Ants on the Binary Tree

The main point of the suggested approach is that the experiments provide a situation in which ants have to transmit information quantitatively known to the researcher in order to obtain food. This information concerns the sequence of turns toward a trough of syrup. We used the laboratory set-up called “binary tree” where each “leaf” of the tree ends with an empty trough with the exception of one filled with syrup. The simplest design was the tree with two leaves. It represents one binary choice that corresponds to one unite of information introduced by Shannon, the bit. In this situation a scouting animal should transmit one bit of information to other specimens: to go to the right (R) or to the left (L). In other experiments the number of forks in one branch increased to six. Hence, the number of bits necessary to choose the correct way is equal to the number of bifurcations, in this case six. (Fig. IX-5).

The experiments with binary tree were conducted with 10 laboratory colonies of six ant species. Ants were housed in plastic transparent nests that made it possible to observe their contacts. All the ants were labelled with an individual colour mark. The composition of ants’ groups was revealed during preliminary stages of the experiments. In Formica ants the small cliques within the colonies were discovered which were composed of a “scout” and 5-8 “recruits” (foragers). The total of 335 scouts were used in the main trials. A scout was placed on a trough containing food, and then it returned to the nest by itself. The scout had to make up to four trips before it was able to mobilize the group of foragers. In all the cases of mobilization we measured (in seconds) the duration of the contact between the scout and the foragers. When the group began moving to the maze the scout was isolated by tweezers and the foragers had to search for the food by themselves. To prevent access to the food by a straight path, the set-up was placed in a water bath, and the ants reached the initial point of the binary tree by going over a bridge.

The experiments were devised so as to eliminate all possible ways helpful to finding food, except information contact with the scout. To avoid the use of an odour track, the experimental set-up was replaced by an identical one when the scout was in the nest contacting to its group. The fresh maze contained all troughs full with water in order to avoid the possible impact of the smell of syrup. If the group reached the correct leaf of the binary tree they were immediately presented with the food (Fig. IX-6).

The long – term experiments revealed information transmission based on distant homing within small constant cliques consisting of a scout and foragers in Formica s.str. Not all of the scouts managed to memorize the way to the correct leaf on the maze. The number of such scouts decreases with the complication of the task. In the case of two forks, all active scouts and their groups (up to 15 per colony) were working, while in the case of six forks, only one or two coped with the task. It turned out that able scouts could transmit information on several different sequences of turns during one daily experiment.

Evidence of information transmission from the scouts to the foragers came from two sets of data: first, from the statistical analysis of the number of faultless findings of the goal by a group, and second, from special series of control experiments with “uninformed” (“naive”) and “informed” foragers.

The statistical analysis of number of faultless findings of the goal was carried out by comparing Hypothesis H₀ (ants find the “right” leaf randomly) with Hypothesis H₁ (they find the goal thanks to obtained information), proceeding from that the probability of a chance finding of the correct way with i number of forks is (1/2) ⁱ Experimenters analyzed different series of experiment (338 trials in sum), separately for 2, 3, 4, 5 and 6 number of forks. In all cases H₀ was rejected in favour of H₁, P < 0.001 (see: Ryabko and Reznikova, 1996).

The special experiments were performed in which naive ants were tested in the maze. Researchers compared searching results in the ants that had and had not previous possibility to contact with the scout (the “informed” and “naive” scouts, correspondingly.) The naive and “informed” ants were allowed to search for the food during 30 min. In more agile F. pratensis the time spent on searching the trough by “informed” and “uninformed” specimens were compared (Reznikova, Ryabko, 2003; Novgorodova, 2005, 2006). For example, in

F. pratensis almost all “naive” foragers were able to find food on their own but they spent 10-15 times more than those ants which entered the maze after the contact with the successful scout. Average values as well as amounts of sampling are given in the Table IX-1. For every trial we used Wilkokson’s non-parametric test (see: Hollander, Wolf, 1973) to examine a Hypothesis H₀ (data from both samples follow the same distribution) against H₁ (they follow different distributions) at significance level 0.01. It turns out that the duration of searching time is essentially more in those ants that previously contacted with the successful scout.

