This report presents data on monitoring investigations of birds carried out during spring and autumn 2003 in relation to the Nysted offshore wind farm at Rødsand.
Information presented covers the spring migration period of 2003 during which the wind farm was under construction. Observations therefore reflect the reactions of migrating birds to the presence of foundations and the support ships in the vicinity.
Data are also presented for autumn 2003 by which time all the turbines had been erected and were in operation.
Migration routes were mapped using the same techniques as those used during the base-line investigations of 1999-2002, combining radar techniques by day and night with specific species identification during daylight hours using telescopes. Radar tracks were entered to a GIS platform to compare the base-line with subsequent monitoring results. Emphasis was placed upon three key variables:
- the orientation of autumn migration routes for waterbirds and terrestrial species to measure potential avoidance responses and response distances,
- the probability that waterbirds will pass through the wind farm area during autumn and spring, to measure waterbird responses to the entire wind farm,
- migration intensity, measured by the number of bird flocks that pass the eastern and northern edge of the wind farm area, to measure the effect of avoidance responses on the volume of migration within the wind farm area.
Comparisons of these key variables between individual base-line years were undertaken by controlling for various factors such as weather conditions, season and time of day using multi-factor ANOVA and regression analyses.
In addition, the distribution and abundance of waterbirds, which included spring migrants and wintering birds, were monitored in and around the wind farm by sequential aerial surveys as during the base-line years.
The wind farm area is situated on a major waterbird migration route, used by up to 300,000 individuals in autumn. The intense migration of waterbirds was confirmed by base-line observations in 1999-2002 and during the construction and operational phases in 2003.
Autumn migrating waterfowl showed significant differences in their mean orientation within the approaching area of the wind farm between all four years of investigation. The analyses of the orientation of individual bird flocks in relation to their distance from the wind farm showed that the year-effect differed across years dependent on the distance from the wind farm. Due to small sample sizes and certain wind conditions the wind effects found in the baseline studies could not be incorporated into the 2003 analyses. It was therefore not possible to demonstrate a convincing change in migration orientation at a specific distance from the wind farm following construction of the wind farm. However, the standard deviation of migration orientation increased significantly during the daytime at distances closer than 3000 m to the wind farm in 2003 and closer than 1000 m during the night. These results support the hypothesis that migrating birds show a response to the wind farm, specifically reacting by increased lateral avoidance to the north and south of the wind farm. They also conform to the predictions under the hypotheses that (i) the deflection will occur close to the wind farm and (ii) that the deflection will occur closer to the wind farm at night than during the day.
Observations in autumn 2003 offered no support for a severe avoidance response to the wind farm, in terms of a substantial reverse migration of birds turning back eastwards from the eastern edge of the wind farm compared with the base-line.
Base-line studies showed that between 24% (2002) and 48% (2000) of tracks registered in autumn by radar passed the eastern border of the proposed wind farm area. After the wind turbines were erected in 2003, significantly less (9%) tracks of waterbird flocks registered by radar passed the eastern border. This result was confirmed controlling for the effects of cross-winds, time of day (4-7% by day compared to 11-24% by night) and latitudinal position.
Generally, the major spring migration route of waterbirds lies north of the wind farm area. During spring 2003, 11% of all migrating waterfowl tracks passed the eastern edge of the wind farm area, less than in 2001 (16%) and 2002 (25%), but the difference was not significant during the day. Hence, during daylight hours there was no support for the hypothesis that birds avoided the wind farm area during the construction phase in spring 2003.
Data relating to migrating land birds are not discussed in any great detail in this report. Temporary suspension of the studies inautumn 2002, extremely low spring migration intensities of raptors, pigeons and passerines in all years provide insufficient information for a full analysis of the effects on these species during construction and operation of the wind farm.
Waterbird migration intensity within the wind farm area varied considerably with weather conditions both locally and on a flyway scale making predictions at a local scale difficult to model, and statistical comparisons complex. Nevertheless, the results from spring and autumn 2003 clearly demonstrated reduced intensity of migration in the wind farm area, based on density of radar tracks. These responses were undoubtedly partly the results of the "shadow effects" cast on the radar screen by individual turbines. However, because many of the tracks appear beyond such turbines shadow and emerge from the outer side of the wind farm, the intensity of migration in the area between the erected turbines was considerably less than at outside and compared with base-line years. This is confirmed from individual tracks identified to species, which showed that many flocks and individuals take avoidance action and fly around the park without ever venturing between the turbines. Furthermore, those that do continue into the wind farm adjusted their flight trajectories and tended to fly down the visually clear corridors between the rows of turbines. Both these features combine to explain the overall reduction in migration track densities within the wind farm.
Despite general support for the hypotheses outlined above, it is important to stress that these results provide little evidence for or against the effects of the construction of wind turbines on migrating waterbirds. The data were collected in just one year and construction phase of the turbines extended over a relatively short period. It is therefore difficult to draw many reliable conclusions from the single case study. Although the results suggest substantial avoidance (and provide data on the nature of that avoidance) by autumn migrating waterbirds of the newly constructed wind farm, it is important to stress that these results come from one single monitoring year. They are gathered under the particular conditions prevailing in that year and before any likely effects of longer-term habituation or other behavioural responses to its presence.
Four aerial surveys of staging and wintering birds were conducted in the study area in 2003, one in each of the months of January, March, April and December. Thus a total of 25 surveys has been performed since August 1999.
The main construction phase of the Nysted windfarm was defined to commence by January 2003 and the operational phase by August 2003. Thus the three early surveys of 2003 represented the construction phase data set, while the December survey comprised the first data set from the operational phase of the wind farm.
At the time of the January 2003 survey ice cover was observed in large parts of the survey area, causing changes in bird distribution patterns. Therefore data from this survey were excluded from part of the data analyses.
The most numerous species recorded in 2003 were Tufted Duck (12,205), Eider (3,142), Mute Swan (2,882) and Long-tailed Duck (2,797). Of these only Eider and Long-tailed Duck occurred frequently in the offshore areas.
To ensure maximum compatibility between base-line data and construction data only data from March and April of the base-line phase were used to analyse construction activity impact on bird distributions. Given the small number of surveys during construction, no firm conclusion can be drawn about the construction phase. However, long-tailed duck and eider showed reduced preference for the wind farm area during construction, whereas the relative number of herring gulls increased slightly in the wind farm area.
Since only one survey exists during operation of the wind farm results must be supported by further surveys before conclusions concerning habitat loss for staging and wintering birds at Nysted Wind Farm can be drawn.