Turbulent Odor Plumes Research Articles

Dynamics of odor-source localization: Insights from real-time odor plume recordings and head-motion tracking in freely moving mice.

Animals navigating turbulent odor plumes exhibit a rich variety of behaviors, and employ efficient strategies to locate odor sources. A growing body of literature has started to probe this complex task of localizing airborne odor sources in walking mammals to further our understanding of neural encoding and decoding of naturalistic sensory stimuli. However, correlating the intermittent olfactory information with behavior has remained a long-standing challenge due to the stochastic nature of the odor stimulus. We recently reported a method to record real-time olfactory information available to freely moving mice during odor-guided navigation, hence overcoming that challenge. Here we combine our odor-recording method with head-motion tracking to establish correlations between plume encounters and head movements. We show that mice exhibit robust head-pitch motions in the 5-14Hz range during an odor-guided navigation task, and that these head motions are modulated by plume encounters. Furthermore, mice reduce their angles with respect to the source upon plume contact. Head motions may thus be an important part of the sensorimotor behavioral repertoire during naturalistic odor-source localization.

Dynamics of odor-source localization: Insights from real-time odor plume recordings and head-motion tracking in freely moving mice.

Animals navigating turbulent odor plumes exhibit a rich variety of behaviors, and employ efficient strategies to locate odor sources. A growing body of literature has started to probe this complex task of localizing airborne odor sources in walking mammals to further our understanding of neural encoding and decoding of naturalistic sensory stimuli. However, correlating the intermittent olfactory information with behavior has remained a long-standing challenge due to the stochastic nature of the odor stimulus. We recently reported a method to record real-time olfactory information available to freely moving mice during odor-guided navigation, hence overcoming that challenge. Here we combine our odor-recording method with head-motion tracking to establish correlations between plume encounters and head movements. We show that mice exhibit robust head-pitch motions in the 5-14Hz range during an odor-guided navigation task, and that these head motions are modulated by plume encounters. Furthermore, mice reduce their angles with respect to the source upon plume contact. Head motions may thus be an important part of the sensorimotor behavioral repertoire during naturalistic odor-source localization.

A reinforcement learning fuzzy system for continuous control in robotic odor plume tracking

AbstractIn dynamic outdoor environments characterized by turbulent airflow and intermittent odor plumes, robotic odor plume tracking remains challenging, because existing algorithms heavily rely on manually tuning or learning from expert experience, which are hard to implement in an unknown environment. In this paper, a multi-continuous-output Takagi–Sugeno–Kang fuzzy system was designed and tuned with reinforcement learning to solve the robotic odor source localization problem in dynamic odor plumes. Based on the Lévy Taxis plume tracking controller, the proposed fuzzy system determined the parameters of the controller based on the robot’s observation and guided the robot to turn and move towards the odor source at each searching step. The trained fuzzy system was tested in simulated filament-based odor plumes dispersed by a changing wind field. The results showed that the performance of the proposed fuzzy system-based controller trained with reinforcement learning can achieve a similar success rate and higher efficiency compared with a manually tuned and well-designed fuzzy system-based controller. The fuzzy system-based plume tracking controller was also validated through real robotic experiments.

Learning to predict target location with turbulent odor plumes.

Animal behavior and neural recordings show that the brain is able to measure both the intensity and the timing of odor encounters. However, whether intensity or timing of odor detections is more informative for olfactory-driven behavior is not understood. To tackle this question, we consider the problem of locating a target using the odor it releases. We ask whether the position of a target is best predicted by measures of timing vs intensity of its odor, sampled for a short period of time. To answer this question, we feed data from accurate numerical simulations of odor transport to machine learning algorithms that learn how to connect odor to target location. We find that both intensity and timing can separately predict target location even from a distance of several meters; however, their efficacy varies with the dilution of the odor in space. Thus, organisms that use olfaction from different ranges may have to switch among different modalities. This has implications on how the brain should represent odors as the target is approached. We demonstrate simple strategies to improve accuracy and robustness of the prediction by modifying odor sampling and appropriately combining distinct measures together. To test the predictions, animal behavior and odor representation should be monitored as the animal moves relative to the target, or in virtual conditions that mimic concentrated vs dilute environments.

Active sensing in a dynamic olfactory world

This Perspective highlights the shift from the classic picture of olfaction as slow and static to a view in which dynamics play a critical role at many levels of sensing and behavior. Olfaction is now increasingly seen as a “wide-bandwidth temporal sense” (Ackels et al., 2021; Nagel et al., 2015). A parallel transition is occurring in odor-guided robot navigation, where it has been discovered that sensors can access temporal cues useful for navigation (Schmuker et al., 2016). We are only beginning to understand the implications of this paradigm-shift on our view of olfactory and olfacto-motor circuits. Below we review insights into the information encoded in turbulent odor plumes and shine light on how animals could access this information. We suggest that a key challenge for olfactory neuroscience is to re-interpret work based on static stimuli in the context of natural odor dynamics and actively exploring animals.

