Offshore wind farms and their effects on birds

Abstract

Exploiting wind energy at sea offers an attractive source of renewable energy avoiding problems on land, but what are the consequences for birds? We review the Danish and European experience of offshore (i.e. marine) windfarms and the effects and impacts which we consider they may have on birds, primarily through barriers to movement, displacement from ideal feeding distributions and collision mortality. We use case studies to demonstrate examples of displacement effects among species such as Red-throated Diver Gavia stellata, Common Scoter Melanitta nigra and Long-tailed Duck Clangula hyemalis but are unable to determine their causes or whether these patterns have population level impacts, assessment of which remains a major challenge. There is accumulating evidence for widespread avoidance of offshore turbines by large-bodied birds at macro-, meso-and micro-scales, but we accept that our knowledge for smaller bird species is less adequate. We conclude that careful siting during the planning phase can avoid a multitude of potential conflicts with avian populations and that despite generally inadequate post-construction monitoring (especially during periods of unusual weather), experience shows low levels of collision rates, especially among long-lived large-bodied bird species considered most at risk. We lack any understanding of the impacts of barrier effects and displacement from favoured feeding areas, but on a single project basis, these impacts to date are considered insignificant at the population level because of the relatively small numbers of birds so affected. Based on experiences from multiple single site studies, it is essential that site specific impact assessments continue to be undertaken to establish the potential effects and impacts of each project development. However, we also urge a more strategic national and international approach to identification, assessment and selection process for sites for potential future development of offshore windfarms. Despite low-level impacts on an individual windfarm basis, cumulative impacts of multiple offshore windfarm development (especially spanning the length of population flyways) have yet to be adequately determined. Developing effective mechanisms to deliver such assessments remains an urgent requirement for the immediate future.

Full text

Oshore wind farms and their eects on birds
Anthony D. Fox AnD Ib KrAg Petersen
Dansk Orn. Foren. Tidsskr. 113 (2019): 86-101
(Med et dansk resumé: Havvindmøller og deres påvirkning af fugle)
Abstract Exploiting wind energy at sea oers an attractive source of renewable energy avoiding problems on land, but what are
the consequences for birds? We review the Danish and European experience of oshore (i.e. marine) windfarms and the eects
and impacts which we consider they may have on birds, primarily through barriers to movement, displacement from ideal feeding
distributions and collision mortality. We use case studies to demonstrate examples of displacement eects among species such as
Red-throated Diver Gavia stellata, Common Scoter Melanitta nigra and Long-tailed Duck Clangula hyemalis but are unable to deter-
mine their causes or whether these patterns have population level impacts, assessment of which remains a major challenge. There
is accumulating evidence for widespread avoidance of oshore turbines by large-bodied birds at macro-, meso- and micro-scales,
but we accept that our knowledge for smaller bird species is less adequate. We conclude that careful siting during the planning
phase can avoid a multitude of potential conicts with avian populations and that despite generally inadequate post-construction
monitoring (especially during periods of unusual weather), experience shows low levels of collision rates, especially among long-
lived large-bodied bird species considered most at risk. We lack any understanding of the impacts of barrier eects and displacement
from favoured feeding areas, but on a single project basis, these impacts to date are considered insignicant at the population level
because of the relatively small numbers of birds so aected. Based on experiences from multiple single site studies, it is essential
that site specic impact assessments continue to be undertaken to establish the potential eects and impacts of each project de-
velopment. However, we also urge a more strategic national and international approach to identication, assessment and selection
process for sites for potential future development of oshore windfarms. Despite low-level impacts on an individual windfarm basis,
cumulative impacts of multiple oshore windfarm development (especially spanning the length of population yways) have yet to
be adequately determined. Developing eective mechanisms to deliver such assessments remains an urgent requirement for the
immediate future.
Introduction
The unexpected pace of climate change and the corre-
sponding search for renewable energy resources to re-
duce CO2 release into the atmosphere have fuelled the
rapid development of wind energy, especially within Eu-
rope. Renewable energy supplied 30% of Europe’s elec-
tricity in 2017, of which 54.6% was provided by wind po-
wer (Agora Energiewende & Sandbag 2018). Twenty per
cent of this installed wind power generation is situated
oshore (Pineda & Tardieu 2018) and a further 25 GW of
oshore capacity is projected by 2020 in Europe (com-
pared to 15.7 GW currently in operation; Pineda 2018).
Land-based windfarms cause impacts to the visual and
sound landscapes, to birds, bats and the human and na-
tural environment in general, while the wind characteri-
stics of the sea oers higher wind speeds and lower tur-

87
Oshore wind farms and birds
bulence levels more suited to sustained and consistent
electricity generation (Esteban et al. 2011). Despite the
greater engineering and economic challenges of gene-
rating power in the sea (because of the extra complexity
and costs of foundations, energy collection networks,
construction and operation in the generally more corro-
sive marine environment), the benets from such clean
renewable energy generation to society, largely out of
sight of land, are therefore likely to be great. But what
of the hidden cost to bird life? The construction of large
aggregations of tall, solid infrastructures with large and
rapidly moving rotor blades in the marine environment
constitutes a series of novel threats to birds, which have
been used to an empty ocean, safe from such threats.
Denmark was the rst country in the world to con-
struct eleven 450 kW wind turbines in the sea o Vinde-
by, Lolland, in 1991, followed by ten 500 kW turbines at
Tunø Knob in Aarhus Bay in 1995, the latter subject to
considerable avian impact assessment (e.g. Guillemette
et al. 1999, Larsen & Guillemette 2007). Following these
developments, the Danish government embarked upon
an ambitious plan to develop ve major oshore wind-
farms, which ultimately resulted in the construction of
the rst two major oshore windfarms at Horns Rev, west
of Blåvands Huk in west Jutland (80 × 2 MW turbines
completed in 2002) and at Nysted, south of Lolland (72
× 2.3 MW turbines completed in 2003). Currently, there
are 13 o- or nearshore wind farms in Denmark (Vindeby
having been decommissioned in 2017 having generated
243 GWh), with a combined installed capacity of 1295
MW (Tab. 1). Many of these projects were subject to con-
siderable environmental impact assessment, including
their impacts upon birds. Because of the long-estab-
lished importance of Danish marine waters for their
breeding, staging, moulting and wintering birds (e.g. Jo-
ensen 1974, Stone et al. 1995, Laursen et al. 1997, Skov et
al. 2011), we in Denmark have had considerable experi-
ence in assessing the impact of oshore wind generation
on birds, which we summarise here with experiences
from elsewhere in Europe, where the majority of devel-
opment to date has occurred in the world. In this review,
we rst consider the ways in which birds may be aected
by the construction of oshore windfarms, causing dis-
turbance or disruption to normal patterns of behaviour
or by collision and then present the results of studies that
provide information on the actual eects.
T
hroughout this article, we dierentiate “eects”
(which are the responses birds show to the presence of
wind turbines, such as avoidance) from “impacts” (which
are the results of these responses on populations, for in-
stance, if displaced terns fail to breed because of loss of
feeding grounds or birds of prey suer increased mortal-
ity because of collision deaths). Hence, eects may then
impact upon populations, in the sense that reductions
Tab. 1. List of oshore windfarms in Denmark, detailing their positions, power rating capacity, number of turbines, commissioning
year, water depth range and distance from shore.
Liste med havvindmølleparker i Danmark, deres positioner, kapaciteter, antal turbiner, idriftsættelsesår, vanddybde og afstand fra land.
Name
Sted
Location
Længde/
bredde
Capacity
(MW)
Kapacitet
(MW)
Number
of turbines
Antal
møller
Completion
year
År
færdiggjort
Water depth
range (m)
Vanddybde
(m)
Distance from
shore (km)
Afstand fra kyst
(km)
Anholt 56°36’N 11°13’E 400 111 2013 15-19 23
Avedøre Holme 55°36’N 12°28’E 11 32009 0-2 0.1
Frederikshavn 57°27’N 10°34’E 8 4 2003 1-3 0.3
Horns Rev I 55°32’N 7°54’E 160 80 2002 10-20 18
Horns Rev II 55°36’N 7°35’E 209 91 2009 9-17 32
Middelgrunden 55°41’N 12°40’E 40 20 2000 3-6 4.7
Nissum Bredning Vind 56°40’N 8°15’E 28 42018 1-6 1.1
Nysted (Rødsand I) 54°33’N 11°43’E 166 72 2003 6-9 11
Rødsand II 54°34’N 11°33’E 207 90 2010 6-9 9
Rønland 1 56°40’N 8°13’E 17 82003 0-2 0.1
Samsø 55°43’N 10°35’E 23 10 2003 10-13 4
Sprogø 55°20’N 10°58’E 21 72009 6-16 10
Tunø Knob 55°58’N 10°21’E 510 1995 3-7 6
Vindeby* 54°58’N 11°08’E511 1991 2-4 1.8
*) Vindeby was decommissioned in 2017 Vindeby blev nedtaget i 2017

