Jelle Atema - US grants

This is a project to study some of the physical factors involved in sensing of chemicals. The basic issue being approached is how the chemicals that are being sensed are transported to the receptor organs, and how the processes are modified by evolutionary specializations in morphology and behavior. The animal model of chemical sensing that is used is the detection of amino acids by lobster olfactory receptors. In these animals the amino acids, which are feeding stimuli, diffuse to the receptors. In doing so, they finally must diffuse through very thin layers of fluid adjacent to the organs themselves. Different chemoreceptor organs, even in the same animal, can have dramatically different morphology that can result in different diffusion layers. In addition, the animal's behaviors in sampling the environment (flicking and sniffing, for example), affect the boundary layer and also help to minimize the confounding effects of turbulence in the environment. The project will include testing a computational model of receptor organ filtering by measuring and manipulating diffusion layers in different chemoreceptor organs and measuring the time courses of responses in the cells in the living animals. Thus, it will result in an understanding of the structure and function of diffusion layers and of their contribution to analysis of chemical signals. The principles of filtering being studied here in the lobster are important components of all chemical sensing systems. They are little studied and poorly understood, and the conclusions drawn in this study should have wide generalizability to other systems as we..

It has never been possible until very recently to measure the natural distribution patterns of odor in the environment. Dr. Atema and his colleagues have developed tools to make such high- resolution measurements based on the odor-sampling equipment of lobsters. The lobster and the aquatic environment are particularly favorable to carry out this fundamental research on odor-sampling. The present research will try to extract from such odor measurements the salient features that allow animals to track turbulent odor plumes and locate their source. Results from this work will open up our human understanding of an almost unknown sensory world, but one used by many animals, both aquatic and terrestrial. Practical applications can be expected almost immediately in the design of odor-tracking underwater vehicles.

9222774 Atema The way in which sensory receptors process rapid changes of stimulus over time has not been studied much in chemical reception (smell and taste), where the emphasis has been on molecular recognition. Yet complex and dynamic odors characterize the natural environment, and are likely to be very important to how the brain integrates chemosensory signals into appropriate behavior. This work develops novel electrochemical measuring techniques for investigating the spatial and temporal resolution of olfactory receptors on the very accessible model system of the antenna of marine crustaceans. A new tightly-controlled stimulus delivery technology allows a variety of stimulus parameters to be used while measuring the magnitude and reliability of response parameters, looking at effects such as excitation and adaptation patterns. The data will be incorporated into a model of chemosensory dynamics, and extended in studies on receptor development and central coding. This work will have impact beyond chemosensory neurobiology into areas of diagnostic tests for smell and taste performance for the food industry as well as medicine, and to the potential biotechnological development of underwater chemosensing robotic devices for ecology and for fisheries exploitation. ***

9315083 Atema Chemical signals play an important role in the lives of many animals for orientation to food, selection of habitats and mates, and avoidance of predators. It has been shown that benthic animals detect and respond to chemical signals produced by other animals or food sources at distances of centimeters to many meters. This study will test the hypothesize that the micro- scale pattern of concentration distribution of odors in a current provides information that is used by animals for orientation toward the odor source. Laboratory and in situ experimental work will be carried out to describe and quantify the distribution of chemical signals in aquatic odor plumes, and the behavioral responses of crustaceans to those plumes. ***

9315791 Atema Neurobiological and behavioral studies on lobsters have taught us that turbulent (underwater) odor plumes contain directional information based on its patchy fine structure. To measure biologically relevant signals, sensor technology capable of high- resolution (30m, 5ms) chemical signal analysis has been developed. Lobsters locate odor sources efficiently based on bilateral sampling with antennules. Chemoreceptor cells in the antennules are capable of frequency analysis similar to (but slower than) acoustic signal processing. This research tests the hypothesis that bilateral analysis of pulse features is the basis for turbulent chemotaxis. Lobsters, computational models and physical models, i.e., untethered robots, will be employed for this task. Lobster-generated hypotheses will be tested in robots equipped with search paradigms that are pre-tested with computational models. Robot failures to locate efficiently will lead to re-examination and testing of lobsters. Further advances in our knowledge of signal processing in lobsters will be used to update computational models and robot search algorithms. 1. Two small underwater robots will be designed and tested to have the size, mobility and chemical sensing capabilities of a lobster (the benefit vehicle) and a fish (the free swimming vehicle). 2. The vehicles will be used to characterize the spatial and temporal dynamics of underwater chemical plumes in two and three dimensions. 3. The vehicles will be used to test chemo-sensing algorithms proposed for lobsters and fish concentrating on algorithms involved with chemical source localization which is the basic behavior underlying the finding of food and mates. Significant advances in underwater robotics design and in understanding chemosensory signal processing is expected. ***