Plume Dynamics Structure the Spatiotemporal Activity of Mitral/Tufted Cell Networks in the Mouse Olfactory Bulb.

Although mice locate resources using turbulent airborne odor plumes, the stochasticity and intermittency of fluctuating plumes create challenges for interpreting odor cues in natural environments. Population activity within the olfactory bulb (OB) is thought to process this complex spatial and temporal information, but how plume dynamics impact odor representation in this early stage of the mouse olfactory system is unknown. Limitations in odor detection technology have made it difficult to measure plume fluctuations while simultaneously recording from the mouse's brain. Thus, previous studies have measured OB activity following controlled odor pulses of varying profiles or frequencies, but this approach only captures a subset of features found within olfactory plumes. Adequately sampling this feature space is difficult given a lack of knowledge regarding which features the brain extracts during exposure to natural olfactory scenes. Here we measured OB responses to naturally fluctuating odor plumes using a miniature, adapted odor sensor combined with wide-field GCaMP6f signaling from the dendrites of mitral and tufted (MT) cells imaged in olfactory glomeruli of head-fixed mice. We precisely tracked plume dynamics and imaged glomerular responses to this fluctuating input, while varying flow conditions across a range of ethologically-relevant values. We found that a consistent portion of MT activity in glomeruli follows odor concentration dynamics, and the strongest responding glomeruli are the best at following fluctuations within odor plumes. Further, the reliability and average response magnitude of glomerular populations of MT cells are affected by the flow condition in which the animal samples the plume, with the fidelity of plume following by MT cells increasing in conditions of higher flow velocity where odor dynamics result in intermittent whiffs of stronger concentration. Thus, the flow environment in which an animal encounters an odor has a large-scale impact on the temporal representation of an odor plume in the OB. Additionally, across flow conditions odor dynamics are a major driver of activity in many glomerular networks. Taken together, these data demonstrate that plume dynamics structure olfactory representations in the first stage of odor processing in the mouse olfactory system.

Spatial, but not temporal, aspects of orientation are controlled by the fine-scale distribution of chemical cues in turbulent odor plumes.

Orientation within turbulent odor plumes occurs across a vast range of spatial and temporal scales. From salmon homing across featureless oceans to microbes forming reproductive spores, the extraction of spatial and temporal information from chemical cues is a common sensory phenomenon. Yet, given the difficulty of quantifying chemical cues at the spatial and temporal scales used by organisms, discovering what aspects of chemical cues control orientation behavior has remained elusive. In this study, we placed electrochemical sensors on the carapace of orienting crayfish and measured, with fast temporal rates and small spatial scales, the concentration fluctuations arriving at the olfactory appendages during orientation. Our results show that the spatial aspects of orientation (turning and heading angles) are controlled by the temporal aspects of odor cues.

Spatial odor discrimination in the hawkmoth, Manduca sexta (L.)

ABSTRACTFlying insects track turbulent odor plumes to find mates, food and egg-laying sites. To maintain contact with the plume, insects are thought to adapt their flight control according to the distribution of odor in the plume using the timing of odor onsets and intervals between odor encounters. Although timing cues are important, few studies have addressed whether insects are capable of deriving spatial information about odor distribution from bilateral comparisons between their antennae in flight. The proboscis extension reflex (PER) associative learning protocol, originally developed to study odor learning in honeybees, was used as a tool to ask if hawkmoths, Manduca sexta, can discriminate between odor stimuli arriving on either antenna. We show moths discriminated the odor arrival side with an accuracy of >70%. Information about spatial distribution of odor stimuli may be available to moths searching for odor sources, opening the possibility that they use both spatial and temporal odor information.This article has an associated First Person interview with the first author of the paper.

Identification of Exploration and Exploitation Balance in the Silkmoth Olfactory Search Behavior by Information-Theoretic Modeling.