88 Oshore wind farms and birds
in breeding success or additive collision mortality could
potentially aect avian population size and trends over
time, so we also need to consider the dierential impacts
of tness consequences on dierent species. Clearly
long-lived birds like divers Gavia spp. with relatively low
reproductive potential are more susceptible to small in-
creases in annual mortality, than passerines, which die in
very large numbers every year, but have the reproduc-
tive capacity to replace lost individuals more rapidly.
Finally, we reect on future needs and in particular
the lack of knowledge about the cumulative eects
of oshore development. This is because we need to
be able to assess the impacts of not just one oshore
windfarm development, but the additive eects on
specic populations of birds of concern of many such
developments in addition to all the other threats and
pressures on such populations at other points in their
annual life cycle.
Eects and impacts of oshore windfarms on
birds
At the very beginning of the work on assessing the im-
pacts of the rst large Danish oshore windfarms, it was
generally agreed that the conceptual eects/impacts
on birds largely fell into three major categories, each of
which were considered to have diering tness conse-
quences (that is impacts on breeding potential and/or
survival rates) for the populations involved. It was con-
sidered that birds encountering an oshore windfarm,
whether for the rst time or not, would be exposed to
three major hazards (after Fox et al. 2006a), namely a vi-
sual stimulus, physical change to their environment and
a risk of collision, as follows:
1. A visual stimulus that may or may not result in an avo-
idance response
Birds may react to encountering a set of wind turbines
Fig. 1. Kernels of space use by
autumn migrating Common Eiders
ying around Gedser and onwards
across the area of the Nysted
Oshore Windfarm o southern
Denmark. The space kernels rep-
resent the intensity of radar tracks
of migrating individuals across the
study area (a) pre-construction, (b)
post-construction and (c) the dier-
ence in space use between (a) and
(b). Darker shading represents the
greatest use, white in (c) indicates
reductions between (a) and (b).
The black dots denote the ultimate
positions of the individual turbines.
Reproduced with permission from
Masden et al. (2009).
Kernel-beregning af passagen af efter-
årstrækket af Ederfugle fra Gedser og
vestpå til Nysted Havvindmøllepark.
Kernel-værdierne repræsenterer inten-
siteten af radar-trækspor af okke af
Ederfugle (a) før opførelsen af Nysted
Havvindmøllepark, (b) efter opførelse
af vindmølleparken og (c) forskellen
imellem intensiteterne før og efter
opførelse af parken. Lyse/hvide områ-
der i (c) repræsenterer områder med
reduceret intensitet mellem (a) og (b).
Sorte prikker angiver positioner for de
enkelte turbiner.

89
Oshore wind farms and birds
by showing an avoidance response as soon as they are
aware of the object, or at distances closer and closer to
the turbines, depending on weather, visibility, species
and other conditions. A very distant avoidance response
by a bird ying towards the windfarm avoids collision
risk totally, because that individual is unwilling to come
anywhere close to a turbine or to risk collision. However,
as a consequence, this response may result in a barrier
to that individual’s movement. For instance, many day-
ying spring-migrating birds of prey approaching the
Anholt Oshore Wind Farm were seen to turn back in
the face of the turbines and return to the safety of the
shore (Jensen et al. 2016). Day-ying waterbirds (mostly
autumn migrating Common Eiders Somateria mollis-
sima, hereafter Eider) also showed evidence of modify-
ing their ight direction at distances up to 3 km away
from the newly constructed Nysted Oshore Windfarm,
although most modication occurred within 1 km (and
less at night; see Kahlert et al. 2004). In the case of mi-
grating birds, this may mean ying horizontally around
the turbines and at night ying up over them, both of
which incurred extra ight costs (e.g. Desholm & Kahlert
2005). This was very clearly the case for migrating Eiders
at Nysted when comparing the maps of the densities
of ight trajectories of these birds before and after the
construction of this windfarm (Fig. 1).
Looking more closely at the radar traces of the routes
taken by migrating waterbirds post-construction shows
individuals or ocks ying along and around the periph-
ery of the windfarm. The very few birds ying between
the turbines did so equidistant between turbine rows
(and always low over the sea and usually took the short-
est possible routes out of the windfarm; Fig. 2).
These responses clearly minimised the risk of collision
posed to otherwise very large numbers of birds passing
through this potentially dangerous area. Furthermore, if
this avoidance occurs only twice each year travelling be-
tween breeding and wintering areas, the extra energetic
costs that result from this detour is biologically trivial (as
was the case for Eider migrating past Nysted, adding just
500 m to a 1400 km ight; Masden et al. 2009). How-
ever, incurred energetic costs may become signicant
if the frequency of avoidance behaviour increases. For
instance, in the case of breeding birds commuting be-
tween oshore feeding areas and a breeding colony to
provision young many times each day, the energetic
costs of avoiding a windfarm suddenly constructed in
their path would be considerably greater and could po-
tentially aect their survival and reproductive success. In
this case, the degree of cost would be determined by the
length and frequency of such ights, as well as the body
mass and ight characteristics of the species concerned,
being highest for seabirds with high wing loadings such
as Great Cormorants Phalacrocorax carbo (hereafter Cor-
morant) and species such as terns that commute most
frequently between oshore feeding grounds and their
nesting colonies (Masden et al. 2010a).
In the case of non-breeding sea ducks at the mer-
cy of the wind and current, these birds may need to
reposition themselves over optimal feeding areas af-
ter being displaced during rest, so they too may show
daily ights between feeding areas and roosting sites
which could be aected by inappropriate positioning
of turbines, although this seems not to be the case at
one studied site (Piper et al. 2008). Nevertheless, Mas-
den et al. (2010a) considered that for all species, costs
of extra ight to avoid a windfarm appeared much less
than those imposed by low food abundance or adverse
weather, but with the growth in size and number of
oshore windfarms, this was likely to become more of
an issue in the future. In addition, there are strong indi-
cations from modelling the behaviours of birds studied
close to turbines, that the geometric arrangements of
turbines in clusters could have considerable eects on
how best to minimise barriers to movement and reduce
collision risks, for instance by creating corridors within
oshore windfarms to allow passage of birds through
large extensive concentrations of such structures (Mas-
den et al. 2012).
Fig. 2. The south-westerly and westerly orientated ight
trajectories of waterbird ocks and individuals based on radar
traces within the Nysted Oshore Windfarm during the initial
post-construction operation of wind turbines at the site. Red
dots indicate the positions of individual turbines, the scale
bar represents 1000 m. Reproduced with permission from
Desholm & Kahlert (2005).
Vandfuglenes sydvestlige og vestlige trækruter baseret på
radarspor indenfor og omkring Nysted Havvindmøllepark efter
opførelse af turbinerne. Røde prikker angiver positionen af de
enkelte turbiner. Målestoksforhold er angivet ved den sorte bjælke
under nord-pilen, som repræsenterer 1000 m.

