Minimization of information processing in communication events - a model

Author(s)

Jose Eduardo Soubhia Natali

• Document type: Master's Dissertation
• Press: São Paulo.
• Institution: Universidade de São Paulo (USP). Instituto de Biociências (IBIOC/SB)
• Defense date: 2009-12-02

Examining board members

José Guilherme de Souza Chaui Mattos Berlinck; Luiz Henrique Alves Monteiro; Gilberto Fernando Xavier

• Advisor: José Guilherme de Souza Chaui Mattos Berlinck

Abstract

Communication between individuals of the same species is present in all known taxa. Focusing in organisms that have nervous system, since there is a limited capacity to process all the environmental input, it is interesting to assume an evolutionary trend to minimize the information processing associated with communication in order to redirect attention to other events. As such a minimization may be associated with a reduction in the amount of energy spent in the process, we may put forward the hypothesis that there is, also, a trend to minimize entropy generation (σ), since this is directly associated with power dissipation during action potentials. The objective of this study is to test the above described hypothesis. To this end, we considered the organisms involved in the communication as oscillators described by differential equations, and communication was defined as the coupling between oscillators. In an initial model, the frequency and the phase were the state variables, whereas in a second model we added two more variables associated with the volition to emit signals. Entropy generation is calculated through the product of the flux (i.e., frequency) by a potential difference (a forcing parameter). Computations of σ were performed for both models, and we compared the conditions of identical with those of non-identical parameters of the oscillators. In both models, the results obtained indicate that identical oscillators are those that minimize entropy generation in relation to systems of non-identical oscillators. Within the context of systems of identical oscillators, we analyzed the regions of the parameters that allow for a further minimization of σ. In the first model, such a minimization occurs when the parameter of frequency coupling tends either to zero or to infinite. In the second model, σ attains minimal values when the parameters associated with internal references of the oscillators are significantly higher than those associated with frequency coupling. We are lead to the conclusion that the best strategy, in evolutionary terms, to minimize the entropy generation in a long-lasting exchange of signals is to keep a high similarity (intra-specific) between the animals that communicate in a given environment. (AU)