Artificial light at night (ALAN) and urban noise are increasing globally and can have a range of impacts on wildlife. While ALAN and noise often co-occur and can affect wildlife in similar ways, their impacts have generally been studied in isolation. Information about possible interactive impacts, which can be more serious, is critical to guide conservation. We studied how noise and ALAN impact a common urban waterbird (Eurasian coot Fulica atra) around the city of Melbourne in south-eastern Australia. We aimed to examine: (1) the individual and (2) interactive impacts of noise and ALAN on abundance, and (3) the relative influence of these stressors and other environmental predictors. To do so, we used data from a large-scale (1,463 surveys across an area of 9,250 kmwith significant heterogeneity in noise and ALAN conditions), long-term (2008-2018) monitoring program, overlaid with georeferenced noise and light data. We used generalized linear mixed effects models and boosted regression trees to model individual and interactive effects of ALAN and noise on abundance. Abundance was negatively correlated with noise and ALAN individually. Furthermore, the two stressors had a negative synergistic effect, ultimately resulting in the absence of coots at the highest observed ALAN and noise levels. We also estimate that the combined influence of the two stressors on abundance was larger than that of other examined environmental factors. Our findings that noise and ALAN have detrimental interactive impacts is worrying for two reasons. First, Eurasian coots are thought to be tolerant to urbanisation, so impacts may be more severe for less tolerant species. Second, noise and ALAN commonly co-occur around cities, so similar impacts are likely elsewhere. By adopting more biologically and ecologically realistic analytical frameworks, future studies can better estimate the cumulative impacts of multiple stressors to facilitate improved conservation and management. 2
