estar: An R package to measure ecological stability

1. Assessing ecological stability across populations or communities is a prime goal in biodiversity monitoring and conservation research. Quantifying stability is not trivial because its different aspects can be measured with various metrics. However, to date, no software enables measuring different stability metrics on ecological time-series data. 2. We present the estar R package that standardises and facilitates the use of ten established stability properties that have been used to assess systems’ responses to press or pulse disturbances at different ecological levels (e.g. population and community). 3. estar provides two sets of functions. The first set corresponds to functions that can be applied to univariate data, i.e., a time series of a system’s state variable (e.g., individual body mass, population abundance, or species richness). The metrics included in this set are: invariability, resistance, extent and rate of recovery, and persistence. The second set of functions can be applied to multivariate data represented by the time series of the abundances of all species in a community. The functions in this set measure the stability of a community at short and long time scales. In the short term, community’s response to a pulse (sudden) perturbation is measured by maximal amplification, reactivity and initial resilience (i.e. initial rate of return to equilibrium). In the long term, stability can be measured as asymptotic resilience and intrinsic stochastic invariability. 4. The package includes vignettes demonstrating the use of all functions and an introduction to the multivariate autoregressive state-space models necessary for the second set of functions. estar constitutes a toolbox with standardised, ready-to-use functions that bridge dichotomies in definitions and enable comparisons across state variables, taxa and scales.