90 Oshore wind farms and birds
It has been suggested that oshore wind farms
act as an attractant to migrating birds of prey, as was
suggested to be the case for the Anholt Oshore Wind
Farm, the site of 111 wind turbines situated in Kattegat,
halfway between northeast Djursland and the island of
Anholt. Modelling suggested that the passage of spe-
cies like Common Buzzard Buteo buteo, Sparrowhawk
Accipiter nisus and Honey Buzzard Pernis apivorus poten-
tially would encounter relatively high collision fatalities
post construction, but observations showed high level
of macro-level avoidance of the wind farm that rather
suggested the alternative problem of a barrier eect.
Although this may prolong migration routes and aect
survival and/or reproduction, this behaviour did at least
reduce the probability of collision mortality among such
relatively long-lived bird species (Jensen et al. 2016).
Another consequence of avoidance responses to the
visual stimulus of novel rotating turbine blades and tow-
ers was to displace birds from ideal feeding distributions.
If, for any reason, birds are unwilling to approach an ac-
tively operational turbine to within a distance that is
half of the distance between adjacent turbines, the area
within the area of a windfarm becomes behaviourally
inaccessible to them, even if the physical feeding habi-
tat and prey are still present, theoretically available and
even potentially improved as a result of the construction
of the wind turbines. In other words, functional habitat
loss results from the behaviour of the birds, because the
food supply and habitat remain intact, but the presence
of the turbines inhibit the birds from approaching and
using such areas. This seems to be the case for certain
species, such as the Common Scoter Melanitta nigra and
Red-throated Diver Gavia stellata, species which seem
to avoid foraging in waters between the turbines in
windfarms in Denmark (Fig. 3; Petersen et al. 2006, 2014),
although small numbers of Common Scoters (at much
lower densities) have been recorded between turbines
on occasions in both Horns Rev I and II windfarms (e.g.
Petersen & Fox 2007). It seems likely that variations in
food supply could explain this variable response, but it
is generally the case that these two species are extreme-
ly reticent to forage between the turbine rows in those
windfarms where they were formerly common. At Horns
Rev, Red-throated Divers which had been present in the
pre-construction windfarm footprint area were not seen
within the newly built windfarm during the post con-
struction monitoring, although a very few individuals
have been seen between turbines in subsequent years.
In the Netherlands, Red-throated Divers were not de-
tected between turbines at one site but did so at anoth-
er Dutch windfarm (Lindeboom et al. 2011) conrming
responses can be variable, even within species. Compar-
ison of the before- and after distributions of Red-throat-
ed Divers in the German Bight suggest a major displace-
ment eect from newly constructed windfarms out to
at least 16 km and reductions in bird densities of more
than 60% in an area within 10 km of the turbines (Men-
del et al. 2019). More perplexing is the case of the Long-
tailed Duck Clangula hyemalis, which foraged in the area
of the subsequent Nysted Oshore Windfarm before its
construction, but has done so since the site became op-
erational in densities much less than those prior to con-
struction and compared to areas where it occurs imme-
diately adjacent to the windfarm (Fig. 4; Petersen et al.
2011). This seems to imply that while some individuals
are willing to forage between the turbines, others that
formerly did so are not now willing to do so, for what-
ever reason (Petersen et al. 2011). Razorbill Alca torda
and Common Guillemot Uria aalge also tend to occur
in lower numbers post-construction of oshore wind-
farms (e.g. Dierschke & Garthe 2006). However, without
understanding the relative importance of a given feed-
ing area and the likelihood, possibility and energetic
costs of shifting elsewhere to feed when displaced by
wind turbines, it is dicult to determine the true costs
(in terms of energy balance, for example) to the individ-
ual of being denied a feeding area in this way or the con-
sequences for the population. It is even more dicult to
assess the impacts on populations from multiple such
developments along waterbird migration corridors (see
the later discussion on cumulative eects). For these
reasons, it can be extremely dicult to conclude about
the true level and impact of displacement of birds from
formerly good feeding areas, as it seems to be a complex
response likely mediated by site and species concerned,
but also potentially on individual responses, levels of ha-
bituation and the richness of the food supply in ways we
have yet to understand.
2. Physical habitat loss/modication or gain
In the case of the well-studied Nysted and Horns Rev
oshore windfarms, the extent of physical loss to turbi-
ne foundations and of habitat modication to anti-scour
protection never amounted to more than 2% of the total
area of the windfarm (Fox et al. 2006a). For most recent
oshore developments, these assessments would be
similar, because anti-scour, foundation and cable dis-
ruptions to the general seabed inside and in the vicinity
of wind turbines tend to be limited to a similar propor-
tional area. Foundations and anti-scour provisions may
also attract the settlement of ora and fauna which are
“alien” to that specic local seabed habitat, as in the case
of providing a solid hard substrate and crevices within
areas of sandy-bottomed habitat which have become
favoured sites for rammed foundation turbines. As a re-
sult, sh and invertebrates that seek shelter in crevices

91
Oshore wind farms and birds
may occur around turbine anti-scour foundations and
blue mussels Mytilus edule may settle on foundations in
densities not formerly possible on sea-beds which for-
merly consisted of bare open sand substrates. Although
not totally insignicant, the areas over which such ef-
fects are manifest constitute a biologically trivial propor-
tion of the total area. Some foraging Eiders (presumably
taking blue mussels) have been seen associating with
turbine bases, but there is no evidence for any signi-
cant eects on bird distributions. Species such as the lar-
ger Larus gull species and Cormorants are undoubtedly
attracted to the superstructure of turbines, meteorolo-
Fig. 3. Map of the Horns Revs 2
Oshore Windfarm Study Area. The
map shows estimated dierences
between pre- and post-construction
densities of Red-throated Diver (A)
and Common Scoter (B) estimated
by generalised additive models in
each 500 × 500 m grid square av-
eraged over the survey period. The
legend denes the colour codes for
the level of changes in each of the
maps. Black plus-symbols indicate
grid squares, which showed statisti-
cally signicant increases, and open
white circles those with signicantly
reduced numbers. Open polygons
indicate the Horns Rev I (construct-
ed prior to this investigation, closest
to land) and Horns Rev 2 oshore
wind farms.
Kort over forandring af tætheder af ra-
stende Rødstrubet Lom (A) og Sortand
(B) ved sammenligning af gennem-
snitlige tætheder for perioden før op-
førelsen af Horns Rev 2-havvindmøl-
leparken og perioden efter opførelsen
af denne. De modellerede værdier blev
beregnet for et kvadratnet med 500
× 500 m celler. Farveskalaen angiver,
hvorvidt tæthederne er forøgede eller
reducerede, røde og gule farver an-
giver forøgelse, grønne og blå farver
angiver reduktion. Sorte krydser angi-
ver celler, hvor en tæthedsforøgelse er
statistisk signikant. Åbne hvide cirkler
angiver celler, hvor en tæthedsredukti-
on er signikant. Åbne sorte polygoner
angiver hhv. Horns Rev 2 og Horns Rev
1 havvindmølleparkerne. Horns Rev 1
blev opført før starten på indsamlin-
gen af data til disse sammenligninger
af før- og efterfordelinger af fugle ved
Horns Rev 2. De neutrale værdier for
Horns Rev 1-området indikerer, at fx
Sortænder var yttet bort fra området
før indledningen af denne undersø-
gelse, og at de ikke inden for perioden
har vist tegn på tilvænning til parkens
tilstedeværelse.

92 Oshore wind farms and birds
gical masts and transformer stations, associated with
oshore windfarms, but Danish studies found no incre-
ased densities of these species post construction (Peter-
sen et al. 2006), in contrast to ndings in the Netherlands
(Lindeboom et al. 2011).
3. Collision mortality
Of all the potential eects of oshore windfarm con-
struction, deaths from collisions have always attracted
most attention as the primary impact on bird popula-
tions. Birds may die by hitting stationary superstructu-
res, the stationary or rotating rotor blades or by being
caught and mortally injured in the vortices created in
the wake of the rotor blades (Fox et al. 2006a). Many
birds (but especially night migrating passerines) collide
with stationary objects on land and at sea (e.g. Kerlinger
2000), especially when these are illuminated, so much
eort has been put into tting navigation lights to os-
hore wind turbines that avoid the need for illumination
and at the same time do not attract birds (Drewitt &
Langston 2008). Studies suggest that birds show avo-
idance of turbines at three spatial scales, the macro-sca-
le (within 3 km of the turbine), the meso-scale (within
the windfarm footprint, i.e. between turbines) and the
micro-scale where birds respond to the proximity of
the blades and the monopole (within 10 m; Skov et al.
2018), so all these need to be carefully considered in any
assessment of potential collision risk at oshore wind-
farms. However, as we report later on, because of the
high levels of avoidance shown by many larger-bodied
seabirds to oshore wind installations, the experience
has generally been that collision rates are low.
Dealing with potential eects/impacts on birds
Species specic impacts
As is the case for any major development aecting the
natural environment, it is important to understand that
not all birds are aected equally by the construction of
oshore turbines, either in terms of the immediate risk
of collision and other eects on their behaviour and
ecology, or the eects on their reproductive success and
survival and ultimately population dynamics (i.e. wheth-
Fig. 4. Map of the Nysted Oshore
Windfarm study area showing
estimated dierences in Long-tailed
Duck numbers within grid cells of
500 × 500 m distributed across the
entire study site generated from
a spatially-adaptive generalized
additive model pre- and post-con-
struction of the windfarm. Estimated
abundances were derived from
combined aerial survey data that
counted birds along transects and
adjusting abundance for detection
probability. Negative dierences
(shades of blue) indicate fewer
individuals in cells post-construc-
tion than prior, positive dierences
(yellow-orange-red grid squares)
indicate increased numbers
post-construction. Black cross sym-
bols indicate statistically signicant
increases and open white circles
indicate statistically signicant
decreases in these numbers when
comparing pre- and post-construc-
tion abundance in these grid cells based on model estimates. Contour lines indicate depth intervals as labelled in metres. The
ultimate position of the windfarm is identied by the light grey polygon outline and aerial transects waypoints are indicated by
the T symbols.
Kort over forskellen mellem tætheder af Havlit før og efter opførelse af Nysted Havvindmøllepark. Forskellen i tætheder er beregnet for
hele undersøgelsesområdet og som gennemsnitlige værdier hhv. før og efter opførelse af parken. Der blev beregnet tæthedsværdier i et
kvadratnet med celler på 500 × 500 m beregnet med rumlig modellering, og baseret på optællinger af fugle fra y langs forudbestemte
transekter, hvorved en kompensation for detektions-sandsynlighed kunne indregnes. Blå farver indikerer reducerede tætheder ved sam-
menligning af tætheder før og efter parkens opførelse, mens røde og gule farver indikerer forøgede tætheder ved samme sammenlig-
ning. Sorte krydser indikerer celler med en statistisk signikant forøgelse af tætheder efter etablering af mølleparken, mens åbne sorte
cirkler indikerer celler med statistisk signikant reduktion i tætheder efter etablering af parken. Den lysegrå rkant midt i guren indikerer
placeringen af Nysted Havvindmøllepark. Sorte T-symboler viser endepunkter for transektlinjer anvendt under optællingerne.

