UC Berkeley

The evolution of the first complex neural systems occurred in a sensory world defined by chemosensory stimuli. Despite its universality across the animal kingdom, the strategies animals use to sample for odors and orient using odors is still not clear. Odor plumes are complex structures. How animals sample their environment for odor stimuli, make decisions, and ultimately find an odorant using their sense of smell has been studied more in invertebrates and birds than mammals, yet olfaction is critical for most mammals, even humans. This dissertation will analyze the orientation and sampling strategies in two species of mammals that differ in their olfactory capacity yet are known to use odors to orient in space. It will examine olfactory navigation in both naturalistic conditions and controlled studies to understand how environmental conditions, individual differences, and olfactory cues affect search strategies, navigation behavior, and sampling behavior.

Chapter One addresses the nature of odors and olfaction, the structure of odor plumes, and a summary of our current knowledge of animal strategies for olfactory navigation through odor plumes. It also discusses how individual differences, such as sex differences and spatial cognitive abilities, may affect olfactory discrimination and subsequently, orientation.

Chapter Two investigates how a species known to be expert in olfactory navigation, the domestic dog, oriented to the outdoor scent trail of a hidden person. Search and rescue dogs were adept at following odor trails to a target person despite a number of meteorological factors affecting odor dispersion along the trail. Meteorological data collected along the odor trail revealed that hot temperatures, low humidity, and high wind speeds caused odors to disperse quickly and widely. Under these conditions, dogs searched farther from a person’s original trail. Studying the dogs’ behavior as they followed the odor trail also revealed that dogs were more likely to sample the air for odors when moving at high speeds. Some results suggested that they slow their speed and searched the ground for odors when they lost contact with the odor plume. This study demonstrated the ease in which animals can track odors over great distances to their source.

Chapter Three describes an experiment with humans that examines sniffing behavior modulation during olfactory navigation. In this experiment, humans in an indoor setting varied their sampling behaviors in response to spatial distance from an odorant as well as to stereo olfactory cues. People sniffed less frequently and longer while initially searching for the odor plume. After detecting the odor, people casted, a behavior characterized by zig zagging trajectories, when they were attempting to definitively locate the odor plume. Once the plume was found and they were following it to the source, they sniffed more frequently and had shorter sniffs. Overall, as people approached the source of the odor, sniffing intensity increased. Sniffs were always slower, shorter, and less intense when using stereo olfaction than when one nostril was blocked. Women also detected odor plumes earlier than men. This is the first study to show people modulate their sniffing patterns while navigating an odor plume.

The final experiment in Chapter Four asked whether people have a mental prototype of an odor plume’s structure. People’s movement patterns and sampling behavior were recorded as they interacted with a virtual odor plume in a virtual environment. This study revealed that people treated the virtual plume as if it were a real odor plume and exhibited similar behaviors as in the real world experiment in Chapter Three. People sampled more frequently while they searched for the stimuli, less frequently when they were following the virtual odor, and then increased their sampling frequency again if they lost contact with the stimuli. People were sensitive to as little as a 1% drop in stimulus intensity, to which they reacted by correcting their trajectory to pursue an increasing stimulus gradient instead. They were also able to successfully modulate their sampling behavior when using stereo or non-stereo cues. Overall, men performed better at this task, meaning they were able to estimate the origin of the virtual odor with higher accuracy than women.

This body of work is one of the first to investigate mammalian odor navigation. It shows that mammals have many strategies to modulate their sampling and orientation behavior to account for meteorological factors, loss of stereo cues, and individual differences. The dissertation concludes with a discussion of these three complementary experiment and suggested future studies.