This paper describes the neuropilar areas of the thoracic ganglia of locusts ( Locusta migratoria ) in which projections from mechanoreceptors terminate. Reference transverse, sagittal and parasagittal sections through a meso- and metathoracic ganglion illustrate the relation of these neuropiles to known structures such as longitudinal tracts and commissures. The ventral association centres (VAC), divided into anterior (aVAC), medial (mVAC), ventralmost (vVAC) and lateral (IVAC) parts, receive terminals from tactile hairs (H; aVAC, vVAC and IVAC) and chordotonal organs (CO; mVAC). Newly defined neuropilar areas are the lateral association centres (LAC) which are divided into an anterior (aLAC) and posterior part (pLAC). These receive projections from hair plates (HP), campaniform sensilla (CS), multipolar sensory cells (MS), strand receptors (SR) and chordotonal organs (CO). In contrast to the VACs they also contain fibres of motorneurons.

During walking, the muscles of a leg undergo a typical sequence of activity, which is partly shaped by phasic sensory feedback. To assess the role played by such feedback, we characterized intracellularly the receptive fields of tarsal motor neurons in the locust Schistocerca gregaria and considered these receptive fields within the context of a step cycle. The depressor motor neurons, active during the stance phase, are excited by ventral tarsal contact or an imposed levation and are inhibited by dorsal contact or an imposed depression. Partial deafferentation of the anterior tarsus reduces this stance phase depressor activity. The levator motor neuron, active during the swing phase, has the opposite receptive field. The retractor unguis motor neurons, synergistic to the depressors, are, like them, excited by ventral contact but, like the levator, are inhibited by afferents which can signal the end of the stance phase. The inhibition of the retractors could constitute a preparation for the swing phase, by reducing the grip on the substrate. The motor neuronal receptive fields thus appear to support the patterns of muscular activity recorded during walking. Excitation of the motor neurons by extero- and proprioceptors is usually direct: hair, canal, campaniform, and chordotonal afferents all evoke 1:1 EPSPs in motor neurons after a central latency of 1-1.5 msec. Inhibition is indirect, as IPSPs occur at least 2 msec later than the EPSPs. The motor neurons of one pool have overlapping but not necessarily identical, receptive fields. Parallel, supplementary excitatory and inhibitory pathways involving nonspiking local interneurons also exist, which can allow gain control of a specific reflex. The weight given to a reflex response will thus depend, first, on the number of motor neurons in a pool affected by the stimulus and, second, on the existence and state of intercalated interneurons.

Projections of primary insect mechanoreceptor fibres in Locusta migratoria and Acheta domesticus are visualised with the cobalt intensification technique after axonal or dendritic filling with cobaltous chloride. Chordotonal proprioceptors of the locust mesothoracic thoracocoxal joint have typical segmental projections and more or less widespread interganglionic processes into the metathoracic, prothoracic and even suboesophageal ganglia. Similar observations can be made on locust abdominal chordotonal organs, from some of which all primary axons project into five preceding neuromeres or ganglia. From locust abdominal stretch receptors and cricket mechanosensitive sternal hairs single afferent axons extend through several ganglia. Locust campaniform sensilla and clavate hair projections of crickets terminate locally. It is concluded that many interganglionic primary afferents contribute to intersegmental control of motor coordination in insects.

No abstract

The central projections of hairplates and groups of campaniform sensilla of various parts of the locust head, neck, and thorax are compared. Receptors with approximately the same peripheral location (e.g., dorsal head, wing, or proximal leg) tend to exhibit similar central projection patterns. Specificity of the projections due to different morphological types of sensory neurones may influence shape and extent of segmental ramifications; this is observed with receptors of the proximal leg joints in particular.