93
Oshore wind farms and birds
er population size changes as a result of wind turbine
impacts). Clearly bird behaviour will aect the chance
of collision mortality, because species that habitually
y at rotor sweep heights will be far more susceptible
than those that y low over the sea. Feeding ecology,
ight height and visual acuity (see Martin 2011) aect
the threats posed to birds by turbines, hence skuas Ster-
corarius spp., Northern Gannets Morus bassanus, gulls
and terns, which y relatively high over the water sur-
face and may be visually distracted by concentrating on
kleptoparasitic pursuit or subsurface prey may be more
susceptible to collisions than, for example, divers Gavia
spp. and auks, or diving ducks such as Common Scoter
and Long-tailed Duck, which tend to y low over the
water surface and feed in the water column or on the
benthos.
Avian species that behaviourally show strong re-
sponses to man-made objects are more likely to avoid
novel structures in the marine environment compared
to species such as some gulls and Cormorants, which
already exploit (and indeed may be attracted to) human
marine architecture throughout their range. Body size
and aerodynamics will also aect the ability of birds to
make last minute avoidance to turbine blades, so small,
highly manoeuvrable birds may have less likelihood of
collision than larger birds that present a large surface
area and show slower avoidance responses (Drewitt
& Langston 2008). Curiously, even the absolute death
rate caused to a species may have dierential eects on
overall population size. Long-lived marine species with
low reproductive turnover, such as divers, are far more
susceptible to even very small increases in adult mor-
tality compared to small passerines, which are short-
lived but which can produce large numbers of young
to replace losses, especially in situations where strong
density dependent eects may aect demographic
rates, so lower breeding densities may enable elevated
reproductive success (Desholm 2009). Hence, it is vital to
consider which bird species are likely to be aected and
in what way by the construction of a specic windfarm.
Site and project specic impacts
The eects of oshore windfarm construction are also
highly dependent upon the characteristics of the site
and the nature of the construction work that is proposed.
Clearly windfarms should not be constructed in areas
where migrant birds of any type are concentrated by
coastal topography (e.g. at the tips of peninsulas where
migrating land birds are classically known to gather;
Desholm et al. 2014), because birds departing on migra-
tion from such “pinch” points will inevitably be highly
concentrated as they funnel out to disperse onwards on
migration. Given these topographical concentrations
of migratory avian trac in specic airspace, avoiding
construction of turbines in these areas will avoid any risk
of collision mortality in particular areas of likely conict.
Likewise, narrow sea passages between landmasses or
promontories rounded by large numbers of migrating
waterbirds may also create concentrations, making such
sites highly unsuitable for the siting of turbines. Feeding
marine bird species are not randomly distributed at sea,
so regular aggregations of seabirds attracted to known
food resources should also be avoided as potential ar-
eas for oshore windfarms. Unfortunately, assessments
of the feeding resources of piscivorous birds may not
be constant, nor simple to predict, although divers (e.g.
Skov & Prins 2001) and Little Gulls Hydrocoloeus minutus
(Schwemmer & Garthe 2006) clearly associate with oce-
anic surface front systems, which despite being ephem-
eral marine features, can show seasonal predictability
in time and space. Even benthos feeding birds, such as
scoters and Eiders, may shift between dierent feeding
areas between years because spat-settlement of their
essentially bivalve prey may result in major dierences
in prey availability between years, due to age and size
class distributions aecting the annual protability of
their food supply. Nevertheless, in Denmark, there is
a presumption to avoid development in very shallow
waters (< 10 m) to avoid major conicts with potential
feeding areas for seabirds feeding on benthos and on
aggregations of organisms in the water column that are
typically most common in such shallow waters. All other
things being equal, the size, layout, distance between
individual turbines, location and siting of turbines will
also aect the likely impact of windfarm construction
of birds using the general area and these also need to
be taken into consideration when attempting to predict
the specic avian impacts of a given development (e.g.
Masden et al. 2012).
Environmentally determined impacts
The interactions between weather and local topogra-
phy also create unique conditions that potentially im-
pact dierentially upon bird species. Mist, rain and snow
showers, especially in situations of rapid meteorological
change, can all result in disorientated migrant birds col-
liding with illuminated structures, potentially causing
catastrophic (if highly infrequent) mortality events that
can aect one or many species (see Newton 2007) and
these considerations are also by nature, site-specic.
Temporal patterns of impacts
Finally, the extent to which impacts may be manifest
for bird populations vary greatly with season. The Eider
neither breeds nor winters in any substantial numbers
in the vicinity of the rst Nysted oshore windfarm.

94 Oshore wind farms and birds
However, the entire breeding population of the north-
ern Baltic (200 000-300 000 birds) passes through this
very restricted area every spring and autumn on annu-
al migration en route to and from the winter quarters
(Desholm & Kahlert 2005). Self-evidently therefore, any
impact assessment of wind turbines constructed at sea
needs to take account of avian movements throughout
the entire annual cycle. Many waterbird species (espe-
cially scoters and Eiders in Danish waters) undergo a
simultaneous wing moult that renders them ightless
for some three or so weeks while remigial feathers are
replaced. At this stage in the annual life cycle, the birds
are highly sensitive to disturbance and show a much
stronger avoidance response to human structures and
activities at sea than at other times of the year (Petersen
& Fox 2009, Petersen et al. 2017). Since the moult period
is a particular energetic bottleneck for these birds and
because of their heightened sensitivity to human distur-
bance at this time of year, particular attention should be
given to siting windfarms in relation to concentrations
of these birds, which often draw birds from along large
expanses of their yway.
For all of the above reasons, it is therefore very ev-
ident that any environmental impact assessment of a
new oshore windfarm needs to take into consideration
the specic challenges of the project and site, especially
with regard to the species presence, their abundance,
sensitivity and conservation status. Such assessments
also need to cover the entire annual cycle to take ac-
count of seasonal changes and should be based on
more than just one year (and ideally more than two) to
assess the degree of within and between year variations
in the patterns observed. They also need to consider the
nature of the proposals, with regard to construction, op-
eration and decommissioning activities, turbine height,
sweep area, numbers and the associated infrastructure
and their impact on the environment (such as trans-
former stations, buried underground cables, lighting,
disturbance from maintenance trac, etc.). Hence, it is
impossible to conclude on a general level about the scale
and magnitude of eects and impacts of oshore wind
farms are likely to have upon the bird populations which
encounter them based on our experience of those con-
structed so far. It is also the case that lamentably few o-
shore wind farms have been adequately monitored for
prolonged periods post construction (rarely more than
two years) to provide a suciently rich record of the
true consequences (rather than the more speculative
pre-construction environmental impact assessments)
to inform future development. Although monitoring is
inevitably costly, the value of such long-term monitor-
ing of eects and impacts cannot be overvalued. Never-
theless, our experience to date enables us to say a great
deal about the general eects (proximate changes in
bird behaviour, local distribution and abundance) and
impacts (dened as ultimate changes in population size
because of reduced reproduction or survival) of the con-
struction and operation of the existing Danish oshore
windfarms, especially with additional experiences from
other countries.
Sequential assessment of eects
Construction and decommissioning phases
There has been hardly any study of the eects on birds
during the period when oshore turbines are being
constructed, but there is no doubt that enhanced ship
and maintenance trac, noise, lighting and concentrat-
ed activity in the development footprint of the wind-
farm are likely to be highly disruptive, and of a dierent
nature, compared to the prior undisturbed situation, as
well the subsequent operational phase. During this pe-
riod, changes to shipping lanes and trac and modica-
tion of shing activity in the vicinity will also come into
eect, while extreme disturbance (e.g. pile driving) can
have profound potential eects on birds, as well as their
prey. On the other hand, the limitations of day length
and availability of good weather tends to constrain con-
struction to a short period of duration in the summer
when there tend to be fewer birds present, with the re-
sult that any potential eects are of very short duration
and of minimal impacts. Unfortunately, there have been
no assessments of the eects of windfarm decommis-
sioning, but these are likely to be of short duration and
of similar nature to the construction phase.
Operational phase
The eects of oshore windfarms on birds during the
initial operation stage have been much studied in rela-
tion to points 1, 2 and 3 above. In the case of studying
the eects of the appearance of turbines in areas of
open sea formerly devoid of such structures, the main
approach to understand avoidance by ying birds has
been to examine the directions of ight before and after
construction using marine surveillance radar, mounted
both vertically and horizontally to generate the intensi-
ty of bird movements in three dimensional space (e.g.
Desholm & Kahlert 2005, Desholm et al. 2006, Petersen
et al. 2006, Krijgsveld et al. 2011, Plonczkier & Simms
2012, Leopold et al. 2013, Skov et al. 2018). These results
generally show major macro-scale adjustments. For in-
stance, migrating Eiders rounding the southern tip of
the Gedser peninsula approaching the Nysted Oshore
Wind farm showed adjustments to ight trajectories to
avoid the turbines at distances up to 3 km away (Kahlert
et al. 2004). Some species were almost never seen ying
between the turbines (Red-throated Divers and North-

