Karen L. Hollis

The biological function of Pavlovian conditioning: the best defense is a good offense

Signaled presentations of a rival male produce an aggressive conditional response in several species of fish. Although conditioning of such species-specific display behavior has been described previously, the biological function of this learning phenomenon has remained unknown. I present experimental evidence that signaled territorial intrusion enables a male blue gourami to defend his territory more aggressively than when the intruder is unsignaled. In Experiment 1, pairs of territorial males, with different conditioning histories, confronted one another for the first time. One member of the pair previously had received Pavlovian conditioning, whereas the other pair member had received explicitly unpaired presentations of the same signal--conditional stimulus (CS)--and rival male--unconditional stimulus (US). In the subsequent encounter, which was signaled by CS presentation, Pavlovian males delivered significantly more bites and tailbeatings than did their control group opponents. Experiment 2 relied upon a different control procedure, a US-only condition, but, again, Pavlovian males enjoyed a significant aggressive advantage. These results suggest an important ecological role for Pavlovian conditioning.

Ants, Cataglyphis cursor, Use Precisely Directed Rescue Behavior to Free Entrapped Relatives

Although helping behavior is ubiquitous throughout the animal kingdom, actual rescue activity is particularly rare. Nonetheless, here we report the first experimental evidence that ants, Cataglyphis cursor, use precisely directed rescue behavior to free entrapped victims; equally important, they carefully discriminate between individuals in distress, offering aid only to nestmates. Our experiments simulate a natural situation, which we often observed in the field when collecting Catagyphis ants, causing sand to collapse in the process. Using a novel experimental technique that binds victims experimentally, we observed the behavior of separate, randomly chosen groups of 5 C. cursor nestmates under one of six conditions. In five of these conditions, a test stimulus (the “victim”) was ensnared with nylon thread and held partially beneath the sand. The test stimulus was either (1) an individual from the same colony; (2) an individual from a different colony of C cursor; (3) an ant from a different ant species; (4) a common prey item; or, (5) a motionless (chilled) nestmate. In the final condition, the test stimulus (6) consisted of the empty snare apparatus. Our results demonstrate that ants are able to recognize what, exactly, holds their relative in place and direct their behavior to that object, the snare, in particular. They begin by excavating sand, which exposes the nylon snare, transporting sand away from it, and then biting at the snare itself. Snare biting, a behavior never before reported in the literature, demonstrates that rescue behavior is far more sophisticated, exact and complexly organized than the simple forms of helping behavior already known, namely limb pulling and sand digging. That is, limb pulling and sand digging could be released directly by a chemical call for help and thus result from a very simple mechanism. However, its difficult to see how this same releasing mechanism could guide rescuers to the precise location of the nylon thread, and enable them to target their bites to the thread itself.

The biological function of Pavlovian conditioning: Learned inhibition of aggressive behavior in territorial fish

Abstract Following Pavlovian discrimination training, stimuli predicting the appearance of a territory intruder (an excitatory conditional stimulus, CS+) or the absence of that event (an inhibitory conditional stimulus, CS−) were presented to pairs of territorial male fish immediately before their first aggressive interaction. Pairmates that both received excitatory stimuli prior to the confrontation were significantly more aggressive than a control group which received the same training but which received neither a CS+ nor a CS− in the test. Pairmates in which both received a CS− were significantly less aggressive than the control group. In these three groups, no differences in aggression were observed between the individual members of a single pair. Two additional groups were composed of pairs whose members received different stimuli prior to the test. Although no differences were found between pairmates in the group in which one member received a pretest CS− while the other member received no stimulus presentation, large differences in aggressive behavior were obtained when one fish received a CS+ and its pairmate received a CS−. We discuss the behavioral ecology of terriorial behavior in fish and attempt to show the potential importance of these results in this naturalistic context. In addition, we discuss the implications of an ecological approach for causal analyses of inhibitory learning.

