Abstract
In the programme ‘Programma Noordzee 2022-2027’, offshore wind farm (OWF) search areas will be designated as development areas for generating wind energy. An assessment was made of the effect of habitat loss due to offshore wind development in these search areas on the following seabird species: diver spec.2, northern fulmar, great cormorant, common eider, common scoter, common guillemot, razorbill and Atlantic puffin. For great cormorant, common eider and common scoter, the effects of habitat loss were assessed based on estimated numbers of casualties. For the other species, population models were used to assess the effect of habitat loss on the population level. For northern gannet and sandwich tern, both the effects of habitat loss and the effects of collision mortality were tested together in the population models. Here, the method and calculation of the numbers of casualties due to habitat loss are described. The estimation of the numbers of casualties due to collision mortality are described in Potiek et al. (2021a). The population effects of offshore wind farms on seabirds were considered for six different scenarios:
- No offshore wind farms
- Basic: All offshore wind farms planned up to 2030
- ‘Rekenvariant I’ (Basic+10.7 GW) ‘Rekenvariant II’ (Basic+12.7 GW)
- ‘Rekenvariant III’ (Basic+16.7 GW)
- International wind farms + Rekenvariant III
The assessment of the effect of OWF search areas on the ten seabird species considered here consisted of the following steps:
1. Data preparation of seabird observations at sea.
2. Calculation of density and dot maps for the seabird species, the choice of map type depends on the data availability.
3. Calculation of the number of casualties due to habitat loss based on the overlap between the bird maps and the wind farm search areas.
4. Development of new population models for northern fulmar and Atlantic puffin and an update of the parameters of the population models for the other five species.
5. Calculation of annual mortality probabilities due to habitat loss for all species. For the northern gannet, an additional estimation of annual mortality from habitat loss was made using an Individual Based Model.
6. Test of the population level effects of annual mortality due to habitat loss and collision mortality for northern gannet and sandwich tern and habitat loss only for the other species. Population level effects were tested against the Acceptable Level of Impact (ALI) thresholds (Potiek et al., 2021a), of which values were defined in a working document by LNV.
For great cormorant, common eider and common scoter, species that reside mostly in nearshore waters, the effects of habitat loss of the wind farm search areas seem negligible as the number of casualties does not increase for the scenarios with new OWF search areas compared to the scenario with the OWFs that are already planned. The estimated total annual numbers of casualties for great cormorant, common eider and common scoter were respectively 3, 4 and 7 for the national scenarios. Most habitat loss casualties in the national scenarios were predicted for common guillemot (annual totals of 588-1419) and razorbill (annual totals of 165-372), followed by northern fulmar (annual totals of 12-33). For the seven species for which population models were used, the number of casualties from habitat loss did not lead to violation of the ALI thresholds. The population growth rates of razorbill and common guillemot were predicted to be most negatively affected by habitat loss. The KEC assessment methodology is continuously in development and updated according to the most recent scientific knowledge. One of the main uncertainties in the methodology are the quality and spatial coverage of the data on bird observations, which vary between species and through time.
Furthermore, the method to predict bird densities could be improved, by using spatial statistical models that predict bird density as a function of relevant covariates. In addition, it is recommended to refine the current method such that it can quantify or qualify uncertainty in all of the assessment steps. Finally, the method to calculate the number of casualties from habitat loss needs to be better integrated with the newly designed ALI method. Ideally, a more sophisticated method based on a mechanistic understanding of habitat loss, should be developed. The estimates in this study give the best estimate possible based on the species distribution data and assessment framework available to us at the current moment and are based on the assumption that seabird distributions will not drastically change in the next 30 years.