IBN: 9723542 PI: ATEMA Smell and taste are senses that detect chemicals in the environment, but they do not work in isolation. In the aquatic environment with high viscosity compared to air, chemical "plumes" are carried in currents and can have complex spatial and temporal structure over long distances. Animals in the water are subject to hydrodynamic stimulation by water currents as well, and it is largely unknown how the tactile stimulation from a swift current may affect the chemosensory input, compared to the same chemosensory signal in gently drifting water. This project uses a multidisciplinary approach incorporating hydrodynamics, ultrastructural anatomy and electrophysiology on a crustacean as a model invertebrate chemosensory system. Morphological types of sensory cells in the appendages such as the antenna will be identified, and tested for physiological sensitivity to combinations of chemosensory signals and mechanosensory stimuli. Results will help determine how peripheral integration of multimodal signals might occur, will be important beyond chemosensory science to sensory neuroscience in general, and will have potential applications for technological development of robotic sensory devices.

Odor plumes are generated by turbulent mixing of odor from a source of release with "clean" air or water. The result is a series of "flavored" eddies intermingled with non-flavored eddies. The spatial distribution of these flavored eddies contains a statistically reliable gradient of information that animals may use to locate distant odor sources. Therefore, they need sensors whose filtering capabilities enhance those signal parameters that provide the best spatial gradients. Both chemical and hydrodynamic dispersal patterns must be considered.

Previous NSF funding supported experiments from which it was determined that for chemical gradients the concentration slopes of odor patches provide good directional information. Neither the equivalent hydrodynamic information nor to which degree these two signal features are actually used by animals and under which circumstances are known. The primary sensory processes of chemo-hydrodynamic signal detection will be studied. The lobster and its well-studied lateral antennule will be used as the model of choice for linking primary sensory information in both modalities.

Electrophysiological recording of action potentials from the antennular nerve, which carries chemical and mechanical information to the brain, will be used. The excised antennule will be exposed to independently controlled chemical and hydrodynamic stimuli with high-resolution measurement of both. The chemical response properties of the antennule are relatively well known. Its hydrodynamic responsiveness now will be determined and the possibility of peripheral bi-modal signal processing will be explored. Any form of bi-modal chemo-hydrodynamic interaction would have consequences for the detection of "flavored eddies". Invertebrates often appear to have hard-wired sensory filters in their peripheral sensory systems.

The results of this work may find application in the design of sensors and algorithms for automated odor plume tracing in such tasks as detection and localization of sources of chemical pollution, both under water and in air.

The larvae of most reef fishes disperse from the natal reef soon after hatching and return to the reef environment after a pelagic stage of several days to weeks. Ocean currents were assumed to disperse larvae widely, resulting in broad genetic connectivity between reef populations. In some recently studied cases, connectivity appears to be more limited and self-recruitment (settlement on natal reefs) more common than previously assumed. This may be the result of favorable hydrographic conditions, perhaps aided by a behavioral component of larvae using swimming and sensory capabilities to stay in proximity to a reef by using appropriate currents. Preliminary data on the genetic structure of the cardinalfish Apogon doederleini show significant differences between reef populations at the geographical scale of 7 to 18 km. Genetic differences between reefs using a hydrographic model would have predicted panmixis. This suggests that larval behavior plays an active role in retention and self-recruitment. This project will determine the genetic population structure of three reef fish species with different larval biology and study mechanisms of dispersal and recruitment behavior. The objectives are: (1) Determine in two consecutive years the degree of genetic substructure and its temporal stability of fish within and among reef clusters at spatial scales of approx. 10 to 20 km based on three species with different dispersal biology. (2a) Determine whether pre-settlement juveniles of the three species differentiate between odors of hydrographically more or less connected reefs and (b) Determine the temporal stability of their reef odor preference. (3) Extend the hydrographic dispersal model for different source populations among the proposed reef clusters and compare the results with connectivity results from genetic analysis and odor preference tests. For genetic analysis, geographic distances and dispersal links and barriers will be studied using DNA microsatellite markers, centered around One Tree Island in the Capricorn/Bunker group of reefs in the Great Barrier Reef, Australia. The hydrographic model of the OTI region will be employed to predict dispersal by advection. Genetic substructure and reef odor preference in settling larvae/juveniles of three species will be compared: Acanthochromis polyacanthus (no pelagic stage), (b) Apogon doederleini (pelagic, weak swimmers) and (c) Pomacentrus coelestis (pelagic, strong swimmers). A mini-flume for odor choice tests will be used on-board ships in which single settlement stage fish have shown significant odor preference for specific reefs.
Intellectual merits are demonstrations of 1) the scale of genetic substructure in coral reef fishes impacting concepts about the scale of reef connectivity, 2) differences between hydrographic dispersal predictions and genetic population structure, 3) olfactory capabilities of reef fish larvae to discriminate between odors of closely linked reefs, 4) possible connections between odor preference and genetic population structure suggestive of early imprinting and retention near natal reefs. Broader impacts are in 1) coral reef conservation and management (connectivity, water quality), 2) training of the next generation of scientists in biological oceanography, behavioral ecology, evolution and (marine) conservation, 3) international collaboration.