95
Oshore wind farms and birds
ern Gannets), others rarely (Common Scoter), while yet
others showed little avoidance (e.g. Cormorants and
large gulls). At Horns Rev, 71-86% of all large bird ocks
heading towards the windfarm at 1.5-2 km distance
avoided entering the wind farm and ying between the
turbine rows (Petersen et al. 2006). The same pattern
was conrmed at Nysted (78%) predominantly amongst
waterbirds, mostly migrating Eiders, but including a
wide range of species (Petersen et al. 2006). The rela-
tively few birds entering the Nysted Oshore Windfarm
also ew midway between turbines rows at low altitude
(below rotor sweep height) and exited the wind farm by
the shortest routes more quickly than could be expect-
ed by chance (Desholm & Kahlert 2005). This resulted
in considerable movement of birds up and down the
periphery of both windfarms as birds preferentially ew
around rather than between the turbines (Fig. 2). Such
avoidance rates were also conrmed at night by radar,
when it was also shown that although the response dis-
tance occurred much closer to the turbines, birds also
tended to y much higher. However, in a few regrettable
cases, impact studies failed to establish the predicted
impact of wind turbines, as in the case of 25 medium
turbines established on eastern port breakwater at Zee-
brugge, Belgium. These turbines were constructed on
a breakwater encircling a breeding colony of Common
Sterna hirundo, Sandwich Thalasseus sandvicensis and
Little Terns Sternula albifrons, and post construction
studies revealed that a mean of 6.7 terns collided per
turbine per year for the whole wind farm (with highest
rates at turbines closer to the breeding colony within 10
m of the nearest turbine). Many gulls were also recov-
ered dead under the turbines conrming the need to
avoid constructing wind turbines close to any import-
ant tern, gull or other sea bird colonies, especially those
associated with frequent foraging ight paths of these
species, because of the high risk of associated collision
mortality (Everaert & Stienen 2006).
Unfortunately, few observational data relating to any
of the study species were obtained during periods of
poor visibility, but generally this was because bird mi-
gration slows and ceases under such circumstances, as
conrmed by radar studies (Petersen et al. 2006). These
studies also conrmed that for large bodied species such
as the larger seabirds, sea ducks (such as Common Scoter
and Eider) and geese (all species particularly susceptible
to additional mortality) as well as migrating dabbling
ducks, there were good grounds to suspect avoidance
behaviour at macro- (< 3 km distance) and meso-scales
(e.g. avoiding ying anywhere near turbines and midway
between rows within wind farms) substantially reduced
the probability of any collisions with turbines.
There is growing evidence for widespread avoidance
of oshore turbines by large-bodied birds, while our
knowledge for smaller bird species is less adequate.
Photo: Ørsted.
Især større fugle har vist sig gode til at undgå kollisioner
med havvindmøller, mens vi har mindre viden om små-
fuglenes kollisionsrisiko.

96 Oshore wind farms and birds
To predictively attempt to estimate the collision rates
of birds, based on the level of avian ying activity re-
corded by radar and other methods in advance of con-
struction, several modelling approaches have been de-
veloped to try and predict the annual numbers of birds
which will collide with turbines ahead of construction
(see Chamberlain et al. 2006 and Masden & Cook 2016
for reviews). All of these models rely ultimately on de-
termining the probability of last minute (i.e. micro-level)
avoidance that birds are able to make when close to the
blades. This parameter is highly dependent upon spe-
cies, weather conditions, visibility etc., and is notorious-
ly dicult to estimate or quantify. Nevertheless, one of
these stochastic models was used to predict that out of
235 000 Eiders passing the Nysted Oshore Windfarm,
0.018-0.020% of these would collide with turbine blades
in an autumn (Fox et al. 2006b). With such a low prob-
ability, it was predicted that the infra-red (i.e. thermal)
video monitoring system set up to detect such collisions
would fail to detect a single collision during 2400 hours
of monitoring, which proved to be the case. The system
detected only 11 birds, all well away from turbine sweep
area, two passing bats, two passing objects (either bats
or birds), a moth and one collision of a small bird with a
turbine blade (Petersen et al. 2006).
Since that time, much eort has been invested in
creating improved models to estimate collision rates
given bird ight trajectories generated from two- and
three-dimensional radar tracking (e.g. Skov et al. 2018).
This has also resulted in much eort measuring ight
heights probabilities to parameterise such collision risk
models (e.g. Johnson et al. 2014, Cleasby et al. 2015,
Fijn et al. 2015). There have also been advances in the
techniques available to enable the eld validation of
collision rates and avoidance of turbines among birds
(and bats) at oshore turbines (Dirksen 2017). Recent
results from monitoring detailed movements of a range
of species previously thought to be at risk (large gull
species and Northern Gannet) show meso- and micro
avoidance behaviours that substantially reduce the risk
of collision and contribute to very low observed colli-
sion rates (Skov et al. 2018).
It is also fair to say that we remain sadly ignorant
of the actual rates of collision of smaller birds with o-
shore turbines. Generally, attention has been focussed
upon the larger bodied species because of their relative
Long-tailed Ducks used to forage in the area of the subsequent Nysted Oshore Windfarm but has done so much less since the
site became operational than prior to construction. Photo: Hans-Henrik Wienberg.
Havlitten er blandt de arter, der er blevet fortrængt fra tidligere fourageringsområder ved opførelse af havvindmølleparker.

