Sensorimotor Integration in the Bat Echolocation SystemHOME | PROJECTS by Melville Wohlgemuth
Sensorimotor transformations are fundamental to perceptually-guided behaviors in all animal systems, and yet a comprehensive understanding of the underlying mechanisms presents a major challenge in the field of neuroscience. With the goal of filling this gap, I am conducting a series of experiments on sensorimotor integration in the bat echolocation system, an animal whose stimulus environment and motor actions can be explicitly measured. My project involves recording neural activity from the midbrain superior colliculus (SC) of the echolocating bat as it produces sonar vocalizations and listens to echo returns from the environment. The SC is a laminated structure receiving superficial innervation from cortical and subcortical sensory structures, and containing premotor neurons in the deeper layers that play a role in directing orientation behaviors (May, 2006). It is hypothesized that sensorimotor integration for orientation occurs in both dorsal and ventral directions across the layers of the SC, with sensory information entering the superficial layers (Cyander et al, 1972; Goldberg and Wurtz, 1972) and motor information emanating from the deeper layers in the form of muscle commands and efference motor information (Dean et al, 1976; Marrocco et al, 1981; Huerta and Harting, 1982; Leigh et al, 1997; Meredith et al, 2001). My research on the neural mechanisms of sensorimotor integration exploits the active components of echolocation. Particularly valuable to my study is the batís ability to adaptively control the rate and duration of its sonar calls. Changes in these behavioral features can be accurately measured, providing a window into the batís ongoing perceptual experiences and processes.
Figure 1: Schematic of insect approach and capture in echolocating bats. a). The three phases of prey pursuit and capture: search, approach and terminal buzz. b). Spectrograms of representative sonar calls during each stage. Note the increase in call rate and decrease in call duration as the bat advances toward the insect.
Acoustic orientation by sonar allows the bat to search and capture insect prey in complete darkness. As a bat approaches an insect, it directs its sonar beam at the target by using information conveyed in the echoes (Ulanovsky and Moss, 2008). There are three distinct phases of insect pursuit and capture (Figure 1), and each phase is characterized by different temporal characteristics of sonar pulses (Surlykke and Moss, 2000). In the search phase, the bat emits long pulses at a rate of 5-10 Hz to scan the environment for insects (Fig.1, Search Phase). Once an insect has been selected, the bat switches to an approach phase that is distinguished by an increase in call rate and decrease in call duration, allowing the bat to sample the position of the target more frequently (Fig. 1, Approach Phase). In the final phase Ė the terminal buzz Ė the bat locks its sonar beam onto the target, and maximizes the rate of very short duration calls to accurately localize the position and velocity of the prey (Fig. 1, Buzz Phase). In each phase, the bat uses incoming auditory information to guide pre-motor commands for future vocalizations.
Several lines of evidence suggest that the SC plays an important role in the sensorimotor integration necessary
for species-specific orienting behaviors (Valentine and Moss, 1997; Valentine et al., 2002; Moss and Sinha, 2003).
In visually guided primates (such as humans), the anatomy and physiology of the SC are dominated by vision (May, 2006).
For bats, the principal method of exploration is echolocation: a behavior involving audition and vocalization (Moss and Sinha, 2003).
Responses throughout the bat SC reflect this natural orienting behavior. Neurons in the intermediate layers of the bat SC primarily respond
to auditory cues, including spatial location and echo delay (Valentine and Moss, 1997); while the deeper layers send efferents
to areas controlling vocalizations (Sinha and Moss, 2007), pinna movements, and head direction (Valentine et al, 2002).
Echolocation therefore involves a computation of stimulus location derived from auditory cues, followed by the production
of vocalizations designed further inform the bat about stimulus position. How these auditory cues are processed
into vocal premotor commands in the bat SC is the central question I am addressing in my research.
Experiment 1: SC responses to echo returns from the batís own sonar calls.Previous research on the auditory response properties of the bat SC have used artificially generated sonar sounds in the passively listening bat (Valentine and Moss, 1997). From this research, two classes of auditory neurons were identified, 2-D and 3-D. Spatial selectivity of 3-D neurons depends on the azimuth, elevation and distance (echo delay) of simulated sonar targets. It has been hypothesized that these neurons code the spatial location of sonar targets to guide behaviorally appropriate orienting behaviors. Although this research has demonstrated 3-D spatial response profiles in the bat SC, the behavioral relevance of these data are limited by the methods employed. In this study, I will use the batís own vocalizations to drive the system, placing the bat in a more naturalistic condition where it is engaged in the task, and biologically relevant auditory computations in the SC can be assayed.
Experiment 2: SC sensorimotor activity during sonar target tracking of single objects.The auditory responses identified in the first experiment will be characterized with respect to motor outputs through this second experiment. Previous research has examined either the superficial layers or the deeper layers of the SC, but few recordings have been made of neural signals from several layers concurrently. Neural activity collected simultaneously from sensory and motor areas in the SC is necessary to generate an understanding of sensorimotor transformations across the structure. The data from this experiment will be informative about the contribution of sensory and motor encoding in each layer of the SC, and will also reveal how neural activity evolves from a sensory signal to a motor signal across the layers of the SC.
Experiment 3: SC sensorimotor activity during sonar target tracking of multiple objects.Behavioral studies in the Moss laboratory on bat echolocation have identified differential control of the rate and duration of echolocation pulses. This result was demonstrated by placing distracting objects in the batís field of view as it echolocates an approaching food reward. It was found that call duration is determined by the closest object, irrespective of the objects reward value; whereas echolocation pulse rate was solely determined by the distance to the food reward. The bat is therefore tracking multiple objects in space, and this behavioral paradigm will be used to identify the role of the SC in the more naturalistic condition of multiple-target tracking. How the SC is involved in the tracking of multiple targets is not well understood. The results of the current experiment will generate new data on the effects of multiple objects on SC activity.
Conclusions and Broader Impacts: