Figure 1.
Survey effort as area (km2) in each 3 km×3 km cell of the prediction grid covering the Seabird Sensitivity Mapping in English Waters study area.
a) JNCC European Seabirds at Sea (ESAS) boat surveys in summer months (April to September inclusive) from 1979 to 2011 and b) in winter months (October to March inclusive). c) WWT Consulting aerial surveys in summer months (April to September inclusive) from 2001 to 2011 (excluding Round 3 wind farm data) and d) in winter months (October to March inclusive).
Figure 2.
Predicted densities of gannets in summer from DSM of ESAS boat and WWT Consulting aerial survey data, with survey observations shown.
DSM used x, y and cdist covariates and a soap film smooth. Note the observations do not take account of coverage so seemingly high use areas may still have low predicted densities if coverage was also high.
Figure 3.
Predicted densities of red-throated divers in winter from DSM of ESAS boat and WWT Consulting aerial survey data, with survey observations shown.
DSM used x, y and cdist covariates with a soap film smooth used only in Liverpool Bay due to the difficulty in locating knots very close to the coast. Note that the model produced unfeasible predictions further offshore where there was no aerial survey coverage which provided most of the red-throated diver records (compare CV map, Figure 5).
Figure 4.
Coefficients of Variation (CVs) of predicted summer densities of gannets from DSM of ESAS boat and WWT Consulting aerial survey data, with survey observations shown.
Data with CVs above 0.5 were excluded from sensitivity mapping. Note the generally higher CVs in Liverpool Bay which was modelled separately to the other areas.
Figure 5.
Coefficients of Variation (CVs) of predicted winter densities of red-throated divers from DSM of ESAS boat and WWT Consulting aerial survey data, with survey observations shown.
Data with CVs above 0.5 were excluded from sensitivity mapping. Note CVs are considerably higher in the areas not covered by aerial surveys which provided most of the red-throated diver data.
Figure 6.
Wind farm sensitivity maps from SeaMaST.
The maps were produced by using highest densities from either the boat or aerial density predictions where the CV was less than 0.3 and excluding predictions with CVs higher than 0.5. The natural log of the density (plus one) was then multiplied by each species wind farm collision sensitivity or displacement score and the resulting value summed across species in each 3 km×3 km grid cell. Note where neither dataset had predicted densities with CVs <0.5 the resulting score is exactly zero and highlights areas where across all species coverage and model fits were poor. Summer and winter maps use the same scale.
Table 1.
Scores used in assessing sensitivity of seabird species to collision and displacement/disturbance risks from offshore wind farms in English territorial waters.
Table 2.
Scores for species’ population vulnerability to collision mortality at offshore wind turbines, with species ranked by overall score.
Table 3.
Scores for English territorial waters marine bird species’ population risk due to displacement by offshore wind farms, ranked by species score.