Further development of the ‘Acceptable Level of Impact’ framework for effects of offshore wind farms on seabirds : extended method development and sensitivity analysis

The Acceptable Level of Impact (ALI) methodology defines acceptable limits for the predicted population effects of mortality imposed by offshore wind farms (OWFs) for marine birds. These population effects are quantified using stochastic population models tailored for specific seabird species. The ALI is defined as: ‘The probability of a population decline of X% or more, 30 years after the impact, cannot exceed Y’. In this definition, the X threshold determines the reduction in population abundance due to OWFs that is considered acceptable, evaluated 30 years after the onset of the impact. The Y threshold of the ALI is derived from a threshold value for the ‘causality probability’ (PC). PC is the probability that a violation of X is caused by the impact from OWFs, instead of being caused by variability or uncertainty inherent to predictions of future population abundance. The ALI methodology was reviewed in November 2021 – January 2022. Reviewers concluded that the approach and assumptions of the proposed ALI methodology were a considerable improvement in relation to existing impact-evaluation frameworks (ORNIS, PBR). However, reviewers also raised concerns about the legal tenability and definition of the causality level PC. Several points put forward in the reviews were addressed within the current report, namely a derivation of the causality threshold PC and results of a sensitivity analysis of the ALI methodology, which contained several components. First, an analysis of the effect of uncertainty on the outcome of the ALI was performed. This analysis was done by: 1) changing the way in which parameters of the population model were sampled to simulate year-to-year variation in demographic rates and, 2) changing the standard deviation of the parameters of the population model. Second, the effect of changing the X threshold value on the outcome of the ALI was evaluated. Lastly, the effect of time-varying, as opposed to constant, mortality levels on the outcome of th