by Zhanna REZNIKOVA, Dr. Sc. (Biol.), Head of the Laboratory of Behavioral Ecology of Communities, Institute of Animal Systematics and Ecology, RAS Siberian Branch, and Department of Comparative Psychology, Novosibirsk State University; Sofiya PANTELEEVA, scientist from the same laboratory, NSU lecturer

The development and manifestation of behavioral traditions in various animal species have been described more than once, but it remains unclear what factors promote and, vice versa, what are the factors preventing prevalence of behavioral forms new for the population. It is therefore essential to understand the role of genetic programs in triggering of stereotypes, in other words, congenital liability to quite definite algorithm of actions. Experimental studies of the hunting behavior of ants led us to a hypothesis on "disseminated social training". We shall speak about its mechanisms below.

SIGNAL HEREDITY

Under natural conditions the animals use three sources of "knowledge": hereditary behavioral programs, individual experience, and social training based on imitation. At the beginning of the 1960s, Mikhail Lobashyov, geneticist from I. Pavlov Institute of Physiology, USSR AS, and Leningrad State University, called the transfer of behavioral characters from one generation to another by non-genetic means the "signal heredity" (signal here implying the continuity of the first and second signal systems concept formulated by Academician Ivan Pavlov). Other authors called the same phenomenon the "cultural continuity" or even "culture". The development of cognitive ethology*, development of new experimental methods promoting the discovery of the intellectual potential of animals gave rise in the last half of the 20th century to a new wave of interest to inculcation of behavioral characters by training. However, attempts to clear out ethological mechanisms of this process led to many controversies. Before we start their discussion, we have to differentiate between the notions "culture" and "behavioral traditions" for animals.

Two situations, described in the middle of the last century, are presented in ethology books as classical examples of "culture". One of them describes British tomtits, who pecked through lids of the bottles, left by milkmen by the doors, and drank the cream. The "technology" of stealing rapidly disseminated in the population living in the south of Great Britain. This phenomenon was first described in 1921. In 1949 English scientists Robert Hind and James Fisher published a map of the entire island showing the populations (their number was more than 30), which mastered this useful habit by imitation. (It is worth mentioning that the last case of cream stealing was noted in 2000: homogenized fatless milk no longer attracted the tomtits). The authors suggested the term "cultural continuity" to describe this phenomenon. Another example concerns vegetable washing "culture" practiced by the Japanese macaques. Young females and their mothers (taught by their enterprising daughters) were seen washing sweet potatoes in sea water and just 10 years later almost all members of the group followed this habit.

Today, when numerous data on the imitating behavior of animals are available, the specialists call the prevalence and inculcation of a new behavioral model in a population the "behavioral tradition". "Culture" implies a block of these traditions. Hence, the above examples should be referred not to "culture", but to behavioral traditions. Quantitative difference is essential, but the gist of these phenomena is the same: "cultural continuity". The main ethological mechanism of dissemination of new behavioral forms is social training, based on experience acquired by studies of the activities of other individuals.

"Behavioral traditions" have been described for various species of monkeys, crows, rats, and other animals capable of intricate mental and social activities. Many discoveries are still to be made in this sphere. One of the recent ones concerns the bottle-nosed dolphins living on the western coast of Australia. They learned to tear pieces of sponges from the substrate and use them for protecting (as if with a glove) the sensitive rostrum (anterior part of the snout, formed by the jaws), facilitating the task of getting benthic animals. Observations and DNA analyses suggest that this tradition is disseminated by social training from mothers to children.

The chimpanzees and orangoutangs revealed not some particular traditions, but real "cultural foci". In

* Cognitive ethology is an interdisciplinary sphere including studies of cognitive processes in animals by ethologists, psychologists, and neurobiologists.--Ed.

1999, a team of primatologists headed by Andrew Viten, English zoopsychologist, published in the Nature journal a paper, which summed up the many-year studies in seven different foci of the chimpanzee residence in Africa. Thirty-nine stable behavioral models were distinguished, differing in the spatially separated populations. The representatives of different "cultures" differently used the "instruments" for getting food, practiced different modes of "courting" and methods for "shelter building". Another team of scientists headed by Carel van Schaik, Swiss ethologist, studied six populations of orangoutangs on the islands of Borneo and Sumatra and distinguished 24 behavioral models, regarded as cultural variants. They included use of branches as instruments, some leaves as "napkins" or "gloves", protecting fingers from thorns and venoms, some other leaves they used to press to the lips to produce specific sounds for communication and frightening away the predators.

