Conditioned Alarm Behavior in Fathead Minnows (Pimephales promelas) Resulting from Association of Chemical Alarm Pheromone with a Nonbiological Visual Stimulus

Fathead minnows (Pimephales promelas) adopt antipredator (alarm) behavior when they detect alarm pheromone released from an injured conspecific. This is an adaptive response since alarm pheromone is generally released only in the context of a predation event. Alarm reactions may also occur in response to chemical and visual stimuli that minnows learn to associate with release of alarm pheromone. Here, we tested if fathead minnows can learn to associate a nonbiological, visual stimulus with predation risk. Minnows were simultaneously exposed to red light and conspecific alarm pheromone, inducing an alarm reaction. When retested using red light alone, small shoals of minnows displayed an antipredator response: dashing movements and disorganized swimming followed by decreased height in the water column and increased shoal cohesion. This resulted from a single-trial exposure to the combined cues and demonstrates a robust ecological mechanism by which minnows learn to recognize indicators of predation risk that may vary in space and time. However, learning to associate risk with biologically irrelevant stimuli may be an ecological liability. How minnows discern between relevant and irrelevant stimuli in nature is not known.     

Learned Recognition of Predation Risk by Enallagma Damselfly Larvae (Odonata, Zygoptera) on the Basis of Chemical Cues

We studied two populations of damselfly larvae (Enallagma boreale): one population cooccurred with a predatory fish (northern pike, Esox lucius); the other did not. Damselflies that cooccurred with pike adopted antipredator behavior (reduced activity) in response to chemical stimuli from injured conspecifics, and to chemical stimuli from pike, relative to a distilled water control. Damselflies from an area where pike do not occur responded only to chemical stimuli from injured conspecifics. In a second set of experiments, we conditioned pike-naive damselflies to recognize and respond to chemical stimuli from pike with antipredator behavior. Damselfly larvae that were previously unresponsive to pike stimuli learned to recognize pike stimuli after a single exposure to stimuli from pike and injured damselflies or pike and injured fathead minnows (Pimephales promelas). The response to injured fathead minnows was not a general response to injured fish because damselfly larvae did not respond to chemical stimuli from injured swordtails (Xiphophorus helleri), an allopatric fish. Taken together, these data suggest a flexible learning program that allows damselfly larvae to rapidly acquire the ability to recognize local predation risk based on chemical stimuli from predators, conspecifics, and heterospecific members of their prey guild.     

Field Verification of Predator Attraction to Minnow Alarm Substance

Fishes such as minnows in the superorder Ostariophysi possess specialized alarm substance cells (ASC) that contain an alarm cue. Alarm substance can only be released by damage to the epidermis; thus, the release of alarm substance is a reliable indicator of predation risk. When nearby minnows detect the cue, they adopt a range of antipredator behaviors that reduce their probability of predation. Predator–predator interactions afford prey an opportunity to escape and, thus, a fitness benefit that maintains alarm substance calls over evolutionary time. Here, we present data from a simple field experiment verifying that nearby predators are attracted to minnow alarm substance because it signals an opportunity to pirate a meal. Fishing lures were baited with sponge blocks scented with either (1) water (control for sponge odor and appearance), (2) skin extract from non-ostariophysan convict cichlids (superorder Acanthopterygii, Archocentrus Cichlasoma nigrofasciatus) to control for general injury-released cues from fish, or (3) skin extract from fathead minnows (superorder Ostariophysi, Pimephales promelas). Predator strike frequency on each sponge type was 1,1, and 7 for water, cichlid, and minnow cues, respectively. These data provide the first field test using fish predators of the predator-attraction hypothesis for the evolution of Ostariophysan alarm substance cells.     

The role of olfaction in chemosensory-based predator recognition in the fathead minnow,Pimephales promelas

Solitary fathead minnows (Pimephales promelas) were rendered anosmic and exposed to chemical stimuli from a predatory northern pike (Esox lucius) to determine the role of olfaction in the minnow's ability to recognize predators on the basis of chemical stimuli. Anosmic fish did not respond to the pike stimuli with a typical fright reaction, while control fish, with intact olfactory receptors, did. These results demonstrate that the olfactory system is necessary for the ability of fathead minnows to recognize northern pike as a predator and that the gustatory and single-celled chemosensory systems are not sufficient for this recognition in the absence of olfactory input. Olfactory impairment was behaviorally confirmed by exposing minnows to alarm substance (Schreckstoff).     

Effects of diet on localized defecation by Northern Pike,Esox lucius

Fathead minnows (Pimephales promelas) are able to detect conspecific alarm pheromone in the feces of northern pike (Esox lucius) and have been shown to avoid areas labeled with the feces of pike that were fed minnows. The minnows did not avoid areas labeled with the feces of pike that were fed swordtails (Xiphophorous helleri), which lack ostariophysan alarm pheromone. In laboratory experiments, pike fed a diet of minnows localized their defecation away from their foraging area. It has been suggested that in doing so, pike may remove chemical cues that label their foraging area as dangerous to prey species. As yet there has been no conclusive evidence to support this hypothesis. In this experiment, we test the effects of different predator diets on localized defecation by pike. Pike were fed minnows, swordtails, or mice (Mus musculus). Swordtails and mice lack ostariophysan alarm pheromones. Area use and location of feces were recorded. Pike fed minnows spent significantly more time in the home area (i.e., area of the test tank where they were fed) and defecated significantly more often in the opposite end of the tank. Pike fed swordtails also exhibited a significant preference for the home area in area use, while those fed mice showed no such preference. When fed either swordtails or mice, there was no significant difference between the proportion of time spent and proportion of feces in each area of the test tank. These data suggest that pike localize their defecation only when consuming prey items containing alarm pheromone. The current findings support the hypothesis that pike localize their defecation to remove chemical cues from the foraging area of the home range in order to avoid chemically labeling their foraging area as dangerous to prey.     

