The insect visual system

The interneurons that process visual information in insects are arranged as a series of neuropiles that lie beneath a compound eye and make up an optic lobe of the brain (Fig. 5.1). Most interneurons connect two layers of neuropile, with their cell bodies and input synapses in one layer and an axon that conducts information towards the brain into the next layer of neuropile. Information, therefore, tends to flow sequentially through different layers of neuropile. But there are also circuits that make lateral connections within one layer, and some neurons convey information outwards, from the brain towards the eye.

The photoreceptors convey a point-by-point representation of the environment to the first neuropile, the lamina. Axons that connect the lamina with the next neuropile, the medulla, cross over one another so that the rear of the medulla is connected with the front of the lamina. The spatial representation of the environment is preserved in an array of columnar elements within the medulla, and the presence of a topographic map of the environment is a universal feature of complex visual systems. (Topographic means 'description of place' in Greek.) In each medulla column, there are several tens of types of neuron. It is likely that different types of neuron in a

Motor Neurons Insect

Figure 5.1 The visual system of a diurnal insect, such as a dragonfly or a locust, showing the main neuronal pathways involved in movement detection. The three optic neuropiles (lamina, medulla and lobula) are indicated by light stippling. The small arrows show the pathway followed by information originating in the uppermost ommatidium that is drawn. The general arrangement is similar in flies, except that the photoreceptors of one ommatidium are not fused, and the lobula contains a distinct neuropile called the lobula plate, which would lie on top of the lobula in this drawing.

Figure 5.1 The visual system of a diurnal insect, such as a dragonfly or a locust, showing the main neuronal pathways involved in movement detection. The three optic neuropiles (lamina, medulla and lobula) are indicated by light stippling. The small arrows show the pathway followed by information originating in the uppermost ommatidium that is drawn. The general arrangement is similar in flies, except that the photoreceptors of one ommatidium are not fused, and the lobula contains a distinct neuropile called the lobula plate, which would lie on top of the lobula in this drawing.

medulla column are dedicated to filtering out particular stimulus features that occur within the visual field of one ommatidium, or of a group of neighbouring ommatidia. Many of the neurons in the next layer, the lobula, are also arranged in topographic columns. It is hard to make electrophysiological recordings from these columnar neurons in the lobula and medulla because they are tiny. However, some particularly large neurons also occur in the lobula, and stable recordings can be made from them so that their response properties can be studied in detail. These neurons have fields of view that incorporate many ommatidia, and they abstract information about particular types of movement which can occur anywhere within a very large field of view. These movement-sensitive cells, therefore, provide excellent examples of feature-detecting neurons.

The first neuropile, the lamina, is divided into an array of discrete neural units called cartridges. Each cartridge receives its input from photoreceptors that share the same field of view. In most species, including locusts and dragonflies, this means that a cartridge receives input from photoreceptors of the ommatidium that lies above it. In advanced flies, however, each cartridge receives input from just one photoreceptor in the ommatidium directly above, and from one photoreceptor from each of five surrounding ommatidia (see Fig. 4.5b, p. 87). Each cartridge contains the same complement of neurons, and the set of connections made within a cartridge is so precisely determined and repeated across the eye that the structure of the lamina has been described as crystalline.

The principal cells of each lamina cartridge are called large monopolar cells. As the name suggests, the axon of a large monopolar cell gives rise to just one process. A tuft of short, brush-like dendrites arises from it and, in blowflies, these dendrites make almost exactly 200 discrete anatomical synapses with each connecting photoreceptor (Nicol & Meinertzhagen, 1982). The large monopolar cell process then becomes a smooth axon which, together with the axons of smaller neurons, conveys information from the lamina to the medulla. From each ommatidium, two central photoreceptors, that are tuned to detect ultraviolet and blue light, pass straight through the lamina and terminate in the medulla. Within the lamina, amacrine cells make lateral connections between adjacent cartridges, and there may also be feedback neurons that carry signals back from the lamina to photoreceptors.

+1 0

Post a comment