An emerging model from molecular studies of reproductive behavior suggests that the generation of reproductive isolation and species formation is, in many cases, a neurological phenomenon. We hope to apply this model of species differentiation to a group of rapidly diversifying Hawaiian crickets. In doing so, we plan to identify a subset of the genes responsible for their extraordinary radiation. Within the Hawaiian archipelago, swordtail crickets have diverged into approximately 173 species. Within this group, divergence within the genus Laupala is particularly well understood. The diversification of Laupala species has been closely linked to the divergence of male calling song, particularly the increase (or decrease) of pulse rate. Building on over fifty years of neurological studies of stridulation in crickets, our work will attempt to identify specific changes in gene expression related to the development of the stridulatory neural circuitry. A cDNA library derived from the relevant ganglia is currently being constructed. In addition to specific candidate genes identified from related model systems, we will use this library to track changes of gene expression related to the development of the stridulatory circuit. This project represents a first step toward understanding the behavioral diversification of Hawaiian crickets at the molecular level

An African wave-type electric fish, Gymnarchus niloticus, ceases its electric organ discharge (EOD) for a prolonged time in response to external electrical signals as a natural behavioral repertoire. During the cessation of EODs, a weak sinusoidal signal (~0.1 mV/cm) near the fish's previous discharge frequency was recorded near the body. The oscillatory potentials at all points on the body surface were synchronized and had a complex spatial distribution. The source of the potential was determined to be within the dermal tissue. Electroreceptive central neurons that responded to a moving target near the fish under normal EODs also responded to the same target during the EOD interruption. This suggests that the fish may be able to electrolocate objects without the discharge from the electric organ

Many insectivorous bats use echolocation to detect, track, and capture flying insects at night. Praying mantids, like many insects, have evolved ears sensitive to bat sonar and perform maneuvers effective in evading capture. One maneuver mantids perform is a power dive, occurring when a bat is close and possibly executing a capture attempt. The timing of the power dive can be as critical as the maneuver itself since an early response will allow the bat to modify its behavior and continue tracking the mantis while a late response may not give the mantis enough time to complete the power dive. Bat vocalization parameters (such as pulse repetition rate-PRR and pulse duration-PD) change predictably as they approach and capture prey. Mantids and other insects could use these changes to assess the level of danger posed by an approaching bat and time the initiation of the evasive response.

One mantis auditory interneuron, 501-T3, possesses several properties consistent with initiating bat-evasive maneuvers and may be important in triggering the power dive. These properties include sensitivity to ultrasound (58 dB SPL at 35 kHz), short response latency (10-14 ms to reach the prothorax), and a large ascending axon (17 micron diameter) with a fast conduction velocity (4 m/s). To determine how well 501-T3 encodes changes in bat vocalizations occurring during capture attempts, we used an electrode implanted in a tethered mantis to record 501-T3 responses extracellularly during actual bat attacks. We conducted experiments in a large flight room using a free-flying bat trained to capture the tethered mantis. In the flight room, the bat attack sequence has three phases: approach (low PRR; > 3 ms PD), buzz I (increasing PRR; decreasing PD), and buzz II (very high PRR, < 1.5 PD). 501-T3 produces multi-spike bursts to every bat vocalization during the approach phase and into the buzz I phase. However, 501-T3 ceases burst responses during the buzz I phase, before the bat enters the buzz II phase. On average, 501-T3 can follow PRRs up to 55 pps. 501-T3 spike bursts cease 255.9 ms (72.0 cm away from the prep) before the bat hit the mantis. The 501-T3 burst cessation time corresponds to the first sign of flight path changes during the mantis power dive (242.2 ms after stimulus onset). These results suggest that the buzz I phase triggers the power dive, possibly timing the power dive to occur during the buzz II phase. The cessation of 501-T3 activity could protect 501-T3 and downstream neural elements important in executing the power dive from habituation during successful escapes.