The role of learning in the aggressive and reproductive behavior of blue gouramis, Trichogaster trichopterus

Like countless other vertebrates and even many invertebrates, blue gouramis, Trichogaster trichopterus, a freshwater tropical fish, can learn to associate environmental cues with the appearance of biologically important events. Research with blue gouramis reveals that this capacity for learning, which psychologists call Pavlovian conditioning, provides the means to enhance their territorial defense as well as reap large reproductive benefits. That is, males are able to defend their territories more efficiently when the appearance of a rival is signaled than when territorial invasion is less predictable. Not only does signaling enable Pavlovian-conditioned males to mount a more aggressive defense of their territories than males that do not have the benefit of signaling, but it helps them to concentrate their efforts at times when, or places where, the territory is most vulnerable. In addition, both males and females can learn to anticipate the appearance of a mate. Signaling of female accessibility enables Pavlovian-conditioned males to attenuate their initial aggressive response to arriving females without compromising their territorial defense. More important, when the arrival of a potential mate is signaled, Pavlovian-conditioned males are able to spawn with females sooner, clasp females more often, and produce more young than males that do not have the benefit of a signal. The ability of blue gouramis, among many other fish, to anticipate the appearance of such different biologically important events suggests that Pavlovian conditioning may play a significant role in their behavioral ecology. In addition, conditioning may provide a useful tool in several practical domains, including animal husbandry, zoological management, and conservation biology.

Rescue behavior: Distinguishing between rescue, cooperation and other forms of altruistic behavior

Reports of rescue behavior in non-human animals are exceedingly rare, except in ants where rescue is well known, but has not been explored experimentally until recently. Although we predict that rescue behavior should be limited to circumstances in which the victim and the rescuer are highly related to one another, or in which unrelated individuals must cooperate very closely with one another, we also predict that it is likely to be far more common than the current literature would suggest. To address this oversight, we propose a rigorous definition of rescue behavior, one that helps researchers to focus on its necessary and sufficient components, at the same time that it helps to differentiate rescue behavior from cooperation and other forms of helping behavior. In this way we also hope to expand our understanding of altruism in particular and kin selection in general.

Novel strategies of subordinate fish competing for food: learning when to fold

In group-living animals, competition for a scarce resource often results in the formation of dominance hierarchies. Dominant individuals garner more of that resource for themselves through the use of aggression, forcing subordinates to rely on nonaggressive strategies to secure it. Stealing from dominants is common, but subordinate thieves can risk attack if caught. How, then, do subordinates decide when and where to steal? We explored whether the opportunity to learn about the location and availability of food might facilitate food-stealing attempts by male blue gouramis, Trichogaster trichopterus , a freshwater tropical fish. Our results show that subordinates were able to use learned cues to mount highly effective ‘sneaky’ tactics. By anticipating food location and availability, subordinates reduced the number of attacks they received as well as the amount of time they spent fleeing from dominants. More importantly, however, our results revealed unexpected plasticity in the cognitive mechanisms underlying this learning: like privately trained dominants, privately trained subordinates learned to approach food cues directly; however, on the very first occasion that subordinates were tested with a dominant present, they dramatically altered their previously learned direct approach to food cues, immediately adopting a different, sneaky tactic instead. Thus, subordinates were able to combine information about previous learning, namely an ‘expectancy’ of food, with the probability of future attack and adjust their food-getting strategy accordingly.

Toward a Behavioral Ecology of Rescue Behavior

Although the study of helping behavior has revolutionized the field of behavioral ecology, scientific examination of rescue behavior remains extremely rare, except perhaps in ants, having been described as early as 1874. Nonetheless, recent work in our laboratories has revealed several new patterns of rescue behavior that appear to be much more complex than previously studied forms. This precisely-directed rescue behavior bears a remarkable resemblance to what has been labeled empathy in rats, and thus raises numerous philosophical and theoretical questions: How should rescue behavior (or empathy) be defined? What distinguishes rescue from other forms of altruism? In what ways is rescue behavior in ants different from, and similar to, rescue in other non-human animals? What selection pressures dictate its appearance? In this paper, we review our own experimental studies of rescue in both laboratory and field, which, taken together, begin to reveal some of the behavioral ecological conditions that likely have given rise to rescue behavior in ants. Against this background, we also address important theoretical questions involving rescue, including those outlined above. In this way, we hope not only to encourage further experimental analysis of rescue behavior, but also to highlight important similarities and differences in very distant taxa.

