In this talk, I will present two subfields of quantitative ecology that require modelling the population dynamics of one or several species in random environments: - Theory: what is the effect of increased variability in a forcing on population dynamics (e.g., can increased climate variability imperil bird populations?) - Statistics: how to estimate interactions between species using multivariate autoregressive models and time series of counts. This endeavour connects to the definition of statistical causality (e.g., Granger causality), debated in neuroscience & econometrics as well. For both themes, I will highlight both ongoing work in ecology and related fields, as well as opportunities for further research at the applied probability/statistics - ecology interface.