[Supported by: NSF IBN-9808859 (DDY) and NRSA F31MH12025 (JDT)]

Crocodilians are an ancient monophyletic group that has existed since before the time of the dinosaurs. Living crocodilians are divided into alligatorids, crocodiles and gavials. All have an amphibious life-style, spending their time on land, underwater and on the interface of the two media, although some extinct forms were strictly terrestrial or aquatic. Behavioral, anatomical and physiological data shown here demonstrate that these predatory animals have evolved a unique sensory organ that mediates orientation to disruption of the water surface. Pressure waves created by these disruptions stimulate dome pressure receptors (DPR) on the crocodilian skin. Removing DPRs abolishes the orienting behavior. The ancient nature of these sensory organs is reflected in the fossil record. Typical patterns of foramina in jaw bones associated with DPR innervation appear in extinct specimens in the early Jurrassic. These osteological markings are present only in animals believed to have had an amphibious life style and are absent in the extinct fully terrestrial or aquatic forms.

Contrast sensitivity functions (CSFs) have been obtained from a wide variety of mammals and some birds. These data consistently have shown that avian species have much lower maximum contrast sensitivity than do a wide range of mammals and fishes. Unfortunately the avian data are extremely sparse compared to the mammalian data. In a meta-analysis (Hodos et al., 1997) the relationship between contrast sensitivity and various stimulus, optical, and retinal variables were investigated in a large number of species of mammals and several fish and bird species. Examples of these variables include target luminance, target area, posterior nodal distance, cone-photoreceptor density, and ganglion-cell density. While many of these variables had high correlations with peak spatial frequency and maximum contrast sensitivity (characteristics of the peak of the CSF) within mammals, fishes and birds, none of these variables could account for the differences in contrast sensitivity between the avian group and mammals/fishes group.

To determine whether low contrast sensitivity is a general phenomenon among birds, the CSFs of 8 different species of birds are being obtained using the pattern electroretinogram (PERG). These 8 species represent 8 different taxonomic orders, and are adapted to diverse ecological habitats. This research may confirm a fundamental difference in contrast processing between mammals and birds, in showing that this contrast sensitivity difference found thus far is not due to sampling error.

In vocal learning birds and mammals auditory feedback is necessary for the acquisition and maintenance of learned vocalizations. In mammals, auditory information is conveyed to the cortex via the thalamocortical pathway, terminating in the primary auditory cortex. In birds two auditory pathways convey information from the brainstem to the telencephalon: the thalamotelencephalic pathway terminating in Field L and an isthmofrontal pathway interconnecting the intermediate nucleus of the lateral lemniscus with the nucleus Basalis (Bas) in the frontal telencephalon. In vocal learning oscine songbirds pathway tracing and functional mapping studies using the immediate early genes (IEG) Zenk and c-fos demonstrate that Field L, but not Bas, provides auditory feedback necessary for vocal learning during the early sensitive period. In psitticine species, which are capable of lifelong vocal learning, such as the budgerigar, both Bas and Field L project to areas of the frontal telencephalon which are interconnected with vocal control circuits. Functional mapping studies in budgerigars using the IEG Zenk, however, suggest that only the thalamotelencephalic pathway exhibits gene induction in response to warble song stimulation. Presently, we show that in response to a repeating unfamiliar contact call c-fos is rapidly and robustly induced throughout Bas. C-fos expression is depolarization dependent, transcribed in response to CREB binding at its promoter and associated with learning and memory processes. The current finding demonstrates that in Bas activity dependent gene regulation can be driven by species specific auditory stimulation. Thus this study supports previous pathway tracing studies, indicating that auditory information reaching the vocal control circuitry in budgerigars is conveyed through both the thalamotelencephalic and isthmofrontal pathways, which to our knowledge is a unique adaptation among vocal learning animals. This research was supported in part by training grant MH-20048-01A1 from NIMH (T.F.R.) and a NSF grant IBN 9816061 (S.E.B.).

The bat's sonar receiver determines the direction and distance of a target from the features and timing of returning echoes. It uses the sonar echoes to build a 3-D representation of the world that is used to adjust the features of its vocalizations. Echolocation thus requires the dynamic interplay between auditory information processing and adaptive motor responses. Neural mechanisms supporting audiomotor integration in the big brown bat, Eptesicus fuscus, were investigated in the midbrain superior colliculus (SC).