Scientists most often explain such facts by "cultural continuity". The analysis of specific examples leads to many controversies. Useful innovations very often find no imitators and die together with their carriers. In other words, members of the community obviously lack something to become successful students. On the other hand, if we analyze the manifestations of complex behavior effectively fixed in populations, we see that they are often based on hereditary (motor) stereotypes. This is true, for example, for use of stones for playing and getting of food by marmosets and Capuchin monkeys. We may therefore suspect that these traditions are more likely to be fixed in the groups in which the "nonmaterial" (signal) heredity is "materially" supported by individuals--carriers of certain congenital behavioral complexes or at least their fragments. This implies the contribution of not only the "second" (social), but also the "first" (genetic) heredity to the formation of these traditions in the populations, which we will discuss below.

NEW BEHAVIORAL FORMS: HOW ARE THEY INCULCATED?

As a rule, modern specialists speak about the use of imitation and "teaching" as the main ethological mechanism of such inculcation. They are most complex forms of bilateral social training process: accurate copying of actions or their sequences, leading to the desired goal; rough copying mode for attaining the goal; and deliberate transfer of existing habits to the "student" with obligatory spending of resources (efforts, time) of the "teacher".

Admitting the contribution of the "cultural" constituent in the transfer of definite behavioral models from generation to generation in some highly social animal species, we should like to emphasize that even for anthropoids an important factor of such process can turn to be a respective genetic component--hereditary liability. In this case there is no need in complex forms of social training-just a simple "social facilitation" is sufficient, when manifestation of this or that behavioral form is facilitated in the presence of kinsmen.

We can offer many examples to support our hypothesis. Some of them have been observed in groups of chimpanzees, when potentially useful innovations died together with their "inventors". For example, monkeys living in the national parks of Thai (Côte d'Ivoire) and Bossou (Guinea) use stones as hammers and anvils to crack nuts. They are even called "nutcracker" populations, while chimpanzees in Mahala and Gomba (Tanzania) never think of using similar instruments, despite plenty of stones and nuts there. Jane Goodall, the discoverer of these behavioral forms, a well-known primatologist (U.K.) noted two cases of use of stone instruments in the Gomba groupping and suggested that this technology would disseminate in the population. However, nothing of the kind happened throughout the 30 years that followed, and members of this groupping did not become "nutcrackers". Presumably, congenital liability to acquisition of appropriate behavioral forms based on hereditary motor stereotypes is different in different chimpanzee populations. That is why some behavioral models are easily disseminated, while others die with their carriers.

The problem of differentiating between the "cultural continuity" based on inculcation of innovations and formation of genetically based behavioral traditions is important. This will bring us closer to solution of the main problem of studies of animal "culture": what kind of factors promote or prevent the dissemination of new behavioral forms in the community.

THE ANT'S POINT OF VIEW ON "CULTURE"

We have suggested a hypothesis of disseminated social training. Experimental data supporting it are published in Acta Ethologica and Information Bulletin of All-Russia Society of Geneticists and Selectioners (2008). According to the hypothesis, a few carriers of integral stereotypes are sufficient for dissemination of complex behavioral forms in a groupping of animals,

if its other representatives have incomplete genetic programs triggering these stereotypes. The presence of these "dormant" fragments creates congenital liability to perform a definite sequence of actions. Such a simple form of social training as the above mentioned "social facilitation" is sufficient to complete an integral stereotype. We have called this behavioral form "disseminated social training", as it presumably implies complete behavioral programs and their fragments, prevailing in the members of the population. This does not exclude other modes of inculcation of new behavioral models, including those based exclusively on social training and involving imitation and "teaching". On the other hand, our approach suggests a simpler explanation of many situations in the observed "culture".

We have tested this hypothesis by analyzing the development of complex behavioral stereotypes in Myrmica rubra L. ants in the process of hunting. The ants catch Collembolas, mobile insects easily slipping off. Up till now it was assumed that Myrmicas, numerous residents of the forest bedding and soil, are collectors of dead or immobile (after molting) invertebrates, and cannot hunt actively. The Collembolas have a spring and rapidly modulate the direction of their movement, and are therefore not easy bag. Therefore, Collembola hunting was regarded as an exotic phenomenon for ants. The ants are mainly hunting Collembolas and other small jumping insects using their clapping jaws. They are residents of South America, tropical Asia, Australia, and Africa. We first demonstrated the ability of several species of small ants (including Myrmicas) prevalent in the forest zone to hunt Collembolas.

