Effects of displacement from marine renewable developments on seabirds breeding at the Isle of May

The project has produced a model which estimates the consequences of displacement and barrier effects on the time/energy budget of breeding seabirds.

4. Discussion

4.1. Displacement model

The model presented in this report represents a significant step forward towards understanding the implications of displacement and barrier effects of wind farms on seabirds breeding at SPAs. To our knowledge, this is the first attempt to model how breeding individuals in a seabird population use their foraging landscape, and, crucially, how this changes when a component of the population is displaced or has to travel round a wind farm development.

In all scenarios, the addition of the Neart na Gaoithe wind farm resulted in an increase in the average costs of foraging. This result is important since it suggests that displacement effects merit further consideration. The impact of displacement was driven by two main processes:

  • the increased travelling costs incurred by the subset of the population that is displaced or for which the wind farm forms a barrier to movement, and
  • the reduction in average prey densities in the remaining habitat due to intensified intra-specific competition, affecting not just displaced birds but the population as a whole.

Whilst there were some differences amongst the scenarios tested, the effect of the wind farm on time/energy budgets was consistent, suggesting that the displacement effect was apparent at different levels of prey aggregation and degrees of interference between individual guillemots.

Whilst the number of birds that did not achieve their daily energy requirements in the scenarios was comparatively small, the purpose of this project was to provide a proof of concept of the modelling approach. As such, formal examination of the absolute values is not justified since various aspects of the model are not realistic ( e.g. 24 hour duration, population density). Future work would focus on developing the approach to explore patterns across a whole season for the SPA population as a whole. The potential effects of displacement over these time scales are hard to predict from the values presented in the model, since the relationships are unlikely to be linear, and may involve a range of outcomes - see section 4.2 part 2).

Although we used the Isle of May guillemot population and the Neart na Gaoithe wind farm to showcase the model, the framework is extremely flexible and could readily be updated, for example to take advantage of improved data on foraging behaviour or prey distribution. Similarly it could be adapted for different seabird or prey species by incorporating appropriate information on foraging behaviour and distribution respectively. Moreover, this model could be adapted to situations where no empirical data are available ( e.g. flight direction could be modelled as random), subject to the appropriate caveats. The model could also be scaled up to the whole SPA, and incorporate interannual and seasonal variation including the effects of displacement on wintering birds. With further development, it could also explore outcomes for multiple species simultaneously, enabling inter- as well as intra-specific competition to be accounted for in the calculations. This modification is likely to be particularly insightful since seabird breeding colonies typically comprise several species and inter-specific facilitation and competition among multi-species feeding flocks are well known. Finally alternative scenarios of array location and design could be explored within this framework, and identifying designs that maintain energy balance above a defined threshold ( e.g. that maintained population level effects below a threshold as agreed in consultation with interested parties - see next section) could be particularly useful, while sensitivity analyses could be employed to identify minimum data requirements and highlight priorities for future monitoring. The model could also be adapted to other case studies, and is designed so that cumulative effects from multiple developments could be estimated, with a view of informing marine spatial planning as well as decisions on a case by case basis.

Changes in the time/energy budgets of breeding seabirds can have important population consequences. This is because such changes may impact on the body condition of adult breeders which, in turn, can affect breeding success (through abandonment of young), adult survival and, ultimately, population size. Additionally, breeding success may be affected directly if provisioning rates alter significantly. There is an urgent need to estimate these more realistic population consequences of displacement, to provide improved assessments of likely adverse effects on SPA populations. Below, we describe how the outputs of the model can be used to parameterise population models that estimate the population consequences of displacement.

4.2. Population consequences of displacement

The most appropriate method of estimating the population consequences of displacement is to link time-energy budget models of foraging with population models under a range of plausible scenarios of displacement (Figure 24).

Figure 24. Flow diagram illustrating the linking of time-energy budget and population models to estimate population consequences of displacement.

Figure 24. Flow diagram illustrating the linking of time-energy budget and population models to estimate population consequences of displacement

The framework can be split into three components:

1) Time-energy budget model: The time-energy model in the absence of a wind farm, presented in this report over a 24 hour period, would be expanded into a cumulative profitability surface estimated over the course of the breeding season. This seasonal model would be quantified in a range of environmental conditions from optimum to severe, based on the range of conditions experienced in the region and forecasted in climate models.

Sustained time/energy deficits may have consequences both for breeding success and adult survival, two critical drivers of population dynamics. Life history theory predicts a trade-off between investment in current breeding and self-maintenance (Ylönen et al. 1998). Available data on the relationships between energy balance and these two parameters would be utilised; alternatively, theoretical understanding would be used ( e.g. allometric scaling relationships). To breed successfully, one member of a pair of most species of seabird, including guillemots, needs to be constantly present at the nest site. Thus, increased time required for foraging can result in temporary unattendance of eggs or young which increases the likelihood of failure (Harris & Wanless 1997; Ashbrook et al. 2008). These fitness consequences of time and energy budgets underpin the subsequent displacement scenarios.

2) Consequences of displacement on breeding success and adult survival: The impact of changes in time/energy budgets on breeding success and adult survival would be estimated using the same approach as the baseline time-energy budget model. There are a number of challenges in reaching a satisfactory conclusion, given the lack of information on fitness consequences of foraging energetics and the effects of displacement on these parameters. Thus, where possible available empirical data would be used, or theoretical understanding together with experience of the species' ecology. At this stage, the most appropriate set of outcomes (for each set of environmental conditions) may be a matrix of severity against likelihood.

3) Population consequences of displacement. A stochastic time-specific matrix model (Caswell 2001; Frederiksen et al. 2008) using data on breeding success and adult survival and, where available, on age at first breeding, age structure and juvenile survival would quantify the population consequences of displacement. The modelling would be undertaken in three steps:

a. Retrospective analysis: existing time would be evaluated to assess how well they capture observed historical population trends, and what environmental variables correlate with demographic rates.

b. Forecasting population change: models would simulate future population growth rate. These simulations would be driven by change in breeding success and adult survival resulting from predicted environmental change or current trends where possible, or in the absence of these the distribution of historical parameter values.

c. Predicted impacts of displacement: using the forecasts for population change with no displacement provides a baseline population trend against which predicted impacts of displacement from energetic models can be compared. These forecasts can be used to identify required decreases to breeding success and survival necessary for causing pre-determined changes in total population size. Different scenarios for the impacts of displacement on breeding success and survival can then be related to these necessary changes. In this way, resulting population change can be evaluated, and displacement scenarios under which negative impacts on the integrity of the SPA network are predicted can be identified. These scenarios would be run for the expected lifespan of the wind farm and an agreed period afterwards to monitor post-closure population trajectories, as part of the EIA/ HRA process. Specifically, the time taken to recover from any decline as a result of displacement would be determined. Finally, the population consequences of displacement could be modelled alongside other potential effects such as collision to provide an overall assessment of wind farm impacts on populations.


Back to top