Insects rely heavily on olfaction to locate food, find mates and sense danger. Odorant stimuli in their natural environment often consist of mixtures of several chemical compounds, and stimulus concentration in air fluctuates rapidly and unpredictably. Despite the complex nature of natural odorant stimuli, odor identity can be encoded by insects' olfactory systems very quickly. How this can be achieved is still an open area of study. Here, I ask two specific questions: 1. Are the olfactory responses to mixture stimuli qualitatively different from those to pure chemical compounds and are such differences relevant to olfactory coding? and 2. Which types of temporal spike information can potentially be useful for olfactory coding? To address the first question, I extended a standard receptor model to mixtures. Through mathematical analysis of the extended model, I found that first-spike latencies are shorter and activity patterns are more stable across concentrations in olfactory receptor neurons for mixtures than for pure odorants. I then built a computational model of the early olfactory system of honey bees and showed that the above-mentioned mixture effects can also be observed deeper in insects' brain. These results suggest that mixtures can be more efficiently identified by insects than pure odorants. To address the second question, I developed mathematical methods to approximate the probability distribution of the first-spike latency of a leaky integrate-and-fire neuron upon receiving an external input signal with auto-correlated background noise. The approximation is excellent for a wide range of sizes and shapes of the signal, even when the dynamical time scale of the signal is comparable to that of membrane integration. The methods allow efficient evaluation of how reliable certain correlation patterns of spikes from multiple neurons would be formed after a sensory stimulus is applied, which indicates whether these patterns could carry information about the stimulus.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:759590 |
| Date | January 2018 |
| Creators | Chan, Ho Ka |
| Publisher | University of Sussex |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://sro.sussex.ac.uk/id/eprint/79823/ |
Page generated in 0.0016 seconds