Predicting population sizes and dynamics of mobile organisms is important for trend estimation, but this is difficult when the origin of individuals cannot be determined, i.e. residents and migrants are indistinguishable in the field.

Here we examine fluctuations in populations during the non-breeding season (autumn to spring) of seven waterbird species, which co-occur on 122 neighbouring water bodies in south-west Germany. We asked whether site-level abundance patterns can be predicted by weather variables, and whether these effects vary over months and years. We used a spatially explicit hidden Markov model to estimate the effects of local and regional movement on population fluctuations.

Although the species varied in their dynamics, with different amounts of movement estimated between sites, several patterns were common across species: density effects were highly month-specific, but with little variation in the strength of effects over years. The abundance of most species was positively related to temperature, especially in winter.

The common teal Anas crecca and common pochard Aythya ferina were the most site-fidelic species once they were present in the study area, but also exhibited the strongest regional migration. The mallard Anas platyrhynchos, tufted duck Aythya fuligula and great crested grebe Podiceps cristatus each behaved more like a single population, as individuals more frequently moved between sites and abundance fluctuations at sites were not explained by migratory movements alone.

Our study shows that the strength of population parameters and environmental forces can be decomposed into monthly and yearly effects. Estimating the unknown origin and movement of individuals may show that commensurate populations of mobile species may have different underlying dynamics, while responding similarly to environmental factors.