The hunting habits of Myrmicas include finding of the prey, rapid jumping on it from above, capture, and stinging. We called this rather complex behavioral stereotype "jumping attack". It turned out that under natural conditions Myrmica families could completely confine themselves to hunting Collembolas, provided their number was considerable. However, if in an experimental setting the Collembolas are offered to ants of the same species, but residents of areas where this bag is scanty, we see that they cannot hunt.

How not simple skills of attack are spread in ants? Why so many lucky hunters emerge when Collembolas are in abundance? Perhaps, they learn from one another, that is, we deal here with the ant "culture"? Let us discuss another hypothesis: stimuli ants get from potential bag (provided they have a congenital pattern of perception of its "image") liberate their genetically programmed reactions and the hunting behavior stereotype, first imperfect, later on "completed" due to individual experience. Such order of the hunting behavior development has been detected in various animal groups, from insects to mammals. However, it is not realized equally well in all of them. Our experiments have shown that "naive" ants (grown in laboratory conditions), kept together with their potential victims, failed to develop hunting interest and skills to capture the victim, despite hundreds of contacts.

Hence, the stimuli from the bag were insufficient to awake "hunting excitement" in this case, at least in the majority of ant family members.

But presumably we deal with a local "culture": the ants master hunting process in Collembolarich areas by often watching their more lucky fellows? This hypothesis was rejected as well, as we found in one of the "naive" families seven individuals (out of 123), who demonstrated a hunting behavioral stereotype when meeting a Collembola: "jumping attack" and capture. (It is noteworthy that the working ants in Myrmica families are genetically heterogeneous, as they are offspring of several tens of females). Presumably, our "naive" family had several individuals with a congenital program of the integral hunting behavior stereotype and with the bag "image" perception pattern, and a meeting with potential objects of attack triggered manifestation of the integral stereotype.

Other ants in the same family seem to have just fragments of the required program. Many-stage development is needed for its complete formation. As we have already mentioned, potential victims (Collembolas) do not provoke striving to attack them in the individuals, who are not "born hunters". In areas with plenty of this bag virtually all ants are lucky hunters. It is logical to suggest that this striving arouses in them due to their congenital behavioral programs, presumably incomplete, when they witness a small number of "born hunters" (who have an integral stereotype) capturing the jumping bag. This is a frequent situation, presumably leading to a cumulative effect, in areas with plenty of Collembolas. If the ants rarely meet lucky hunters, formation of hunting skills can take months (that is, almost the entire life span).

Hence, the main experimental fact of our studies has become a discovery of two groups in the ant populations: with a complete congenital hunting stereotype and a potential victim perception pattern and those with just fragments of this stereotype. This hypothesis was supported by discovery of carriers of fragments of stereotypes not only in ant families, but also in some vertebrates, including social rodents (gerbils). Thus, grounds for the formation of genetically based behavioral tradition have been found.

"FIRST" OR "SECOND" HEREDITY?

The "disseminated social training" hypothesis seems to explain some cases of inculcation of cultural traditions in communities of different animal species with consideration for not only the "second"--signal heredity, but with consideration for the "first" one (congenital liability to formation of certain associative relationships) as well.

Presumably, not numerous carriers of complex behavioral stereotypes serve as a sort of catalysts for their kinsmen with just incomplete "dormant" fragments of genetic programs. The most unsophisticated forms of social training are sufficient for completion of these programs. This explanation suggests an alternative interpretation of many situations, hardly explained (as we think) from the standpoint of "cultural" transfer of behavioral traditions. We think that insufficient differentiation between the species-specific stereotypes and innovations, as well as underrating of genetic factors lead to disputable interpretations of ethological mechanisms of new behavioral forms prevalent in animals. Let us illustrate it by several examples.

Scientists from the University of St. Andrews (U.K.) and the Primatological Center (USA) have shown recently that new food-getting technology was rapidly propagated in groups of adult chimpanzees on the basis of imitation of the leader. Andrew Viten and his colleagues used "artificial fruits"--boxes with favorite food, difficult to open. According to the authors, their results contradict the commonly accepted opinion on the important role of a critical period in early childhood for mastering complex behavioral models. However, we believe that there are no contradictions. The existence of a period for mastering complex forms has been demonstrated by the results of experiments carried out in Russia by Leonid Firsov, Dr. Sc. (Med.), zoopsychologist from I. Pavlov Institute of Physiology, USSR AS, and his colleague Stella Bruer from the U.K. at the end of the 1970s. The chimpanzees whose early childhood was spent under natural conditions, later on, getting to man, easily mastered the skills of using instruments and construction of nests, while animals brought up in captivity from the neonatal period were later unable to do these things. It seems that after a certain age the primates are no longer able to master some key behavioral models.