97
Oshore wind farms and birds
vulnerability to elevated death rates (primarily from
collision) and because larger bodied birds are easier
to monitor using techniques such as infra-red vide-
ometry and radar. This is not to say that there is (or is
not) a major problem with smaller species, merely that
to date, they have not been subject to robust levels of
monitoring. Generally, it is considered that there is no
major problem with other species, and infra-red vide-
ometry at Nysted conrmed this to be the case at that
site. However, there remains the minimal risk that under
certain (likely very rare) prevailing weather conditions,
circumstances may conspire to cause major collision
mortality and we would urge more low-key long-term
monitoring to better determine the levels of such risks.
If there are conditions under which unacceptable levels
of collision deaths occur for any species, we should be
thinking in terms of developing forms of mitigation, for
instance implementing early warning devices to warn of
the approaching risk and potentially using remote sens-
ing to detect bird movements close to rotor blades to
cease electricity generation under such circumstances
(Dirksen 2017).
Finally, it is important to remember that wind tur-
bines require regular maintenance and irregular repair,
necessitating support vessels, cranes, helicopters and
operating crews being active in waters which were of-
ten not subject to frequent ship trac pre-construction.
Although designation of windfarms as “no-go” shing
areas may reduce the physical presence of boats in an
area of constructed turbines, intense maintenance traf-
c in formerly undisturbed areas and along routes to
and from their home harbours may add substantially
to the sources of surface anthropogenic disturbance to
seabirds out in the open sea. This is most likely to have
eects on the distributions of birds foraging in the area
but will also aect other species.
The greatest future challenges
It is very evident from where we are now that we need
to take a more strategic national and international ap-
proach to the identication, assessment and selection
process for the selection of areas suitable for future o-
shore windfarm developments. However, our greatest
challenge for assessing the impacts of oshore wind-
farms on birds is an assessment of their so-called ”cu-
mulative impacts”. As clearly recognised here, individual
windfarms may have minor eects on the environment,
but collectively, many of these developments, especial-
ly spread out to confront individuals from a migratory
avian population along the entire length of its migra-
tion corridor may have a signicant eect. This eect
may be far greater than the sum of the individual parts
acting alone, especially if contributing adversely to the
tness of many individuals. EU Directives 85/337/EEC
(as amended) and 2001/42/EC both require that a cu-
mulative impact assessment is undertaken as part of
an environmental impact assessment of an individual
proposed oshore windfarm development. However, to
date, we still lack detailed guidance about how to tackle
such cumulative assessments and those that have been
attempted have generally been inadequate and not
subject to retrospective review. We therefore remain
remarkably ignorant about the cumulative impacts of
many oshore windfarms on bird populations, although
happily there continue to be new attempts to create a
conceptual framework for such analysis (e.g. Masden et
al. 2010b, Poot et al. 2011, Busch et al. 2013, May et al.
2018). In our humble opinion, this remains one of the
single most important areas to address in the future,
as we see more and more development of oshore po-
tential for electricity generation. As our seas become
increasing enclosed and covered with turbines, there
clearly will be major cumulative eects on bird popula-
tions of which we remain ignorant at the present time.
Conclusions
The hazards presented to birds by the construction of
oshore windfarms remain primarily (i) the barrier they
present to movement, (ii) loss, gain and enhancement of
habitat and (iii) collision risk. Most studies to date have
used radar and thermal infra-red monitoring as well as
range-nding and visual observations to conrm that
most of the more abundant and especially large bodied
birds show major avoidance to oshore windfarms, mi-
nimising the probabilities of collision. Slightly extended
migration distances are unlikely to have consequences
for these species. Eects on breeding and wintering
birds interrupted during their commuting ights re-
main less well studied, but avoidance of conict is eas-
ily achieved by siting oshore wind turbines well away
from important concentrations of breeding and winter-
ing seabirds and their respective feeding areas.
Avoidance also extends to some species of birds
which aect their feeding distributions (usually outside
of the breeding period). Such physical displacement as
a result of individuals avoiding to feed in the vicinity of
turbines means that the species suers eective hab-
itat loss, even though the habitat and even the food
supply may remain intact. From studied locations, this
seems to be the case for Red-throated Divers, Common
Scoters, Long-tailed Ducks, Razorbills and Common
Guillemots, but for most species, we lack sucient data
of sucient quality to make a judgement. While it has
been possible to demonstrate such eects, it remains
a major challenge for the future to understand how in-
creasing displacement from ideal foraging habitat may

98 Oshore wind farms and birds
impact upon population processes, especially as a result
of cumulative eects along the yways of the migratory
waterbirds concerned.
Avian avoidance at long distances reduces the col-
lision risk to individuals and this seems to be the case
for many study species. Although this has mostly been
studied for the large-bodied bird species considered
most at risk, we suspect this to be the case for small-
er bird species as well. Recent results from monitoring
detailed movements of a range of species previously
thought to be at risk (large gull species and Northern
Gannet) show meso- and micro avoidance behaviours
that also substantially reduce the risk of collision and
contribute to very low observed collision rates. Howev-
er, we lack long-term and intensive eective monitoring
of the numbers of bird collisions at oshore turbines un-
der a vast range of diering seasonal and weather con-
ditions and at dierent sites to be truly condent that
this impact is as minimal as all studies suggest they are.
Still, our experience to date has provided a very solid
foundation upon which to propose robust impact as-
sessments following specic methods to determine the
eects on bird populations from the proposed develop-
ment of new windfarms in oshore waters.
One of the greatest historical challenges in the early
days of oshore windfarm development was the rather
piecemeal nature of the development. Windfarms were
proposed in areas which were good for windfarms (in
the sense that the wind proles, suitability of substrates,
connections into the electricity grid and economic con-
siderations mitigated in favour of their construction),
but for which we lacked good biodiversity information
(including birds). This meant, for example, that biolog-
ical assessments undertaken as part of the impact as-
sessment of windfarms in Britain discovered previously
unknown concentrations of wintering Common Scoters
and Red-throated Divers that ultimately stopped or
caused major modication to the proposed construc-
tion of windfarms, at great expense to the developers.
In Denmark, we are now in a better position to combine
strategic marine planning layers that describe shipping
routes, buried submarine cables, military restriction
areas, shing banks, protected areas and other sites of
important biodiversity interest (including historical bird
distributions) and other features of stakeholder interest
to look more strategically at where best to site wind-
farms to avoid conicts with other users of the marine
environment at a preliminary stage. However, it remains
essential to undertake detailed bird surveys to deter-
mine the true current importance of areas proposed for
windfarm development and to set the derived knowl-
edge in the context of the potential eects on their y-
way populations.
EU Directive 2001/42/EC requires a strategic environ-
mental assessment (SEA) of national wind energy plans
and programmes that have potential adverse eects
on biodiversity, which would also help guide marine
wind power developments, both nationally and inter-
nationally. International coordination and collaboration
is required under the United Nations Espoo Convention
(UNECE 1991) where there are potential transboundary
eects regarding the placement of oshore windfarms.
While obligatory EIA legislation (EU Directive 85/337/
EEC and 97/11/EC) requires project level environmental
impact assessment, these tend to take account of eects
on birds only at the local geographical scale. The SEA
and EIA Directives require assessment of the cumulative
eects of each proposal (including associated on- and
oshore infrastructure development, such as road im-
provements, power lines, etc.) in conjunction with other
projects and factors (not necessarily only other oshore
windfarms, so including pollution, sheries, ship trac,
mineral extraction from the sea bed, etc.) that impact
upon the same yway populations of birds. These re-
quirements make it even more essential that we use
our current knowledge to become better able to model
the cumulative eects of many such windfarm develop-
ments in the context of the many other development
pressures that currently threaten our bird populations.
In the meantime, our planet warms and the pressure to
provide renewable non-fossil fuel electricity increases.
There is no doubt that oshore windfarms can make
a major contribution to providing such power, and we
therefore need to nd innovative solutions to ensure we
do not save the planet at cost to migratory bird popu-
lations.
Acknowledgements
We would like to dedicate this review to the memory of our
good friend and late colleague Sjoerd Dirksen who died sud-
denly recently, with whom it was always inspiring to work with
on oshore windfarm issues. We are extremely grateful to fun-
ding from very many sources that supported our oshore wind-
farm impact assessment work in Denmark, fully acknowledged
among the very many reports and publications referenced here.
We thank the sta of the very many development companies
and statutory agencies with whom we have worked on these
projects, colleagues at the former National Environmental Re-
search Institute, Denmark, and Aarhus University (especially
Thomas Kjær Christensen, Ib Clausager, Johnny Kahlert, Mark
Desholm) and around the world who have contributed to im-
proving our work in this eld and to the many reviewers of our
studies for their help and support over the years. Thanks to
Oxford University Press (order 4443100278867) and the Royal
Society of London for their written copyright clearance permis-
sions to reproduce Figs 1 and 2, respectively and to the Aarhus
University Graphic Workshop for producing the gures. Finally,
thanks to two referees for their constructive suggestions for im-
provements to an earlier draft.

