:71-79. W. Loher (eds.), Cricket behavior and neurobiology, Schildberger, K. 19846. Temporal selectivity of iden- pp. 262-309. Cornell University Press, Ithaca, tified auditory neurons in the cricket brain. J. New York. Comp. Physiol. A 155:171-185. Labhart, T. 1988. Polarization-opponent interneu- Schildberger, K. and M. Horner. 1988. The function rons in the insect visual system. Nature 331:435- of auditory neurons in cricket phonotaxis. I. 437. Influence of hyperpolarization of identified neu- Labhart, T., B. Hodel, and I. Valenzuela. 1984. The rons on sound localization. J. Comp. Physiol. A physiology of the cricket's compound eye with 163:621-631. particular reference to the anatomically special- Schildberger, K. and F. Huber. 1988. Post-lesion ized dorsal rim area. J. Comp. Physiol. A 155: plasticity in the auditory system of the cricket. In 289-296. H. Flohr(ed.),Poj//f«o)! neural plasticity, pp. 565- Larsen, O. N., H.-U. Kleindienst, and A. Michelsen. 575. Springer-Verlag, Berlin, Heidelberg, New 1989. Biophysical aspects of sound reception. In York, London, Paris, Tokyo. F. Huber, T. E. Moore, and W. Loher (eds.), Schildberger, K., F. Huber, and D. W. Wohlers. Cricket behavior and neurobiology, pp. 364-390. 1989. Central auditory pathway—neuronal cor- Cornell University Press, Ithaca, New York. relates of phonotactic behavior. In F. Huber, T. Loher, W. and M. Dambach. 1989. Reproductive E. Moore, and W. Loher (eds.), Cricket behavior behavior. In F. Huber, T. E. Moore, and W. Loher and neurobiology, pp. 423-458. Cornell University (eds.), Cricket behavior and neurobiology, pp. 43-82. Press, Ithaca, New York. Cornell University Press, Ithaca, New York. Schildberger, K. and H.-U. Kleindienst. 1989. Sound Moiseff, A. and R. R. Hoy. 1983. Sensitivity to ultra- localization in intact and one-eared crickets, Gryl- sound in an identified auditory interneuron in /«» bimamlntuf. J. Comp. Ph\siol. A 165:615-626. the cricket: A possible neural link to phonotactic Schmitz, B. 1989. Neuroplasticity and phonotaxis in behavior. J. Comp. Physiol. 152:155-167. monaural adult female crickets (Cnllus bimacu- Moiseff, A., G. S. Pollack, and R. R. Hoy. 1978. lntu\ DeGeer).J. Comp. Physiol. A 164:343-358. Steering responses of fl) ing crickets to sound and Schmit7, B., H.-U. Kleindienst, K. Schildberger, and ultrasound: Mate attraction and predator avoid- F. Huber. 1988. Acoustic orientation in adult, ance. Proc. Xatl. Acad. Sci. U.S.A. 75:4052-4056. female crickets (Gnllu* Inmaculatus DeGeer) alter NERVE CELLS AND INSECT BEHAVIOR 627

unilateral foreleg amputation in the larva. J. Ulagaraj, S. M. and T. J. Walker. 1973. Phonotaxis Comp. Physiol. A 162:715-728. of crickets in flight: Attraction of male and female Selverston, A. I. 1980. Are central pattern genera- crickets to male calling songs. Science 182:1278- tors understandable? Behav. Brain Sci. 3:535- 1279. 571. Walker, T. J. 1982. Sound traps for sampling mole Selverston, A. I., H.-U. Kleindienst, and F. Huber. cricket rights (Orthoptera, Gryllotalpidae: Scap- 1985. Synaptic connectivity between cricket teriscus). Fla. Entomol. 65:105—110. auditory interneurons as studied by selective pho- Weber, T. 1989. Acoustic and visual orientation in toinactivation. J. Neurosci. 5:1283-1292. crickets. In K. Wiese (ed.), Sensory systems and com- Stabel.J. 1988. Der Mechanimus der Richtungsbe- munication in arthropods. Birkhauser, Basel. (In stimmung und seine Beziehung zur Gesangser- press)

kennung bei der Phonotaxis der Grille (Gryllus Weber, T., G. Atkins, J. F. Stout, and F. Huber. 1987. Downloaded from https://academic.oup.com/icb/article/30/3/609/225053 by guest on 30 September 2021 bimaculatus DeGeer). Ph.D. Thesis, Universitat Female Acheta domesticus track acoustical and visual Koln. targets with different walking modes. Physiol. Stabel, J., G. Wendler, and H. Scharstein. 1989. Entomol. 12:141-147. Cricket phonotaxis: Localization depends on rec- Weber, T. andj. Thorson. 1989. Phonotactic behav- ognition of the calling song pattern. J. Comp. ior of walking crickets. In F. Huber, T. E. Moore, Physiol. A 165:165-177. and W. Loher (eds.), Cricket behavior and neuro- Stout,J. F., G. Atkins, T. Weber, and F. Huber. 1987. biology, pp. 310-339. Cornell University Press, The effect of visual input on calling song attrac- Ithaca, New York. tiveness for female Acheta domesticus L. Physiol. Weber, T., J. Thorson, and F. Huber. 1981. Audi- Entomol. 12:135-140. tory behavior of the cricket. I. Dynamics of com- Stout, J. F., C. H. deHaan, and R. W. McGhee. 1983. pensated walking and discrimination paradigmns Attractiveness of the male Acheta domesticus call- on the Kramer treadmill. J. Comp. Physiol. 141: ing song to females. I. Dependence on each of 215-232. the calling song features. J. Comp. Physiol. 153: Wendler, G., M. Dambach, B. Schmitz, and H. Schar- 509-521. stein. 1980. Analysis of the acoustic orientation Stout, J. F., G. Gerard, and S. Hasso. 1976. Sexual behavior in crickets, Gryllus campestns L. Natur- responsiveness mediated by the corpora allata and wissenschaften 67:99-100. its relationship to phonotaxis in the female cricket, Wohlers, D. W. and F. Huber. 1982. Processing of Acheta domesticus L. J. Comp. Physiol. 108:1-9. sound signals by six types of neurons in the pro- Thorson, J., T. Weber, and F. Huber. 1982. Audi- thoracic ganglion of the cricket, Gryllus campestris tory behavior of the cricket. II. Simplicity of call- L. J. Comp. Physiol. 146:161-173. ing-song recognition in Gryllus and anomalous phonotaxis at abnormal carrier frequencies. J. Comp. Physiol. 146:361-378. Downloaded from https://academic.oup.com/icb/article/30/3/609/225053 by guest on 30 September 2021