Insect walking has often been used as a model system for the study of rhythmic behavior, and research over the past several decades has revealed much about the insect locomotor system. First, in common with other rhythmic behavior, the alternating action of antagonistic muscle pairs (leg extensors and flexors) is generated by networks of neurons that can produce rhythmic output without needing any external timing signals. Such networks are called central pattern generators (CPGs). There is at least one CPG per leg; some evidence suggests that there may even be one per joint. The CPGs are coupled together via interneurons so that the activity of one can influence the activity of another. Second, sensory feedback from the moving legs is in most cases critical for the proper coordination of the legs during walking. In some insects (such as stick insects like Carausius morosus) the requirement for sensory feedback is so strong that coordination virtually disappears if it is eliminated. In other insects (such as cockroaches like Periplaneta americana), the need for sensory feedback may be reduced during fast walking.

My research has focussed on two main issues. First, how are CPGs for different legs coupled together? This question can be answered in part by experiments in which leg movements are disrupted in some way. Results suggest that whereas ipsilateral (same side) CPGs can change coordination (i.e., coupling is loose), contralateral pairs of CPGs are tightly coupled and hence always show the same coordination relative to one another. Second, what sensory feedback from the legs is most critical for coordination, and how does this feedback affect coordination? Research in several labs has suggested that in insects cuticular mechanoreceptors (campaniform sensilla) that respond to strain in the cuticle, are critical.