Table IX-1. Comparison of duration of search for the trough by “uninformed” (U) F. pratensis ants and individuals that previously contacted with the successful scout (“Informed”, I); July-August, 2003.

From: Reznikova and Ryabko, 2003

In sum, the obtained data confirm information transmission in three species belonging to subgenus Formica s.str. and exclude any orientation mechanism, except the use of information transmission by the scouts.

No information transmission based on distant homing was observed in the rest two species. In singly foraging F. cunicularia the foragers learned the way towards the maze, while making up to 30 trips per day, but they did not try to recruit other members of their colony. Not more than 5 ants were active in the maze per day. M. rubra workers used olfactory cues, but when we changed the maze, they had to do without odour trails. In these cases they resorted to only solitary foraging, just as F. cunicularia.

Evaluation of information transmission rate in ants is based on the fact that the quantity of information (in bits), necessary for choosing the correct way toward the maze, equals i, the number of forks. One can assume that this duration (t) was ai + b, where i is the number of forks, a – coefficient of proportionality, equals to the rate of information transmission (bit/min), and b is an introduced constant, since ants can transmit information not related directly to the task, for example, the simple signal “food”. Besides, it is not ruled out that a discovering ant transmits, in some way, the information on its route to the nest, using acoustic or some other means of communication. In this connection, it is important that the way from the maze to the nest was in all experiments approximately the same and, therefore, the time before the antennal contact with the foragers in the nest, which the scout could hypothetically use for the message transmission, was approximately the same and did not depend on the number of forks.

From the obtained data, the parameters of linear regression and the sample correlation coefficient (r) can be evaluated. The rate of the information transmission (a) derived from the equation t = ai + b, was 0.738 for F. sanguinea and 1.094 for F. rufa. These values are not considered species-specific constants, they probably vary. Note that the rate of information transmission is relatively small in ants.

In order to estimate the potential productivity of ants “language”, let us count the total number of different possible ways to the trough. In a simplest binary tree with one fork there are two leaves and therefore two different ways. In a tree with two forks there are 2² ways, with three forks, 2 ³ ways, and with six forks, 2 ⁶ ways; hence, the total number of different ways is equal to 2+2²+2³+…2⁶=126. This is the minimal number of messages the ants must possess in order to pass the information about the food placed on any leaf of the binary tree with 6 forks.

Another series of experiments was based on a basic concept of Kolmogorov complexity and was aimed to check whether highly social ant species specimens possess such an important property of intelligent communicators as the ability to quickly grasp the regularities and to use them for coding and “compression” of information. Thus the length of the text should be proportional to the complexity of the information. This idea is the basic concept of Kolmogorov complexity. This concept is applied to words (or text) composed of the letters of an alphabet, for example, of an alphabet, consisting of two letters, L and R. Informally, the complexity of a word (and its uncertainty) equals the length of its most concise description, according to Kolmogorov. For example, the word “LLLLLLLL” can be represented as “8L”, the word “LRLRLRLR “ as “4LR”, while the “random” word of shorter length “ LRRLRL” probably cannot be expressed more concisely, and this is the most complex of the three and has the greatest uncertainty.

Reseachers analysed the question of whether ants can apply simple “text” regularities for compression (here “text” means the sequence of the turns toward the maze). As proven by Kolmogorov (1965), there is no algorithmically computable quantitative measure of text complication. So, strictly speaking, we can only verify whether ants have the “notion” of simple and complex sequences.

In the special series of experiments, ants F. sanguinea were presented with different sequences of turns. Comparison of the main Hypothesis H₀ (the time of the information transmission does not depend on the text complexity) with Hypothesis H₁ (this time actually depends on it) showed that the more time ants spent on the information transmission, the more information - according to Kolmogorov- was contained in the message. It is interesting that the ants began to use regularities to compress only quite large “texts”. Thus, they spent from 120 to 220 sec.

to transmit information about random turn patterns on the maze with 5 and 6 forks and from 78 to 135 sec. when turn patterns were regular. On the other hand, there was no essential significance when the length of sequences was less than 4 (Table IX-2).

