The active dynamics of biological systems represents the paradigmatic example of an efficient information transfer, which often leads to the emergence of collective behavior. Despite the proliferation of models and theories that aim to reproduce such behaviors, the secret behind the mechanism of collective behavior is still not totally understood.
Within this context, can field experimental data make the difference?
In this talk, I will present two examples where experimental data and advanced techniques were fundamental to validating and developing novel theories. Features found in experimental data may be used to discriminate between realistic and unrealistic theories. This is the case with modeling the mechanism of controlling the speed in starling flocks. We validated a novel theory based on marginal speed confinement showing that, unlike other models, it is able to reproduce the scale free correlation and the low fluctuations of the speed that we found on 3D data. But data cannot be relegated to just verifying models. Data themselves may drive the development of new theories. With an unprecedented set of 3D data, we could estimate the dynamic exponent of natural swarms of midges. The value of 1.37, obtained with enormous effort both in terms of data collection and data analysis, suggested that in the ferromagnet approximation (dynamic exponents = 2), we were missing crucial ingredients: the high activity and the underdamped inertial relaxation that we found in natural swarms. It is actually by introducing these two factors, in the RG calculation, that we were able to predict a dynamic exponent of 1.34 compatible with natural swarms, which definitively proves the power of RG for the quantitative description of collective behavior.