99
Oshore wind farms and birds
Resumé
Havvindmøller og deres påvirkning af fugle
Udfordringerne for fugle ved etablering af havvindmølleparker
kan samles i tre hovedkategorier, nemlig 1) barriereeekt i for-
hold til fuglenes bevægelser, 2) forandring af habitatet, der kan
medføre tab, forbedring eller udvidelse af areal og 3) kollisions-
risiko. Langt de este undersøgelser, der har anvendt radar og
infrarød overvågning kombineret med laser-afstandsmålende
kikkerter og menneskelige observatører til at registrere fugles
reaktion på møllerne, har kunnet konstatere, at talrige fuglear-
ter, og specielt de større fugle, undgår havvindmølleparkerne
på ret stor afstand og reducerer på den måde risikoen for kol-
lision med turbinerne. Det er derimod mindre grundigt belyst,
hvordan ynglende og overvintrende fugle kan blive påvirket
under deres – ofte daglige – yvninger, men sådanne påvirk-
ninger kan let undgås ved at projektere nye vindmølleparker på
afstand af vigtige yngle- eller overvintringsområder, og dermed
undgå eller reducere potentielle barriereeekter.
Undvigende adfærd omfatter imidlertid ikke bare forbitræk-
kende fugle, men kan også omfatte tab af fourageringsområder
(oftest uden for yngleperioden). En sådan reaktion, forårsaget
af fuglenes uvilje til at fouragere tæt på turbiner, forårsager et
eektiv habitattab, også selv om det marine habitat og den til-
gængelige føderessource forbliver uændrede. På grundlag af
undersøgte etablerede havvindmølleparker er der stærke indi-
kationer på, at det er tilfældet for lommer, Sortand, Havlit, Alk
og Lomvie, men for hovedparten af disse arter mangler vi data
af tilstrækkelig robust karakter til at foretage en tilstrækkelig
velunderbygget vurdering. Og selv om det har været muligt at
konkludere sådanne eekter for nogle få arter, forbliver det en
stor udfordring at undersøge, hvordan stadigt stigende tab af
habitat fra foretrukne fourageringsområder kan have en eekt
på den samlede ywaybestand af en given art og artens demo-
gra – i særdeleshed når man tager de kumulative eekter af
mange vindmølleparker langs en arts trækrute i betragtning.
Fuglenes afvigereaktion på stor afstand af turbinerne redu-
cerer risikoen for kollision, og dette ser som nævnt ud til at være
tilfældet for en lang række arter. Selv om dette hovedsagelig har
været undersøgt for større fuglearter, der betragtes som mere
i risiko for kollision end små arter, forventer vi, at det samme vil
være tilfældet for mindre fuglearter. Nylige detaljerede moni-
teringsundersøgelser af passage af arter, der tidligere blev be-
tragtet som værende i risiko for kollision (større mågearter og
Sule), viste undvigelse overfor turbinerne på mellem- og kort af-
stand, hvilket samtidig reducerer risikoen for kollision markant
og gav meget lave antal observerede kollisionsrater. Vi mangler
imidlertid moniteringsprogrammer med større varighed og in-
tensitet til at beskrive antallet af kollisioner ved turbiner til havs,
og som strækker sig over forskellige årstider og vejrmæssige
forskelligheder og fra geogrask forskellige områder for at få
vished for, at denne indydelse på fuglene er så beskeden, som
de foreliggende studier antyder, at de er. Vores hidtidige erfa-
ringer har imidlertid givet os et solidt grundlag for at denere
specikke undersøgelsesmetoder til at beskrive de potentielle
eekter af etableringen af nye havvindmølleparker.
En af havvindmølleparkernes tidlige udviklingsmæssige
udfordringer var den ”bid for bid”-udvikling, som man var nødt
til at gennemgå. Vindmølleparker blev projekteret i områder,
Many bird species most often y low over the water and thereby out of risk; here Barnacle Geese. Photo: Lars Maltha Rasmussen.
Mange fuglearter yver oftest lavt over havoveraden og dermed udenfor fare fra møllevingerne, som disse Bramgæs.

100 Oshore wind farms and birds
der var gunstige for havvindmølleparker, dvs. steder, hvor vind-
prolen var optimal, hvor havbundens sediment var velegnet til
fundering af turbinerne, og hvor der var mulighed for tilkobling
til aftagende el-netværk, og hvor de økonomiske betingelser
var optimale. I England betød det blandt andet, at de biologiske
undersøgelser, der blev foretaget som del af VVM-redegørelser-
ne, opdagede hidtil ukendte koncentrationer af overvintrende
Sortænder og Rødstrubede Lommer, hvilket i sidste ende satte
en stopper for udviklingen af projekter eller afstedkom store
ændringer, med store økonomiske konsekvenser for projekt-
holderne. I Danmark er vi med tiden blevet bedre til at foretage
marin planlægning ved at kombinere informationer om sejlru-
ter, nedgravede kabler, militære restriktionsområder, skeriin-
teresser, råstondvinding, beskyttede områder og andre om-
råder med vigtige biologiske forekomster (inklusive historiske
informationer om vigtige fugleforekomster) samt beskrivelser
af andre interessegruppers interesser. Med disse er der skabt
mulighed for på mere strategisk vis at undgå konikter med
andre interesser ved etablering af nye havvindmølleparker. Det
er ikke desto mindre af stor vigtighed at gennemføre grundige
undersøgelser af fugleforekomster forud for etablering af nye
havvindmølleparker for at kunne beskrive den aktuelle betyd-
ning af et områdes ornitologiske kvaliteter og sætte disse infor-
mationer i relation til et vindmølleprojekts potentielle eekt på
ywaybestanden af en given fugleart.
EU-direktiv 2001/42/EC fordrer, at der i forbindelse med na-
tionale havvindmølleplaner, der kan have negativ indvirkning
på biodiversiteten, gennemføres en strategisk miljøkonse-
kvensvurdering (SEA). Sådanne strategiske undersøgelser kan
reducere potentielle eekter af vindmølleprojekterne, til glæde
for både industri, administration og generelle brugere af vores
omgivelser. FN’s Espoo-konvention (UNECE 1991) fastsætter
bestemmelser om nationale nabohøringer for projekter, hvor
grænseoverskridende eekter kan komme på tale, fx i forbin-
delse med havvindmølleparker. Når vi taler om trækfugle, så kan
Espoo-høringer blive aktuelle for en række nabolande. VVM-di-
rektivets (EU Directive 85/337/EEC og 97/11/EC) bestemmelser
om miljøkonsekvensvurderinger på projektplan har tendens til
udelukkende at forholde sig til eekten på fuglearter i et meget
afgrænset geogrask område, selv om der er krav om at evaluere
potentielle kumulative eekter. Evaluering af kumulative eek-
ter skal inddrage eekten af aedte aktiviteter, såsom etablering
af ny infrastruktur både til havs og på land. Den skal samtidig
vurdere bidrag til potentielle eekter fra helt andre menneske-
lige aktiviteter som fx forurening, skeri, skibstrak og råstond-
vinding langs en given arts yway. Sådanne krav nødvendiggør,
at vi bliver bedre til at udnytte vores nuværende viden til at vur-
dere eekten af mange havvindmølleparker i kombination og
kombineret med eekten af andre menneskelige påvirkninger
af vores fuglefauna. Samtidig fortsætter de globale temperatu-
rer med at stige, og der er et akut og stigende behov for gene-
rering af fossilfri energi. Der er ingen tvivl om, at havvindmøl-
leparker kan bidrage markant til sådan en CO2-neutral energi.
Det er vores klare overbevisning, at det kan opnås til gavn for det
globale klima og – med grundig strategisk planlægning – uden
at påvirke vores trækfuglebestande unødigt.
References
Agora Energiewende & Sandbag 2018: The European Power Sec-
tor in 2017. State of aairs and review of current develop-
ments. – Sandbag, London and Agora Energiewende, Berlin.
Busch, M., A. Kannen, S. Garthe & M. Jessopp 2013: Consequenc-
es of a cumulative perspective on marine environmental im-
pacts: oshore wind farming and seabirds at North Sea scale
in context of the EU Marine Strategy Framework Directive. –
Ocean Coast. Manage. 71: 213-224.
Chamberlain, D.E., M.R. Rehsch, A.D. Fox, M. Desholm & S. An-
thony 2006: The importance of determining avoidance rates
in relation to the use of wind turbine collision risk models to
predict bird mortality. – Ibis 148: S198-S202.
Cleasby, I.R., E.D. Wakeeld, S. Bearhop, T.W. Bodey, S.C Votier & K.C.
Hamer 2015: Three‐dimensional tracking of a wide‐ranging
marine predator: ight heights and vulnerability to oshore
wind farms. – J. Appl. Ecol. 52: 1474-1482.
Desholm, M. 2009: Avian sensitivity to mortality: Prioritising mi-
gratory bird species for assessment at proposed wind farms.
– J. Environ. Manage. 90: 2672-2679.
Desholm, M. & J. Kahlert 2005: Avian collision risk at an oshore
wind farm. – Biology Lett. 1: 296-298.
Desholm, M., A.D. Fox, P. Beasley & J. Kahlert 2006: Remote tech-
niques for counting and estimating the number of bird-wind
turbine collisions at sea: a review. – Ibis 148: S76-S89.
Desholm, M., R. Gill, T. Bøvith & A.D. Fox 2014: Combining spatial
modelling and radar to identify and protect avian migratory
hot-spots. – Curr. Zool. 60: 680-691.
Dierschke, V. & S. Garthe 2006: Literature review of oshore
wind farms with regard to seabirds. Pp. 131-198 in C. Zucco,
W. Wende, T. Merck, I. Köchling & J. Köppel (eds.): Ecological
Research on Oshore Wind Farms: International Exchange of
Experiences. PART B: Literature Review of Ecological Impacts.
BfN-Skripten 186. Bundesamt für Naturschutz, Bonn.
Dirksen, S. 2017: Review of methods and techniques for eld vali-
dation of collision rates and avoidance amongst birds and bats
at oshore turbines. – Commissioned Report to Rijkswatersta-
at WVL by Sjoerd Dirksen Ecology, Utrecht.
Drewitt, A.L. & R.H.W. Langston 2008: Collision eects of wind‐
power generators and other obstacles on birds. – Ann. N. Y.
Acad. Sci. 1134: 233-266.
Esteban, M.D., J.J. Diez, J.S. López & V. Negro 2011: Why oshore
wind energy? – Renew. Energ. 36: 444-450.
Everaert, J. & E.W.M. Stienen 2006: Impact of wind turbines on
birds in Zeebrugge (Belgium). In: D.L. Hawksworth & A.T. Bull
(eds): Biodiversity and Conservation in Europe. Topics in Biodi-
versity and Conservation, vol 7. Springer, Dordrecht.
Fijn, R.C., K.L. Krijgsveld, M.J. Poot & S. Dirksen 2015: Bird move-
ments at rotor heights measured continuously with vertical
radar at a Dutch oshore wind farm. – Ibis 157: 558-566.
Fox, A.D., M. Desholm, J. Kahlert, T.K. Christensen & I.K. Petersen
2006a: Information needs to support environmental impact
assessment of the eects of European marine oshore wind
farms on birds. – Ibis 148: S129-S144.
Fox, A.D., T.K. Christensen, M. Desholm, J. Kahlert & I.K. Petersen
2006b: Chapter 7. Birds. In: Danish Oshore Wind – Key Envi-
ronmental Issues. – Energistyrelsen, København.
Guillemette, M., J.K. Larsen & I. Clausager 1999: Assessing the im-
pact of the Tunø Knob wind park on sea ducks: the inuence
of food. – NERI Technical Report No. 263, Danish Ministry of
Environment and Energy.
Jensen, F.P., R. Ringgaard, J. Blew & E.M. Jacobsen 2016: Anholt
Oshore Wind Farm. Post-construction Monitoring of Bird
Migration. – Dong Energy Report, Dong Energy, Fredericia,
Denmark.
Joensen, A.H. 1974: Waterfowl Populations in Denmark 1965-
1973. A Survey of the Non-breeding Populations of Ducks,
Swans and Coot and their Shooting Utilization. – Dan. Rev.
Game Biol. 9 nr. 1.
Johnston, A., A.S. Cook, L.J. Wright, E.M. Humphreys & N.H. Burton
2014: Modelling ight heights of marine birds to more accura-