Table IX-2.

Duration of transmitting information on the way to the trough

by F.sanguinea scouts to foragers (no.1-8 regular turn pattern;

no. 9-15 random turn pattern). From: Ryabko and Reznikova, 1996.

3 2. 2. Evaluation of ants’ “language”

The majority of models that describe processes of self-organisation in colonies of social insects consider cognitive skills redundant for effective information transmission. However, another set of arguments comes from the data concerning excellent learning abilities of social insects. Bulk of the obtained results concerns orientation and memory, mainly in bees and wasps. Probably, studying communicative means is one of the most effective tools to comprehend limits of intelligence in social insects at the individual level. For this, appropriate species have to be chosen as well as the adequate set of tasks for the colony to be solved.

From early experiments of Schneirla (1946) and Thorpe (1950) it was known that some ants and solitary wasps perform almost as well as rats and dogs in maze-learning and detour tasks. Mazokhin - Porshnyakov (1969, 1984) experimentally demonstrated that the honey bees and social wasps are capable of abstraction, extrapolation and solving rather sophisticated discrimination tasks at the level of dogs and monkeys. In particular, several individually trained bees were able to distinguish between chains consisting of paired and unpaired small figures and thus capable for concept formation. Guirfa et al. (1996) found bees to be able to form concepts about symmetry versus asymmetry.

As we have already seen in Parts IV, V, and VI, bees, wasps and ants possess several specific mechanisms which help them in the efficient perception and learning (Collet et al., 1993; Menzel, Mhller, 1996). For example, memory of landmarks passed by on flights between the hive and the feeding place is organized sequentially in honeybees, so they have to “count landmarks” (Chittka and Geiger, 1995). Behavioural flexibility allows social insects to switch their learning patterns in accordance with changeable environment. For example, when the bees are prevented from learning landmark cues on arrival to a hive, they match for learning them during specific “turn-back-and-look” flight maneuvers (Lehrer and Bianko, 2000).

Ants are known to do many clever things such as using cognitive maps (Wehner, 1990), learning by observation (Reznikova, 1982, 2001) and even counting (Reznikova and Ryabko, 2001). Seemingly to bees, from which only one species, the honey bee, possesses symbolic language, from about 10000 ant species, only few highly social species use sophisticated system of communication that has been revealed by applying methods of information theory. Among these highly social species red wood ants (Formica s.str.) belong to the most intelligent ones. Rosengren and Fortelius (1987) characterized them as “replete ants” storing not lipids in their fat-bodies but habitat information in their brains. In comparison to many sympatric species, red wood ants have hundred times more individuals in their colonies and huge feeding territories. They face complex tasks every day; for example, in order to obtain honey dew, their basic food, red wood ants have to memorise and possibly pass each other information about locations of thousands aphid colonies within such a huge three-dimension space as a tree is for an ant (Reznikova and Novgorodova, 1998). Tasks of this kind require communicative means for distant homing. It is natural to expect that ant species with different colony design develop communication systems of different levels of complexity. Thus, Robson and Traniello (1998) found a major difference between mass-recruitment ant species that are characterized by collective actions of simple individuals, and the group retrieving ones. In the latter case, a colony design is based on complexity rather than simplicity. The process of recruitment is so organized that the removal of the discovering ant leads to the dissolution of the retrieval group.

Elaborating possible approach for comparative investigation of communication, we found out that the most likely situation to observe task distribution and behavioural flexibility at the individual level and to evaluate potential properties of communication in group- retrieving ants is to force them to solve a complex search problem. In this situation hidden processes of information transmission would become observable.