Understanding how animals navigate under water is not only fascinating in its own right, it also contributes to instrumentation designs of underwater vehicles, robots and surface vessels, impacts management of fisheries, and helps protect the marine environment. Sharks have been chosen to demonstrate how they navigate. While they can not detect a drop of blood a mile away, as often stated, sharks do have impressive prey tracking capabilities. Sharks are important in fisheries worldwide and have been severely depleted in recent decades, often taken as unwanted by-catch in other fisheries. Yet, they are essential top predators needed to maintaining a healthy ecosystem. This research project will show how sharks use all their senses in hunting behavior, starting with initial prey detection, through tracking and locating, and ending with striking their prey. For more complete understanding, we compare a few shark species that appear to use their senses differently mostly because they specialize in different prey in different habitats. A team of experts in sensory and shark biology, using unique testing facilities in Massachusetts and Florida, has been assembled including graduate students being trained in the many technical approaches needed for work on live sharks. The research directly involves undergraduate and high school students and provides extensive outreach to other students of all ages and to the public in general. The accumulated knowledge should lead to a model of shark navigation and predation that can be used for the conservation of sharks, protection of humans, and the engineering design of underwater steering algorithms. The inevitable presentation of this research in future television programs and video documentaries will disseminate new knowledge to the public at large, both of sharks and of rigorous science.

Understanding how far young fish move away from their parents is a major goal of marine ecology because this dispersal can make connections between distinct populations and thus influence population size and dynamics. Understanding the drivers of population dynamics is, in turn, essential for effective fisheries management. Marine ecologists have used two different approaches to understand how fish populations are connected: genetic methods that measure connectivity and oceanographic models that predict connectivity. There is, however, a mismatch between the predictions of oceanographic models and the observations of genetic methods. It is thought that this mismatch is caused by the behavior of the young, or larval, fish. The objective of this research is to study the orientation capabilities of larval fish in the wild throughout development and under a variety of environmental conditions to see if the gap between observations and predictions of population connectivity can be resolved. The project will have broader impacts in three key areas: integration of research and teaching by training young scientists at multiple levels; broadening participation of undergraduates from underrepresented groups; and wide dissemination of results through development of a website with information and resources in English and Spanish.

The overall objective of the research is to investigate the role of larval orientation behavior throughout ontogeny in determining population connectivity. This will be done using the neon goby, Elacatinus lori, as a model system in Belize. The choice of study system is motivated by the fact that direct genetic methods have already been used to describe the complete dispersal kernel for this species, and these observations indicate that dispersal is less extensive than predicted by a high-resolution biophysical model; E. lori can be reared in the lab from hatching to settlement providing a reliable source of larvae of all ages for proposed experiments; and a new, proven behavioral observation platform, the Drifting In Situ Chamber (DISC), allows measurements of larval orientation behavior in open water. The project has three specific objectives: to understand ontogenetic changes in larval orientation capabilities by correlating larval orientation behavior with developmental sensory anatomy; to analyze variation in the precision of larval orientation in different environmental contexts through ontogeny; and to test alternative hypotheses for the goal of larval orientation behavior, i.e., to determine where larvae are heading as they develop.