No abstract

SUMMARY Odor plumes are complex, dynamic, three-dimensional structures used by many animals to locate food, mates, home sites, etc. Yet odor itself has no directional properties. Animals use a variety of different senses to obtain directional information. Since most odor plumes are composed of dispersing odor patches and dissipating vorticity eddies, aquatic animals may localize odor sources by simultaneous analysis of chemical and hydrodynamic dispersal fields, a process referred to as eddy chemotaxis. This study examines the contributions of olfaction, mechanoreception and vision to odor source localization in a shark, the smooth dogfish Mustelus canis. Two parallel, turbulent plumes were created in an 8 m flume: squid rinse odor and seawater control. Minimally turbulent `oozing' sources of odor and seawater control were physically separated from sources of major turbulence by placing a brick downstream from each oozing source, creating two turbulent wakes, one or the other flavored with food odor. This created four separate targets for the sharks to locate. Animals were tested under two light conditions(fluorescent and infrared) and in two sensory conditions (lateral line intact and lateral line lesioned by streptomycin). Intact animals demonstrated a preference for the odor plume over the seawater plume and for the source of odor/turbulence (the brick on the odor side) over the source of the odor alone(the odor-oozing nozzle). Plume and target preference and search time were not significantly affected by light condition. In the light, lesioning the lateral line increased search time but did not affect success rate or plume preference. However, lesioned animals no longer discriminated between sources of turbulent and oozing odor. In the dark, search time of lesioned animals further increased, and the few animals that located any of the targets did not discriminate between odor and seawater plumes, let alone targets. These results demonstrate for the first time that sharks require both olfactory and lateral line input for efficient and precise tracking of odor-flavored wakes and that visual input can improve food-finding performance when lateral line information is not available. We distinguish between rheotaxis: orientation to the large-scale flow field (olfaction, vision and superficial lateral line),eddy chemotaxis: tracking the trail of small-scale, odor-flavored turbulence(olfaction and lateral line canals), and pinpointing the source of the plume(lateral line canals and olfaction).

The direction of an odor signal source can be estimated from bilateral differences in signal intensity and/or arrival time. The best-known examples of the use of arrival time differences are in acoustic orientation. For chemoreception, animals are believed to orient by comparing bilateral odor concentration differences, turning toward higher concentrations. However, time differences should not be ignored, because odor plumes show chaotic intermittency, with the concentration variance several orders of magnitude greater than the concentration mean. We presented a small shark species, Mustelus canis, with carefully timed and measured odor pulses directly into their nares. They turned toward the side stimulated first, even with delayed pulses of higher concentration. This is the first conclusive evidence that under seminatural conditions and without training, bilateral time differences trump odor concentration differences. This response would steer the shark into an odor patch each time and thereby enhance its contact with the plume, i.e., a stream of patches. Animals with more widely spaced nares would be able to resolve smaller angles of attack at higher swimming speeds, a feature that may have contributed to the evolution of hammerhead sharks. This constitutes a novel steering algorithm for tracking odor plumes.

The underwater sensory world and the sensory systems of aquatic animals have become better understood in recent decades, but typically have been studied one sense at a time. A comprehensive analysis of multisensory interactions during complex behavioral tasks has remained a subject of discussion without experimental evidence. We set out to generate a general model of multisensory information extraction by aquatic animals. For our model we chose to analyze the hierarchical, integrative, and sometimes alternate use of various sensory systems during the feeding sequence in three species of sharks that differ in sensory anatomy and behavioral ecology. By blocking senses in different combinations, we show that when some of their normal sensory cues were unavailable, sharks were often still capable of successfully detecting, tracking and capturing prey by switching to alternate sensory modalities. While there were significant species differences, odor was generally the first signal detected, leading to upstream swimming and wake tracking. Closer to the prey, as more sensory cues became available, the preferred sensory modalities varied among species, with vision, hydrodynamic imaging, electroreception, and touch being important for orienting to, striking at, and capturing the prey. Experimental deprivation of senses showed how sharks exploit the many signals that comprise their sensory world, each sense coming into play as they provide more accurate information during the behavioral sequence of hunting. The results may be applicable to aquatic hunting in general and, with appropriate modification, to other types of animal behavior.