Chemical labeling of northern pike (Esox lucius) by the alarm pheromone of fathead minnows (Pimephales promelas)

In previous experiments, chemical stimuli from northern pike (Esox lucius) elicited fright responses from pike-naive fathead minnows (Pimephales promelas) only if the pike had recently eaten conspecific minnows. We used a behavioral assay to determine if the fright response is the result of the incorporation of the minnow alarm pheromone into the chemical signature of the pike. Because the alarm substance cells (epidermal club cells) of fathead minnows are seasonally lost by males, we used chemical stimuli from pike that had eaten breeding male minnows as a control stimulus. In independent laboratory and field experiments, pike-naive minnows exhibited fright reactions (e.g., increased shelter use, avoidance) when exposed to water from tanks containing pike that had eaten nonbreeding fatheads (with alarm substance cells), but not to water from tanks containing pike that had eaten breeding male fatheads (without alarm substance cells). These data indicate that the fathead minnow alarm pheromone chemically labels northern pike as dangerous to pike-naive receivers.     

Population and sex differences in antipredator responses of breeding fathead minnows (Pimephales promelas) to chemical stimuli from garter snakes (Thamnophis radix andT. sirtalis)

We conducted a predator bite survey on a population of fathead minnows (Pimephales promelas) considered to be under substantial predation pressure by western plains garter snakes (Thamnophis radix). Scarring, due to failed predation attempts by garter snakes and crayfish (Orconectes virilis), was observed significantly more often in breeding males than in breeding females and nonbreeding minnows. Likely, territorial nest defense under the edges of rocks along the water's edge, a habitat occupied by crayfish and frequented by snakes, caused the breeding males to be differentially vulnerable to predation. Under controlled laboratory conditions, breeding males from this population exhibited an antipredator response to chemical stimuli from live snakes (T. sirtalis andT. radix) significantly more often than breeding female minnows from the same population and breeding minnows of both sexes from a population that was presumed to be under lower predation pressure from snakes.     

Responses of Mud Snails and Periwinkles to Environmental Odors and Disaccharide Mimics of Fish Odor

Estuarine snails, periwinkles (Littoraria irorata), and mud snails (Ilyanassa obsoleta) were tested for behavioral responses to aqueous extracts of tissue macerates, odors of living intact organisms, and to disaccharides derived from heparin. Extracts included salt-marsh cordgrass (Spartina alterniflora), pink shrimp (Penaeus duorarum), crushed periwinkles, and crushed mud snails. Odors included live periwinkles, mud snails, stone crab (Menippe mercenaria), striped hermit crab (Clibanarius vittatus), tulip snail (Fasciolaria hunteria), and mummichog (Fundulus heteroclitus). Responses were determined upon contact by snails of a ring of solution in the bottom of an otherwise dry bowl and by presenting snails in seawater with 25 l of solution on a cotton swab. In each test, the response of 30 individuals was determined. Behaviors were recorded as fed, alarm, and no response. The strongest alarm response (>80% of all snails tested) in both snails was elicited by crushed mud snails. The strongest feeding response was to shrimp and periwinkle extract for mud snails (>70%), and salt-marsh cordgrass extract attracted periwinkles. High percentages of mud snails and periwinkles fled in response to the odors of intact snail predators, stone crabs, tulip snails, and mummichogs. Mud snails and periwinkles did not flee in response to nonpredator odors, including each other's odor, as well as that of hermit crabs, shrimp, and marsh cordgrass. Heparin disaccharides were tested because other studies reported that biological activity of predatory fish odor is due to similar disaccharides originating from fish mucus. Mud snail responses to disaccharides indicate that they respond to the same kinds of molecules as crustacean larvae and that the modified amine on the disaccharide is essential for activity. The Q-tip assay is appropriate for bioassay-directed purification of alarm signals originating in fish body odor.     

Avoidance of conspecific injury-released chemical cues by free-ranging Gammarus lacustris (Crustacea:Amphipoda)

Behavioral responses to chemical cues have been demonstrated for a range of aquatic animals. Injury-released chemical alarm cues from conspecifics are released when a prey's predator is actively foraging. Detection of these cues elicits antipredator behaviors that reduce the probability of predation. Amphipod crustaceans in the genus Gammarus are widespread denizens of ponds and streams. Antipredator responses by Gammarus to conspecific alarm cues, and subsequent reduction of predation risk, are known from experiments in the laboratory. Here, we verify laboratory findings by demonstrating an avoidance response to alarm cues using a field population of G. lacustris. We used small traps baited with sponge blocks containing either water (control) or injury-released cues from Gammarus. We repeated the experiment twice. In both experiments, significantly fewer Gammarus were captured in traps with alarm cue sponges than in traps with water sponges. Predatory leeches Dina parva were attracted to Gammarus traps in the first experiment but not the second experiment. In the second experiment, we measured the individual weight of captured amphipods. Two size classes were present; small (1-5 mg) and large (35-108 mg). Both sizes contributed to the avoidance response. Within the large size class, small individuals were significantly less responsive to the alarm cue than large individuals, implying that small adult Gammarus may have different cost/benefit decision criteria for risk assessment than large Gammarus.