Using ideas of Information Theory to design experiments enables researchers not only to reveal information transmission from scouts to foragers within cliques but also to estimate the rate of this process and to suggest that group – retrieving ants are able to memorize and use simple regularities, thus compressing the information available. This, in turn, gives a possibility to assume flexibility of communication system in highly social ant species based on individual cognitive capabilities that help colonies to solve many daily, but non-standard, problems.

The hypothesis of flexibility of ants’ “language” is supported by the results of “counting” experiments of Reznikova and Ryabko (2000, 2001) that were described in Part VI. It was suggested that ants are able to invent several new “words” in order to denote frequently encountered landmarks. Applying ideas of Kolmogorov complexity to study ants’ “language” enabled researchers to consider ants to be able to “compress” both perceived information and transferred messages on the basis of grasped regularities of serial turns to the desired goal. All these data allow to suggest that group retrieving ant species possess a flexible communication system which makes the process of transferring information less costly for them.

CONCLUDING COMMENTS

During last decades, common efforts of scientists applying different experimental approaches revealed some features of communication systems of non-human species that were earlier attributed exclusively to humans. Among them one can list animals’ ability to use referential signals organised by “proto-grammar” rules, to transfer messages in abstract “symbolic” form, to create messages about things and events distant in time and space, to “translate” messages of other species and to extract meaningful parts from strangers’ signals.

However, one can find many points of discontinuity with the communicative practice of animals. Although members of several species demonstrate understanding of grammatical rules when using artificial intermediary languages, there is no evidence of syntax in the natural communication of animals. There is also little evidence for learning and modification in the natural signals of animals. There is much do be done to reveal evolutionary roots of such sophisticated system of communication as human language.

Each of three methodological approaches that have been discussed in this part has its specific power. Direct deciphering of animals’ communication is the oldest from the considered methods. By now some of the researchers who have tried to crack wild codes by means of recording signals and playback experiments became masters of at least segments of King Solomon’s Ring. Dictionaries, although very fragmentary, have been compiled for several species of mammals, birds and insects. The decoded “words” concern alarm calls, calls for cohesion, and signals about food. The honeybees’ dance language remains the most complex among animals’ communication system that was decoded up to date. Beside this unique system, other communications are difficult to decipher because of many methodological barriers, among which low repeatability of standard living situations in wild life is important to note.

A number of studies with captive animals that have been trained with human-designed artificial communication systems has revolutionary changed our imagination about animals’ linguistic abilities and language-related cognitive skills. Language-training experiments are of great help for discovering potentials of animal language behaviour and for studying roots of human predisposition for the development of sophisticated language. Animals from very far systematic taxa that were taught very different artificial languages have met the same criteria of language rules and demonstrated their competence in syntax and semantics. But the evolutionary significance of these animals’ ability would be difficult to evaluate if species’ naturally existing syntactic abilities that have evolved in the context of their natural communicative behaviour were not revealed. At the same time, only narrow and limited syntactic abilities discovered in natural communicative systems restrict our possibilities to judge about potential power of species’ language behaviour and related cognitive abilities.

For this, the use of ideas of Information Theory opens new horizons. This approach is designed to study quantitative characteristics of natural communicative systems and important properties of animal intelligence. Applying this method, there is no need to crack animals’ codes. All we need is to ask animals to pass messages of definite lengths and complexities. By measuring time duration which the animals spend on transmitting messages with desired conditions, we can judge about potentials of their communicative system. By means of this approach it has been demonstrated that communicative system of highly social ants meets several fundamental criteria of language such as productivity, specialisation, and displacement.

From the view of information theory at least two important standards should be added to a list of criteria characterising language. First, the length of a message should correlate with the quantity of information contained in this message. Second, the ability to grasp the regularities and to use them for coding and “compression” of information should be considered one of the most important properties of language and its carriers’ intellect. These properties were revealed in language behaviour of Formica ants. The use of ideas and methods of information theory led to the discovery of flexibility of communication systems in ants which may be considered one of the most complex properties of animal “languages”. It is a challenge to apply the information theory approach for studying communication in a wide variety of social animals.