Scottish Marine and Freshwater Science Volume 5 Number 13: Population consequences of displacement from proposed offshore wind energy developments for seabirds breeding at Scottish SPAs

Report on a project which aimed to develop a model to estimate the population consequences of displacement from proposed offshore wind energy developments for key species of seabirds breeding at SPAs in proximity to proposed Forth/Tay offshore wind farm d

Executive Summary

  • The aim of this project was to develop a model to estimate the population consequences of displacement from proposed offshore wind energy developments for key species of seabirds breeding at SPAs in proximity to proposed Forth/Tay offshore wind farm developments.
  • The steering group identified five seabird species for which displacement modelling was required in support of HRA/ AA for Forth/Tay developments: black-legged kittiwakes Rissa tridactyla; common guillemot Uria aalge; razorbill Alca torda; Atlantic puffin Fratercula arctica; northern gannet Morus bassanus.
  • The steering group agreed that the SPAs to be considered in this report were Buchan Ness to Collieston Coast SPA, Fowsheugh SPA, Forth Islands SPA and St Abb's Head to Fastcastle SPA.
  • We considered impacts of displacement on population size operating via two main processes: reduced survival of offspring during the breeding season, and reduced body mass of adults leading to lower survival in the following winter.
  • Displacement effects are of two main types: the effects of displacement of birds that intended to forage in the wind farm, and the effects of the wind farm acting as a barrier to movement of birds intending to forage beyond the wind farm.
  • The principal requirement was to develop time and energy models of foraging to estimate consequences for demographic rates. A simulation model was developed that modelled the time/energy budgets of breeding seabirds during the chick-rearing period. The model was parameterised from information available in the literature or, where this was unavailable, from expert judgment. The model simulated foraging decisions of individual seabirds under the assumption that they were acting in accordance with optimal foraging theory. Each individual selected a suitable location for feeding during each foraging trip from the colony based on bird density maps derived from fitting a generalized additive model ( GAM) to empirical location data (obtained from GPS tracking) for each species. Subsequent behaviour of birds was then simulated incorporating realistic assumptions and constraints derived from observed behaviour. Fundamentally, we assumed that the foraging behaviour of individual seabirds was driven by prey availability, travel costs, provisioning requirements for offspring, and behaviour of conspecifics.
  • Choice of foraging location was dependent upon assumptions regarding the spatial distribution of prey, for which we had no direct empirical data. Therefore two scenarios were considered: an assumption that prey was distributed uniformly throughout the foraging range ("homogeneous" prey), and an assumption that prey was distributed proportional to the estimated distribution of birds after adjustment to account for the effects of availability due to proximity to colonies ("heterogeneous" prey).
  • Baseline simulations from the model, in the absence of wind farms, were parameterised based on empirically estimated values for foraging time, flight time, adult body mass and chick survival from studies of these or closely related species from CEH's long term study of seabirds on the Isle of May or elsewhere.
  • The impacts of the proposed wind farms were assessed by comparing simulated values of adult and chick survival in models that included the wind farms against the baseline simulations. Impacts were assessed separately for each of the proposed wind farms (Neart na Gaoithe, Inch Cape, Round 3 Alpha, Round 3 Bravo) and for the cumulative effect of all four wind farms.
  • Models were initially run using relatively small numbers of simulated birds (1000 per species) for a relatively large number of different scenarios (66). The scenarios reflected possible assumptions regarding food availability (good, moderate or poor), the spatial distribution of prey (homogeneous or heterogeneous), the percentage of birds affected by barrier and displacement effects (assumed to be 100% when looking at individual wind farms, but scenarios of 50% displacement/50% barrier; 0% displacement/100% barrier; 100% displacement/0%barrier were considered when looking at cumulative effects) and the effect of the width of the buffer around the wind farm that is included within the wind farm footprint (values of 0km, 0.5km and 1km were considered for Neart na Gaoithe; values of 1km were used for all other wind farms).
  • These exploratory simulations helped to identify those scenarios and SPA-by-species combinations that were of greatest interest. They also identified the fact that a substantial amount of uncertainty was introduced by running the models using relatively small numbers of birds. The final simulations therefore used much larger sets of simulated birds (20000 rather than 1000), but, in order to prevent the computational cost becoming prohibitive, focused on a smaller number of scenarios (ten). These scenarios involved looking at the effect of homogeneous and heterogeneous prey for each of the four individual wind farms ('full' and 'fast' models, see below) and for all four wind farms in combination ('full' model only). The final simulations assumed moderate food availability, a 1km buffer around each wind farm, and that 60% of birds experienced displacement and barrier effect (except for kittiwake, where the percentage was assumed to be 40%).
  • The final simulations used both a "full" and "fast" version of the foraging model. The primary benefit of the "fast" model was the fact that the energetic consequences of barrier effects were included in a more realistic way. The fast model was also less computationally expensive, and was therefore used in performing a sensitivity analysis. Some aspects of the fast model are less biologically realistic than the full model, however, so the final estimates of wind farm effects were based on combining the "full" and "fast" model outputs in a way that captured the strengths of both models.
  • The simulations indicated three species-by- SPA combinations for which declines in adult survival of more than 0.5% seemed to be possible - Forth Island kittiwakes, Fowlsheugh kittiwakes, and Forth Island puffins - with the largest declines corresponding to the two combinations involving the Forth Islands SPA. The results did not show evidence of declines in adult survival of more than 0.5% for gannets, razorbills or guillemots, or for kittiwakes at St. Abbs or Buchan Ness. The declines in kittiwakes at Forth Islands appeared to be driven by all four wind farms, but with the largest impact coming from Neart na Gaoithe. The declines in kittiwakes at Fowlsheugh appeared to be primarily driven by Alpha. The results for Forth Island puffins were sensitive to assumptions about the distribution of prey: if prey were assumed to be spatially homogeneous then the estimated declines were larger than for any other species- SPA combination, and driven primarily by Inch Cape and Alpha. If prey were assumed to be heterogeneous - and linked to the distribution of birds - then the estimated declines were much smaller and were primarily driven by Neart na Gaoithe. Estimates of cumulative effects were generally approximately equal to the sum of effects from individual wind farms.
  • The results for breeding success were qualitatively similar, but were generally of lower magnitude (if we assume that a 1% decrease in adult survival is roughly equivalent, in demographic terms, to a 5% decrease in chick survival). Breeding success results were also harder to disentangle from the effects of stochastic noise, probably due to threshold effects in the model relating to the consequences of nest unattendance by adults on offspring survival. Only cumulative estimates for Forth Island puffins with homogeneous prey showed a decrease in chick survival of more than 2.5% and none showed an estimated decline of more than 5%.
  • This study is, to our knowledge, the most comprehensive assessment of the population level consequences of displacement for seabirds to date. Displacement effects have been considered to potentially impact on chick survival. What has been less widely appreciated is that impacts on adult survival are also possible, mediated via changes in body condition. Declines in adult and chick survival were recorded for some species/wind farm/ SPA combinations that matched expectations in terms of foraging range, foraging costs and wind farm location relative to SPAs. The model had to make a number of assumptions that would benefit from parameterisation with local data, in particular prey distribution, behaviour of seabirds in response to wind farms (including habituation) and effects of adult body mass change on subsequent survival.


Back to top