On the usefulness of bilateral comparison to tracking turbulent chemical odor plumes

This article reports on the usefulness of bilateral comparison of chemosensory information to an animal or agent tracking an odor to its source. Instantaneous concentration fields of a chemical plume diffusing in a fully developed, turbulent, open channel flow are measured using planar laser-induced fluorescence. The plume is released isokinetically 25 mm above the smooth bed (z 90), thus transport is mainly due to advection and ambient turbulence. A spatial cross-correlation function in the spanwise direction gives a strong indication of the relative position of the centerline and distance from the source. The relative direction of the plume centerline can be estimated from the instantaneous concentration, provided the sensors are separated by a distance larger than the spanwise integral length scale of the concentration field. Many aquatic and terrestrial animals rely on sensory cues to track turbulent odor plumes in order to locate food and mates (Vickers 2000; Zimmer and Butman 2000). Very long time records would be required to accurately determine traditional statistical measures of odor concentration, such as time-averaged concentration and variance (Elkinton et al. 1984; Moore and Atema 1991; Murlis et al. 1992). But it has been observed that animals, such as blue crabs, do not wait long enough at a location to accurately assess these time-averaged quantities and their small spatial variations. Thus, these animals must be using instantaneous observations of the odor plume to make tracking decisions. The instantaneous spatial and temporal structure of a turbulent odor plume is complex and depends on transport within the flow. Measurements of chemical plumes released into turbulent boundary layers in the laboratory (Fackrell and Robins 1982; Bara et al. 1992) and field (Murlis and Jones 1981; Jones 1983; Murlis 1986; Hanna and Insley 1989) show highly intermittent concentration time records. Time-averaged properties, such as the mean and variance of concentration, are typically calculated from the time records and compared to models based on turbulent diffusion. Although these data and models may be useful for predicting time-averaged distributions, they provide no information about instantaneous quantities that may be useful to an animal tracking the plume. There is evidence that rapidly moving foragers are not using sequential sampling to acquire orientation cues (Webster and Weissburg 2001); thus, our attention now turns to the ‘‘information content’’ in the instantaneous spatial distribution of odor concentration. Little evidence is available about the nature of this information content. It should be pointed out, however, that benthic crustaceans are rarely successful in locating the odor source in the absence of mean flow and turbulent transport (Weissburg and Zimmer-Faust 1994). Thus, measurements of the fine-scale spatial distriAcknowledgments The authors gratefully acknowledge the valuable contribution of Philip Roberts to this project, helpful discussion with Marc Weissburg, and the financial support of ONR/DARPA under the direction of Keith Ward and Regina Dugan. bution of a chemical tracer are needed to evaluate the usefulness of spatial sampling. In this study, tracer concentration fields of a plume diffusing in a turbulent open channel flow were measured using the planar laser-induced fluorescence (PLIF) technique. With the rapid development of modern lasers, digital cameras, and computer acquisition systems, long sequences of high-resolution concentration fields can now be obtained and stored. Information on the instantaneous concentration fields and spatial correlations that were previously unattainable can be readily obtained. The objective of this article is to report on the value of a sensory cue, namely bilateral comparison, available to animals such as blue crabs in a turbulent odor plume.