CONICET | Buscador de Institutos y Recursos Humanos

In order to correctly apply fractal analysis to characterize a temporal locomotor pattern it is important to evaluate the presence of periodic oscillations and trends in the data time series. Also, to infer power-law (fractal) scaling a straight line in a double logarithmic plot must be observed. This linearity must not be assumed a priori but has to be established by the estimation of local slopes. The local slopes then have to be evaluated for consistency in a sufficient range. In this study we evaluate temporal locomotor patterns in two animal models (Japanese quail and mosquito larvae) that widely differ phylogenetically, in their living environments, ecology and mainly in their type of locomotion. In this work we evaluate the temporal locomotor pattern of 8 male and female quail in their home cage (43 x 41.7 x 46cm) during the 14 hours of day light, and of 7 Culex quinquefasciatus mosquito larvae in glass boxes (4 x 4 x 16cm, length x width x height) during 3 hours, by registering their locomotor activity at 0.5s intervals. Power spectrum analyses showed no periodic oscillations in any of the time series. Time series were analyzedwith Detrended Fluctuation Analysis (DFA). Trends found in data were eliminated with DFA of order 3. Clearly DFA and also the analysis of the frequency histograms of the duration of immobility and locomotor events showed a fractal pattern and long range autocorrelations. Interestingly, the temporal locomotor pattern of quail and mosquito larvae were clearly fractal; however, important differences were found in their organization, complexity and predictability. While quail showed the same fractal pattern independently of the temporal scale, larvae showed two distinct levels of organization depending on the temporal scale of observation. This is the first study that shows two levels of behavioral organization at different temporal scales in an animal species.

enviar mensaje