So, why have adult animals in the above-mentioned experiments with "artificial fruits" so easily mastered and widely used (by imitation) new food-getting technology? We suppose that the presence of a critical period for mastering certain species-specific stereotypes indicates a significant role of the genetic component in their development. That is, the impact of congenital stereotypes does not allow the animal shift significant-

ly in the course of development of certain behavioral forms, which are (like in the above situations with chimpanzees) a result of joint effects of congenital behavioral programs, imprinting, imitation, and individual experience. By the way, the further from the species-specific stereotype is the behavioral form to be mastered--the easier the animals (not depending on the congenital stereotypes) learn it. Such was the situation in the above-mentioned experiments. In them the primates mastered food-getting methods quite unlike natural situations (the same box with locks), while in the observations, which the authors compared with their results, the animals exhibited elements of species-specific behavior (use of stones and branches as instruments).

Other experiments carried out by Leonid Firsov illustrate our hypothesis on the role of the genetic component. The chimpanzee brought up by humans from the neonatal period could not build nests and "fish" ants with a branch under natural conditions, but they easily opened the locks and rapidly understood how to use, say, a stick for getting out a rope which fell into water and by this rope, pool the boat towards the bank.

It is noteworthy that the training philosophy and even cognitive potentials are largely similar in vertebrates and invertebrates. That is why we should not hurry to explain the prevalence of complex behavioral forms in groups of dolphins or anthropoids by the development of "cultural traditions". Each of them may be based on genetic liability and, presumably, "disseminated social training", as in our ants.

One of the examples illustrating this hypothesis is a specific behavioral model of grooming* in chimpanzees, called "hand in hand". The pair of these monkeys assumes a characteristic posture, resembling the letter A during grooming: the animals couple two raised hands, while with two other hands they groom the fur on each other. When the group of animals is at rest, their raised hands show their interaction. We think that hereditary liability plays an important role in this behavioral form. The "hand-in-hand grooming" is an obvious example of species-specific behavior with limited prevalence in populations. It seems to differ significantly from innovations disseminated in the communities by "cultural continuity".

The results of our experiments on ants suggest a simpler explanation to "cultural" behavior of animals.

One more example of this fact is formation of habits of instrument handling in crows from New Caledonia. In nature these birds get insects from cracks in tree barks by means of transformed plant parts. Under laboratory conditions they demonstrate amazing intellect while solving instrumental problems: they get fodder using sticks and wire fragments, easily transforming them to fit the task. Based on the well-developed cognitive potentialities of New Caledonian crows, the scientists think that use of instruments in the natural populations of this species is more likely to be determined by training and cultural continuity than by realization of the heredity program. However, high specialization of "instrumental behavior" in nature and the 100 percent "coverage" of complex skills by all members of the population suggest under consideration an important role of the hereditary component in the formation of corresponding basic stereotypes in this species.

This hypothesis is supported by the results of experiments described in the Nature journal in 2005. Scientists of the University of Cambridge (U.K.) bred

* Grooming is an active behavior of animals to clean the body surface.--Ed.

Science in Russia, No. 6, 2011

four "naive" little crows and traced the development of their capacity to use tools. The nestlings were offered branches and leaves similar to those their kinsmen found in nature. One nestling demonstrated an effective sequence of actions (as our hunting ants) and extracted a larva from a fissure using a branch. Later on, he repeated this action more than once. For others, it took many days to master this skill.

These results suggest the presence of carriers of the integral stereotype of "instrumental" behavior in New Caledonian crow populations. Presumably, these carriers serve as catalysts for manifestation of the same mode of activities in the birds, that have just fragments of congenital program. Of course, we cannot rule out the cognitive component of this activity--the authors of studies, which are still in progress, insist on it. However, the base for innovations in this species is most likely a stereotype with a strong hereditary component.

Perhaps, innovations disseminate in populations on the basis of imitation (of the most complex forms of social training), while elements of complex forms of species-specific behavior disseminate by means of social facilitation. But we should remember that the probability of inculcation of innovations in the groups of animals is rather low, as they are "introduced" in a viscous medium of carriers of the stereotypes inherent in the species, and just few individuals can "catch up the banner" and carry it for a sufficiently long time, so that the new behavioral form is transferred from one generation to another by means of signal heredity.

Summing up, let us mention that adaptive potentialities of populations can be extended rather "economically": there is no need for animals to master complex behavioral stereotypes for all life situations; it is quite enough to have just some "blanks" and possess a capacity to the simplest forms of social training. Genetic liability is the best "teacher" for animals, at least for many of them.