Insects search for and find odor sources as their basic behaviors, such as when looking for food or a mate. This has motivated research to describe how they achieve such behavior under turbulent odor plumes with a small number of neurons. Among different insects, the silk moth has been studied owing to its clear motor response to olfactory input. In past studies, the “programmed behavior” of the silk moth has been modeled as the average duration of a sequence of maneuvers based on the duration of periods without odor hits. However, this model does not fully represent the fine variations in their behavior. In this study, we used silk moth olfactory search trajectories from an experimental virtual reality device. We achieved an accurate input by using optogenetic silk moths that react to blue light. We then modeled such trajectories as a probabilistic learning agent with a belief of possible source locations. We found that maneuvers mismatching the programmed behavior are related to larger entropy decrease, that is, they are more likely to increase the certainty of the belief. This implies that silkmoths include some stochasticity in their search policy to balance the exploration and exploitation of olfactory information by matching or mismatching the programmed behavior model. We believe that this information-theoretic representation of insect behavior is important for the future implementation of olfactory searches in artificial agents such as robots.

A spiking neural program for sensorimotor control during foraging in flying insects

Foraging is a vital behavioral task for living organisms. Behavioral strategies and abstract mathematical models thereof have been described in detail for various species. To explore the link between underlying neural circuits and computational principles, we present how a biologically detailed neural circuit model of the insect mushroom body implements sensory processing, learning, and motor control. We focus on cast and surge strategies employed by flying insects when foraging within turbulent odor plumes. Using a spike-based plasticity rule, the model rapidly learns to associate individual olfactory sensory cues paired with food in a classical conditioning paradigm. We show that, without retraining, the system dynamically recalls memories to detect relevant cues in complex sensory scenes. Accumulation of this sensory evidence on short time scales generates cast-and-surge motor commands. Our generic systems approach predicts that population sparseness facilitates learning, while temporal sparseness is required for dynamic memory recall and precise behavioral control. Our work successfully combines biological computational principles with spike-based machine learning. It shows how knowledge transfer from static to arbitrary complex dynamic conditions can be achieved by foraging insects and may serve as inspiration for agent-based machine learning.

Moth-inspired navigation algorithm in a turbulent odor plume from a pulsating source.

Some female moths attract male moths by emitting series of pulses of pheromone filaments propagating downwind. The turbulent nature of the wind creates a complex flow environment, and causes the filaments to propagate in the form of patches with varying concentration distributions. Inspired by moth navigation capabilities, we propose a navigation strategy that enables a flier to locate an upwind pulsating odor source in a windy environment using a single threshold-based detection sensor. This optomotor anemotaxis strategy is constructed based on the physical properties of the turbulent flow carrying discrete puffs of odor and does not involve learning, memory, complex decision making or statistical methods. We suggest that in turbulent plumes from a pulsating point source, an instantaneously measurable quantity referred as a “puff crossing time”, improves the success rate as compared to the navigation strategies based on temporally regular zigzags due to intermittent contact, or an “internal counter”, that do not use this information. Using computer simulations of fliers navigating in turbulent plumes of the pulsating point source for varying flow parameters such as turbulent intensities, plume meandering and wind gusts, we obtained statistics of navigation paths towards the pheromone sources. We quantified the probability of a successful navigation as well as the flight parameters such as the time spent searching and the total flight time, with respect to different turbulent intensities, meandering or gusts. The concepts learned using this model may help to design odor-based navigation of miniature airborne autonomous vehicles.

High Precision of Spike Timing across Olfactory Receptor Neurons Allows Rapid Odor Coding in Drosophila

SummaryIn recent years, it has become evident that olfaction is a fast sense, and millisecond short differences in stimulus onsets are used by animals to analyze their olfactory environment. In contrast, olfactory receptor neurons are thought to be relatively slow and temporally imprecise. These observations have led to a conundrum: how, then, can an animal resolve fast stimulus dynamics and smell with high temporal acuity? Using parallel recordings from olfactory receptor neurons in Drosophila, we found hitherto unknown fast and temporally precise odorant-evoked spike responses, with first spike latencies (relative to odorant arrival) down to 3 ms and with a SD below 1 ms. These data provide new upper bounds for the speed of olfactory processing and suggest that the insect olfactory system could use the precise spike timing for olfactory coding and computation, which can explain insects' rapid processing of temporal stimuli when encountering turbulent odor plumes.

Odor-Driven Spike Synchrony Across Olfactory Receptor Neurons Allows Rapid Odor Identification in Drosophila

In recent years, it has become evident that olfaction is a fast sense, and millisecond short differences in stimulus onsets are used by animals to analyze their olfactory environment. In contrast, olfactory receptor neurons are thought to be relatively slow and temporally imprecise. These observations have led to a conundrum: How, then, can an animal resolve fast stimulus dynamics and smell with high temporal acuity? Using parallel recordings from olfactory receptor neurons in Drosophila, we found hitherto unknown fast and temporally precise odorant‐evoked spike responses, with first spike latencies as short as 3 ms and with a standard deviation (jitter) well below 1 ms. These data provide new upper bounds for the speed of olfactory processing and suggest that the insect olfactory system can use the precise spike timing for olfactory coding and computation, which can explain insects’ rapid processing of temporal stimuli when encountering turbulent odor plumes.