Our earlier studies have shown that the functional organization of the bat SC exhibits specializations that are potentially important to acoustic orientation by sonar. We have shown that a population of auditory neurons in the bat SC shows echo-delay tuning, a response property hypothesized to encode target range. In addition, microstimulation of the SC elicits sonar vocalizations, and anatomical pathways have been identified connecting the SC with pre-motor vocal control nuclei. These findings suggest that the SC may play an important role in the audiomotor integration required for echolocation.

We describe results of neurophysiological recording experiments with awake, behaving bats. Bats were trained to rest on a platform in a dark room, while tracking a food reward using echolocation. The reward was moved in a space that extended 15-45 cm in front of the bat and up to 30° laterally from center, at zero elevation. Multi- and single unit neural activity was recorded from the intermediate and deep layers of the SC using high impedance electrodes, while simultaneously recording the bat's sonar vocalizations. Here, we delineate robust groups of neural activity preceding vocal onsets (group 1 peak activity: -16.2 ± 4.5 ms; group 2 peak activity: -2.8 ± 1.7 ms; n=560 vocalizations). The study of pre-motor and auditory responses in the SC of bats performing echolocation behavior deepens our understanding of the mechanisms supporting sensorimotor integration in bat echolocation.

Used for access to food, mates and shelter, aggression is a normal part of the behavioral repertoire of most animals. Despite its importance, relatively little is know of the biological basis of the behavior. Hormonal substances, including amines like serotonin, peptides like arginine vasopressin and gonadotropin releasing hormone and steroids like testosterone all have been implicated in aggression, but precisely how or where these substances act remains a mystery. Our laboratory has been examining a lobster model of aggression for the past 20 years. Lobsters are excellent animals for such studies because they fight readily and establish long-term dominance relationships, and because anatomical, physiological and molecular methods can be used to bring the analysis of the behavior to the level of identified neurons that are important in the behavior. One focus of our studies has been to define the roles of serotonin and other neurohormones in aggression in lobsters. Results of these studies will occupy the first part of this presentation. The second part of the presentation will deal with recent studies in our laboratory using common laboratory strains of the fruit fly, Drosophila melanogaster, as a model system to explore the genetics of aggression. With their genomes sequenced and with the wealth of mutants available for examination, these organisms offer a powerful experimental system with which to explore complex patterns of behavior.

Mustached bats emit a variety of complex sounds for audiovocal communication. These social calls are distinct from the stereotypic pulses emitted for echolocation. We are interested in the acoustic structure of these complex sounds and how they are processed. We have discovered that in some parts of the auditory cortex, there exist specialized neural mechanisms where cognition of social calls may depend on the peak firing rate of a neuron. The presence of these mechanisms discounts a spatiotemporal population code described previously for cats and nonhuman primates. In other parts of the cortex, the
increased probability of eliciting a temporal pattern of spikes within a neuron may signal the presence of a familiar call. Perhaps, the ultimate recognition of calls is based not purely on either firing rates or spike trains within individual neurons, but
on "microstates" of neural networks in the cortex that can be distinguished by specific patterns of correlated activity within a small group of neurons. We continue to explore these mechanisms to understand the decoding of social calls in mustached
bat's auditory cortex.

The design of visual signals presents a fascinating evolutionary problem. Signals must be easily seen and positively dentifiable by the receiver, and their design must cope with the environments in which they are used (where both lighting and background influence their visibility) as well as the visual system(s) of the intended receiver(s). While it is often hypothesized that visual systems within species coevolve with intraspecific signals, there is little or no evidence that vision is adapted to perceive signals. Instead, visual evolution appears to be controlled primarily by general visual tasks like detection of food or predators, or general imaging of the environment. Biological signals therefore adapt to pre-existing visual systems. In this talk, I will discuss intraspecific and interspecific signal design in two very colorful signaling systems, those of stomatopod crustaceans and poison-dart frogs.