Quantifying and explaining synchrony in ecological fluctuations is a major problem. Spatial synchrony is of interest because of its consequences for the behavior and stability of metapopulations. Identifying driving mechanisms by exploiting the statistical power of data drawn from many locations (or pairs of locations) enables us to understand the causal mechanisms underlying these fluctuations. Synthesizing causes and consequences may enable us to make better predictions about the effects of a changing environment on our ecosystems. The causes of spatial synchrony in ecology have historically been addressed in two scale-dependent ways. One way involves the comparison of temporal spectral characteristics in data (such spectra are commonly ‘reddened’), to identify possible driving mechanisms by their timescale of action. Another way involves the use of correlation-distance curves to study the spatial scale of synchrony effects. For data that are periodically sampled (e.g. on a regular grid or transect, and/or recorded at regular intervals in time) simple Fourier transform methods are available to interrogate scale-dependent information in the data directly.
Results/Conclusions
We show that the spatial Fourier power spectrum of measurements made at a particular point in time encodes the information about scales of variability which produces the correlation-distance curve. We further show that as with the temporal Fourier transform, Fourier phase information is necessary to show a statistically meaningful relationship between two variables, and that this provides the most valuable information about scale-dependent mechanisms which produce observed patterns. Using Fourier-based techniques we can identify drivers of synchrony by the spatio-temporal characteristics of the fluctuations they produce in ecological variables. As mechanically sampled gridded data becomes more widely available (e.g. from satellite measurements) these techniques should become vital to understanding the scale-dependent dynamics of ecosystems.