Scottish Marine and Freshwater Science Volume 5 Number 12: Strategic assessment of collision risk of Scottish offshore wind farms to migrating birds

6. Conclusions

6.1. The collision estimates for non-seabird species on migration through Scottish offshore wind farms are typically very small in relation to their populations. The estimates have been made on the basis of comparatively little species specific information regarding migratory movements and flight heights, and therefore several of the assumptions used are likely to have been conservative ( i.e. we expect to have over-estimated collision rates).

6.2. Therefore, at a strategic level the populations of non-seabird species which pass through Scottish waters do not appear to be at risk of significant levels of additional mortality due to collisions with Scottish offshore wind farms.

6.3. The collision estimates for seabirds are higher than those for non-seabird species. This reflects the narrower migration routes used in this analysis which lead most species to cross several wind farms in relatively high numbers. Not surprisingly, the species with the higher levels of mortality are those which typically fly higher: gull species and gannet. An indicative threshold value of 1% of the passage population was used to highlight species of potential concern at the strategic level (it should be noted that this 1% value has been presented to guide interpretation and does not have a specific biological basis). Using the Option 1 collision model with the far offshore coastal strip and uniform flight distribution six species had maximum total collision estimates which exceeded 1% of the passage populations: cormorant (1.2%); common gull (1.2%); lesser black backed gull (1.4%); herring gull (1.7%); great black-backed gull (1.9%) and common tern (1.03%). When Option 1, far offshore negative binomial flight distribution was applied five species had maximum total collision estimates which exceeded 1% of the passage populations: cormorant (1.1%); common gull (1.6%); lesser black backed gull (2%); herring gull (2.3%)and great black-backed gull (2.7%). However, using the Option 3 collision model only two species had maximum total collision estimates which just exceeded 1% of the passage populations: great black-backed gull (1.3%) and herring gull (1.1%) and in both cases these mortality levels were only obtained from one of the six models (negative binomial and far-shore). Both species had much lower estimated collision rates for the other migration corridor combinations While the option 3 collision model is accepted as a more realistic representation of bird interactions with wind turbines, outputs from this model are dependent on the distribution of flight heights used and methods for estimating seabird flight heights remains an area of active research. The 1% of the passage populations has been agreed as an indicative tool and not a threshold.

6.4. It should be noted that the use of Forrester et al. (2007), which is the most recently available comprehensive dataset for UK seabirds, does have limitations due to the age of the dataset. Whilst numbers impacted could be proportional to population scales there is the risk that where populations have increased e.g. northern gannet or decreased e.g. black-legged kittiwake the potential impacts could be under or overestimated.

6.5. Therefore, at a strategic level only ( i.e. not for EIA or HRA purposes at a project or CIA level), populations of almost all seabirds on migration through Scottish waters do not appear to be at risk of significant levels of additional mortality due to collisions alone with Scottish offshore wind farms. The possible exceptions identified by this study are large gulls, cormorant and common tern. However, these species are already recognised as ones which warrant detailed development specific assessment and this result is therefore not surprising. Furthermore, this result is likely to be precautionary due to the need to apply large scale simplifications suitable for this strategic level analysis.