Can population time-series data be combined into a single data structure? Insights from developing the Popler database

Have you ever wanted to synthesize data from hundreds of ecological population abundance time-series? The first step of such analysis would be to put all datasets in the same format, but this seems impossible to do without mayor loss of information: ecological time-series data are incredibly heterogeneous. Here, we show that most population time-series share characteristics that allow devising a common data structure capable of retaining most of the original information. To illustrate these characteristics, we use examples from our work in developing a publicly available database of population time series – popler.