Sensory memory is a short-lived persistence of a sensory stimulus in the nervous system, such as iconic memory in the visual system. However, little is known about the mechanisms underlying olfactory sensory memory. We have therefore analyzed the effect of odor stimuli on the first odor-processing network in the honeybee brain, the antennal lobe, which corresponds to the vertebrate olfactory bulb. We stained output neurons with a calcium-sensitive dye and measured across-glomerular patterns of spontaneous activity before and after a stimulus. Such a single-odor presentation changed the relative timing of spontaneous activity across glomeruli in accordance with Hebb's theory of learning. Moreover, during the first few minutes after odor presentation, correlations between the spontaneous activity fluctuations suffice to reconstruct the stimulus. As spontaneous activity is ubiquitous in the brain, modifiable fluctuations could provide an ideal substrate for Hebbian reverberations and sensory memory in other neural systems.

The antennal lobe plays a central role for odor processing in insects, as demonstrated by electrophysiological and imaging experiments. Here we analyze the detailed temporal evolution of glomerular activity patterns in the antennal lobe of honeybees. We represent these spatiotemporal patterns as trajectories in a multidimensional space, where each dimension accounts for the activity of one glomerulus. Our data show that the trajectories reach odor-specific steady states (attractors) that correspond to stable activity patterns at about 1 second after stimulus onset. As revealed by a detailed mathematical investigation, the trajectories are characterized by different phases: response onset, steady-state plateau, response offset, and periods of spontaneous activity. An analysis based on support-vector machines quantifies the odor specificity of the attractors and the optimal time needed for odor discrimination. The results support the hypothesis of a spatial olfactory code in the antennal lobe and suggest a perceptron-like readout mechanism that is biologically implemented in a downstream network, such as the mushroom body.