Improving Our Understanding of Seabird Behaviour at Sea

This project collated tracking data from five seabird species thought to be vulnerable to offshore wind farms. These data were analysed to understand whether seabird distribution data, usually undertaken in daytime, good weather conditions, were representative of behaviour in other conditions.


Offshore wind farms form a key part of the Scottish Government's strategy to mitigate the impacts of climate change by generating 100% of electricity using renewable energy. However, Scotland also hosts internationally important populations of seabirds during the breeding season and it is important to ensure any offshore wind farms do not adversely affect these populations. At present assessments of the likely impacts of offshore wind farms on seabirds are largely based on data collected using boat and/or digital aerial surveys. However, such surveys are constrained by light levels and weather conditions in when they can be carried out. This leads to concerns that the data used in assessments may be biased towards favourable conditions and may not accurately reflect the conditions experienced by birds. Furthermore, understanding the behaviour of seabirds at sea is key to understanding the potential exposure of seabirds to impacts such as displacement and barrier effects. However, the behavioural data that can be collected using standard survey approaches is extremely limited.

The widespread application of tracking technologies offers an opportunity to investigate the behaviour of seabirds at sea in more detail. We collate tracking data collected from five seabird species thought to be vulnerable to the impacts of offshore wind farms – Northern Gannet, Lesser Black-backed Gull, Black-legged Kittiwake, Common Guillemot and Razorbill – from colonies from across the UK. We analyse these data in relation to the diel cycle and weather conditions in order to understand how seabird distributions may vary between conditions in which traditional survey methods can and cannot be applied. We further analyse these data using Hidden Markov Models in order to classify these data into one of three behavioural states – floating, commuting and foraging. We investigate the spatial patterns in these different behaviours and consider how they may be influenced by the diel cycle and weather conditions.

Our analyses highlight that, during the breeding season, the constraints related to traditional surveys are unlikely to mean data are biased towards particular conditions. However, our analyses also highlighted clear spatial patterns in seabird behaviour at sea. We discuss the implications of these spatial patterns for the assessment of the impacts associated with offshore wind farm.



Back to top