Bird population models

Bureau Waardenburg and Wageningen Marine Research developed species-specific (Leslie-Matrix) population models for use in studies of both collision risk and habitat loss for potentially critical species. The population models are based on the latest scientific knowledge available in the literature as well as expert judgment.

Collision risk
This study looked at collision victims from all offshore wind farms planned pre-2030 in the North Sea. Species of interest were selected in KEC 1.0: they include five gulls, two skuas, two terns, three species of wildfowl and one wader. Enough data were available to establish population models for all species except the Little Gull. These population models give a projected population trend based on the current demographic rates. For each species, the outcome of an impacted scenario (with planned wind farms) was compared with the outcome of a scenario without additional mortality (based on the current situation). In the scenario with additional mortality, the population models assume the number of victims as estimated KEC 3.0. The largest estimated impacts were found for the Lesser Black-backed Gull, Black Tern, Common Shelduck and Curlew. However, the overall impacts associated with the KEC scenario seem to be relatively minor for most species.

An additional scenario was run with the full annual Potential Biological Removal as a measure of additional mortality. This scenario shows how the maximum threshold of additional mortality used at present would impact the population of the selected species. It shows a relatively strong impact for most species, which suggests that this level of additional mortality is too high to maintain a stable population.

This project shows how population models can be used for a population-level impact assessment of mortality due to collisions of birds with turbines. Population models provide a better picture than other methods of the possible effects of offshore wind farms on these species. However, before the models can be used to inform the permit process, thresholds are needed for the metrics they can produce. This is a policy decision rather than a scientific one. The report is in the completion phase and should be published soon.

Habitat loss
A three-step approach was adopted to estimate the effects of bird habitat loss due to offshore wind farms. The effects of habitat loss were assessed for the Red-throated Diver, the Northern Gannet, the Sandwich Tern, the Razorbill and the Common Guillemot. In the first step, the spatial population distribution was estimated with habitat models based on seabird distribution data and abiotic explanatory variables. In the second, the cost of habitat loss was estimated with an energy-budget model for individual animals using the habitat model predictions and levels of displacement. The energy-budget model estimates changes in survival rates of the birds caused by the spatial redistribution of the bird population (in the context of a loss of foraging area as a result of offshore wind farms). Finally, the population models were used to estimate how the reduction in survival rates may affect the future development of the bird populations.

Two scenarios were examined for the effects of the displacement of the birds by wind farms: a best estimate and a worst case scenario. In the first, the best available estimate for the level of displacement was used; complete displacement was assumed in the worst case. Both scenarios were run for situations without wind farms, for Dutch wind farms only, for foreign wind farms only and for all wind farms combined.

^{Density map based on habitat suitability for Common Guillemot (WMR)}

The habitat suitability maps for three of the five species (Northern Gannet, Razorbill and Common Guillemot) covered the entire North Sea. Given data limitations, the habitat suitability maps for the Red-throated Diver and Sandwich Tern were limited to the Dutch Continental Shelf. The overlap between the bird populations and Dutch offshore wind farms was generally found to be low (0.5-2% overlap, depending on the species). The outcomes for affected and unaffected populations were found to be very similar in broad terms: it is therefore unlikely that the effects of displacement by wind farms will be distinguishable from normal fluctuations in the populations (resulting from either natural and/or anthropogenic causes). The strongest effect was found for the Northern Gannet population, probably because the birds are confronted with wind farms in both the breeding and non-breeding seasons: the other species make only seasonal use of the areas where wind farms are planned.
 
The simulation modelling approach provides an estimate of bird displacement but there are still large sources of uncertainty that may influence the outcome. The results should be seen as an example of what can be achieved with this approach but definite conclusions cannot be stated at present. The report is in the completion phase and should be published soon.