Tracking Odorant Plumes.

Chemical signals are primarily distributed throughout aquatic environments by processes that are affected by turbulence. Turbulence continually stirs and mixes chemical odorants into complex, filamentous structures that are sampled by organisms. These odorant signals are critical for survival and/or reproductive success of most aquatic animals, and the time varying spatial structure of velocity and concentration offers valuable guidance cues while navigating in a plume. Two separate techniques are described to simultaneously measure a turbulent odor plume on a scale relevant to the chemosensors and mechanosensors located along the antennules of aquatic organisms. The first, planar laser induced fluorescence (PLIF), is used to quantify odorant concentrations, while the second, particle image velocimetry (PIV), is used to measure turbulent fluid velocities.

The Rate of Concentration Change and How It Determines the Resolving Power of Olfactory Receptor Neurons.

The response characteristics of olfactory receptor neurons (ORNs) and their corollary, the differential sensitivity and the resolving power, are fundamental to understand olfactory coding and the information extracted from a fluctuating olfactory signal. Previous work has focused on the temporal resolution of odor pulses presented for very brief periods at varying concentrations. The time course of the odor pulses as a stimulus parameter has not been considered. The present study investigated the precision of the ON and OFF ORNs on the antennae of the cockroach to discriminate increments and decrements of continuously rising and falling odor concentrations. Stimulation consisted of ramp-like upward and downward concentration changes in a trapezoid fashion. By varying ramp steepness, we examined the effect of the rate of concentration change. Both ORNs were clearly dependent on continuously rising and falling odor concentrations. As the rate of upward and downward concentration changes increases, differential sensitivity improves. Since the scatter of responses around the stimulus-response functions also increases, the resolving power for concentration increments and decrements deteriorates. Thus, the slower the rate of concentration change, the higher the precision in differentiating small concentration changes. Intuitively, the inverse relationship between the rate of concentration change and the resolving power is not surprising because accuracy requires time. A high degree of precision at slow concentration rates enables the cockroach to use information about the onset and offset slopes of odor pulses in addition to the pulse height to encode the spatial-temporal structure of turbulent odor plumes.

Additional Navigational Strategies Can Augment Odor-Gated Rheotaxis for Navigation under Conditions of Variable Flow.

The navigation strategies animals use to find sources of odor depend on the olfactory stimuli, the properties of flowing fluids, and the locomotory capabilities of the animal. In high Reynolds number environments, animals typically use odor-gated rheotaxis to find the source of turbulent odor plumes. This strategy succeeds because, although turbulence creates an intermittent chemical cue, the animal follows the (continuous) directional cue created by the flow that is transporting the chemical. However, in nature, animals may lose all contact with an odor plume as variations in the direction of bulk flow cause the plume to be rotated away before the animal reaches the source of the odor. Our goal was to use a mathematical model to test the hypothesis that strategies that augment odor-gated rheotaxis would be beneficial for finding the source of an odor plume in such variable flow. The model links a stochastic variable-direction odor plume with a turbulence-based intermittent chemical signal and four different movement strategies, including: odor-gated rheotaxis alone (as a control), odor-gated rheotaxis augmented by further rheotaxis in the absence of odor, odor-gated rheotaxis augmented by a random walk, and odor-gated rheotaxis augmented by movement actively guided by the heading of the flow when the odor was still present. We found that any of the three augmented strategies could improve on strict odor-gated rheotaxis. Moreover, variations in performance caused the best strategy to depend on the speed of movement of the animal and the magnitude of the variation in flow, and more subtly on the duration over which the augmented strategy was performed. For most combinations of parameters in the model, either augmenting with a random walk or following the last-known heading were the best-performing strategies. Overall, our results suggest that marine animals that rely on odor cues to navigate in turbulent environments may augment odor-gated rheotaxis with additional movements that will increase the probability of finding the sources of odors. Moreover, we believe our approach to modeling odor plumes in variable flows is a valuable step toward mathematically capturing the key conditions experienced by animals navigating on the basis of odors carried by flows.

