Minimizing errors in identifying Lévy flight behaviour of organisms.

Area of Science:

• * Movement ecology and theoretical ecology
• * Statistical physics and biophysics

Background:

• * Lévy flights are random walks with scale invariance and superdiffusivity, applied in optimal foraging theory.
• * These flights, characterized by power-law distributed movement lengths, are observed in diverse organisms.
• * Accurate estimation of the Lévy exponent (micro) is key for detection, but methods vary.

Purpose of the Study:

• * To investigate how different plotting methods affect the accuracy of Lévy exponent estimation.
• * To identify the most reliable method for detecting Lévy flight behavior in empirical data.
• * To assess the impact of sample size on the precision of Lévy exponent estimation.

Main Methods:

• * Simulations were used to test various plotting methods for analyzing movement step lengths.
• * Methods evaluated include log transformation, 2(k) logarithmic binning (with and without normalization), and cumulative distribution functions.
• * Empirical data from published studies were reanalyzed using the identified optimal method.

Main Results:

• * The 2(k) logarithmic binning method with prior normalization yielded the lowest error (1.4%) in identifying Lévy flights.
• * Simple log transformation underestimated the Lévy exponent by approximately 40%.
• * Increasing sample size significantly reduced variation in estimated Lévy exponents, enhancing accuracy.

Conclusions:

• * The choice of plotting method critically impacts the accuracy of Lévy exponent estimation and biological interpretation.
• * The 2(k) binning method with normalization offers a reliable approach for detecting Lévy flight behavior.
• * Utilizing a large number of steps in movement analysis improves the accuracy of detecting optimal Lévy flight patterns.