Rescue of newborn ants by older Cataglyphis cursor adult workers.

Cataglyphiscursor worker ants are capable of highly sophisticated rescue behaviour in which individuals are able to identify what has trapped a nestmate and to direct their behaviour towards that obstacle. Nonetheless, rescue behaviour is constrained by workers’ subcaste: whereas foragers, the oldest workers, are able both to give and to receive the most help, the youngest workers, inactives, neither give nor receive any help whatsoever; nurses give and receive intermediate levels of aid, reflecting their intermediate age. Such differences in rescue behaviour across subcastes suggest that age and experience play a critical role. In this species, as in many others in which a sensitive period for nestmate recognition exists, newly enclosed ants, called callows, are adopted by ants belonging not only to different colonies but also to different species; foreign callows receive nearly the same special care provided to resident newborns. Because callows are younger than inactives, which are incapable of soliciting rescue, we wondered whether entrapped callows would receive such aid. In the present study, we artificially ensnared individual callows from their own colony (homocolonial), from a different colony (heterocolonial), and from a different species (heterospecific), and tested each one with groups of five potential C.cursor rescuers, either all foragers or all nurses. Our results show that all three types of callows are able to elicit rescue behaviour from both foragers and nurses. Nonetheless, nurse rescuers are better able to discriminate between the three types of callow victims than are foragers.

Ants and antlions: The impact of ecology, coevolution and learning on an insect predator-prey relationship

A behavioural ecological approach to the relationship between pit-digging larval antlions and their common prey, ants, provides yet another example of how the specific ecological niche that species inhabit imposes selection pressures leading to unique behavioural adaptations. Antlions rely on multiple strategies to capture prey with a minimal expenditure of energy and extraordinary efficiency while ants employ several different strategies for avoiding capture, including rescue of trapped nestmates. Importantly, both ants and antlions rely heavily on their capacity for learning, a tool that sometimes is overlooked in predator-prey relationships, leading to the implicit assumption that behavioural adaptations are the result of fixed, hard-wired responses. Nonetheless, like hard-wired responses, learned behaviour, too, is uniquely adapted to the ecological niche, a reminder that the expression of associative learning is species-specific. Beyond the study of ants and antlions, per se, this particular predator-prey relationship reveals the important role that the capacity to learn plays in coevolutionary arms races.

Organization of rescue behaviour sequences in ants, Cataglyphis cursor, reflects goal-directedness, plasticity and memory

The experimental study of rescue behaviour in ants, behaviour in which individuals help entrapped nestmates in distress, has revealed that rescuers respond to victims with very precisely targeted behaviour. In Cataglyphis cursor, several different components of rescue behaviour have been observed, demonstrating the complexity of this behaviour, including sand digging and sand transport to excavate the victim, followed by pulling on the victims limbs as well as the object holding the victim in place, behaviour that serves to free the victim. Although previous work suggested that rescue was optimally organized, first to expose and then to extricate the victim under a variety of differing circumstances, experimental analysis of that organization has been lacking. Here, using experimental data, we characterize the pattern of individual rescue behaviour in C. cursor by analysing the probabilities of transitions from one behavioural component to another. The results show that the execution of each behavioural component is determined by the interplay of previous acts. In particular, we show not only that ants move sand away from the victim in an especially efficient sequence of behaviour that greatly minimizes energy expenditure, but also that ants appear to form some kind of memory of what they did in the past, a memory that directs their future behaviour.