Evolutionary robotics approach to odor source localization

An Evolutionary Robotics (ER) approach to the task of odor source localization is investigated. In particular, Continuous Time Recurrent Neural Networks (CTRNNs) are evolved for odor source localization in simulated turbulent odor plumes. In the experiments, the simulated robot is equipped with a single chemical sensor and a wind direction sensor. Three main contributions are made. First, it is shown that the ER approach can be successfully applied to odor source localization in both low-turbulent and high-turbulent conditions. Second, it is demonstrated that a small neural network is able to successfully perform all three sub-tasks of odor source localization: (i) finding the odor plume, (ii) moving toward the odor source, and (iii) identifying the odor source. Third, the analysis of the evolved behaviors reveals two novel odor source localization strategies. These strategies are successfully re-implemented as finite state machines, validating the insights from the analysis of the neural controllers.

The impact of odor and ambient flow speed on the kinematics of the crayfish antennular flick: implications for sampling turbulent odor plumes

Before organisms extract relevant information from olfactory cues, animals often actively sample the environment through sniffing, wing beating, or flicking of olfactory appendages. These processes serve to increase the local flow velocity and to facilitate the movement of chemical signals from the environment to receptor sites. Crustaceans flick their antennules to sample. While the kinematics of this process is beginning to be understood, modulation of antennular flicking in response to increases in ambient flow velocity or through stimulation from chemicals is unknown at this time. This manuscript details two studies that were performed to understand the kinematics on this behavior and their implications on sensory perception. In the first study, the kinematics of a flick from a restrained crayfish was analyzed in response to flow velocities of 1, 3, and 5 cm/s and in the presence and absence of odor at these three velocities. The second study sought to determine whether antennular flicking behavior changes as a function of crayfish’s walking speed or in response to variations in odor source concentration. The first study clearly shows that flick kinematic parameters (flick acceleration, flick velocity, and flick angle deflection) are increased as the ambient flow is increased and are increased in response to chemical stimulation. The second study shows that bilateral antennular flicking is not performed during chemical orientation and that antennular flicking behavior does change as a function of walking speed of the crayfish or increases in source concentration.

The spatial and temporal patterns of odors sampled by lobsters and crabs in a turbulent plume

Odors are dispersed across aquatic habitats by turbulent water flow as filamentous, intermittent plumes. Many crustaceans sniff (take discrete samples of ambient water and the odors it carries) by flicking their olfactory antennules. We used planar laser-induced fluorescence to investigate how flicking antennules of different morphologies (long antennules of spiny lobsters, Panulirus argus; short antennules of blue crabs, Callinectes sapidus) sample fluctuating odor signals at different positions in a turbulent odor plume in a flume to determine whether the patterns of concentrations captured can provide information about an animal's position relative to the odor source. Lobster antennules intercept odors during a greater percentage of flicks and encounter higher peak concentrations than do crab antennules, but because crabs flick at higher frequency, the duration of odor-free gaps between encountered odor pulses is similar. For flicking antennules there were longer time gaps between odor encounters as the downstream distance to the odor source decreases, but shorter gaps along the plume centerline than near the edge. In contrast to the case for antennule flicking, almost all odor-free gaps were <500 ms at all positions in the plume if concentration was measured continuously at the same height as the antennules. Variance in concentration is lower and mean concentration is greater near the substratum, where leg chemosensors continuously sample the plume, than in the water where antennules sniff. Concentrations sampled by legs increase as an animal nears an odor source, but decrease for antennules. Both legs and antennules encounter higher concentrations near the centerline than at the edge of the plume.

Detection and Discrimination of Mixed Odor Strands in Overlapping Plumes Using an Insect-Antenna-Based Chemosensor System

Olfactory signals, a major means of communication in insects, travel in the form of turbulent odor plumes. In terrestrial environments, an odor blend emitted from a single point source exists in every strand of the plume, whereas, in confluent plumes from two different odor sources, the strands have some chance of being coincident and comprising a new third odor in those strands. Insects have the ability to detect and interpret necessary olfactory information from individual filamentous odor strands in complex multifilament odor plumes. However, behaviorists have had no way to measure the stimulus situations they are presenting to their temporally acute insect subjects when performing Y-tube olfactometer or confluent pheromone plume wind tunnel assays. We have successfully measured the degree of plume-strand mixing in confluent plumes in a wind tunnel by using a multichannel insect-antenna-based chemosensor. A PC-based computer algorithm to analyze antennal signals from the probe portion of the system performed real-time signal processing and, following a short training session, classified individual odorant/mixture strands at sub-second temporal resolution and a few tens of millimeters of spatial resolution. In our studies, the chemosensor classified a higher frequency of strands of two different odorants emitted from two closely spaced filter papers as being "mixed" when the sources were located only 1 or 2 cm apart than when the sources were 5 or 10 cm apart. These experiments demonstrate the chemosensor's potential to be used for measuring odor stimulus situations in more complex multiple-plume environments.