101
Oshore wind farms and birds
tely assess collision risk with oshore wind turbines. – J. Appl.
Ecol. 51(1): 31-41.
Kahlert, J., I.K. Petersen, A.D. Fox, M. Desholm & I. Clausager 2004:
Investigations of birds during construction and operation of
Nysted oshore wind farm at Rødsand: Results and conclusi-
ons, 2003. – NERI Report.
Kerlinger, P. 2000: Avian mortality at communication towers: a re-
view of recent literature, research and methodology. – Curry &
Kerlinger LLC, Cape May, New Jersey.
Krijgsveld, K.L., R.C. Fijn, M. Japink, P.W. van Horssen, C. Heunks
et al. 2011: Eect studies Oshore Wind Farm Egmond aan
Zee. Final report on uxes, ight altitudes and behaviour
of ying birds. – NoordzeeWind Report OWEZ_R_231_T1
20111110_ux&ight. Bureau Waardenburg, Culemborg, The
Netherlands.
Larsen, J.K. & M. Guillemette 2007: Eects of wind turbines on
ight behaviour of wintering common eiders: implications for
habitat use and collision risk. – J. Appl. Ecol. 44: 516-522.
Laursen, K., S. Pihl, J. Durinck, M. Hansen, H. Skov, J. Frikke & F. Da-
nielsen 1997: Numbers and Distribution of Waterbirds in Den-
mark 1987-1989. – Dan. Rev. Game Biol. 15 nr 1.
Leopold, M.F., R.S.A. van Bemmelen & A.F. Zuur 2013: Responses
of Local Birds to the Oshore Wind Farms PAWP and OWEZ
o the Dutch mainland coast. – IMARES Report No. C151/12,
Wageningen University.
Lindeboom, H.J., H.J. Kouwenhoven, M.J.N. Bergman, S. Bouma,
S.M.J.M. Brasseur et al. 2011: Short-term ecological eects of
an oshore wind farm in the Dutch coastal zone; a compila-
tion. – Environ. Res. Lett. 6: 035101.
Martin, G.R. 2011: Understanding bird collisions with man‐made
objects: a sensory ecology approach. – Ibis 153: 239-254.
Masden, E.A. & A.S.C.P. Cook 2016: Avian collision risk models for
wind energy impact assessments. – Environ. Imp. Asses. Rev.
56: 43-49.
Masden, E.A., D.T. Haydon, A.D. Fox, R.W. Furness, R. Bullman & M.
Desholm 2009: Barriers to movement: impacts of wind farms
on migrating birds. – ICES J. Mar. Sci. 66: 746-753.
Masden, E.A., D.T. Haydon, A.D. Fox & R.W. Furness 2010a: Barriers
to movement: modelling energetic costs of avoiding marine
wind farms amongst breeding seabirds. – Mar. Poll. Bull. 60:
1085-1091.
Masden, E.A., A.D. Fox, R.W. Furness, R. Bullman & D.T. Haydon
2010b: Cumulative impact assessments and bird/wind farm
interactions: developing a conceptual framework. – Environ.
Imp. Asses. Rev. 30: 1-7.
Masden, E.A., R.E. Reeve, M. Desholm, A.D. Fox, R.W. Furness & D.T.
Haydon 2012: Assessing the impact of marine wind farms on
birds through movement modelling. – J. Roy. Soc. Interface 9:
2120-2130.
May, R., E.A. Masden, F. Bennet & M. Perron 2018: Considerations
for upscaling individual eects of wind energy development
towards population-level impacts on wildlife. – J. Environ. Ma-
nage. 230: 84-93.
Mendel, B., P. Schwemmer, V. Peschko, S. Müller, H. Schwemmer,
M. Mercker & S. Garthe 2019: Operational oshore wind farms
and associated ship trac cause profound changes in distri-
bution patterns of Loons (Gavia spp.). – J. Environ. Manage.
231: 429-438.
Newton, I. 2007: Weather‐related mass‐mortality events in mi-
grants. – Ibis 149: 453-467.
Petersen, I.K. & A.D. Fox 2007: Changes in bird habitat utilisation
around the Horns Rev 1 oshore windfarm, with particular
emphasis on Common Scoter. – Report commissioned by Vat-
tenfall A/S. National Environmental Research Institute, Aarhus
University.
Petersen, I.K. & A.D. Fox 2009: Faktorer der påvirker fordelingen af
sortænder i fældningsperioden i Ålborg Bugt. – Rapport re-
kvireret af Vattenfall Vindkraft. Danmarks Miljøundersøgelser,
Aarhus Universitet.
Petersen, I.K., T.K. Christensen, J. Kahlert, M. Desholm & A.D. Fox
2006: Final results of bird studies at the oshore wind farms
at Nysted and Horns Rev, Denmark. – Commissioned Report
by DONG Energy and Vattenfall A/S. National Environmental
Research Institute, Denmark.
Petersen, I.K., M.L. MacKenzie, E. Rexstad, M.S. Wisz & A.D. Fox
2011: Comparing pre- and post-construction distributions of
long-tailed ducks Clangula hyemalis in and around the Nysted
oshore wind farm, Denmark: a quasi-designed experiment
accounting for imperfect detection, local surface features and
autocorrelation. – CREEM Technical Report, no. 2011-1, Univer-
sity of St Andrews.
Petersen, I.K., R.D. Nielsen & M.L. Mackenzie 2014: Post-construc-
tion evaluation of bird abundances and distributions in the
Horns Rev 2 oshore wind farm area, 2011 and 2012. – Report
Commissioned by Dong Energy. Danish Centre for Environ-
ment and Energy, Aarhus University.
Petersen, I.K., R.D. Nielsen, O.R. Therkildsen & T.J.S. Balsby 2017:
Fældende havdykænders antal og fordeling i Sejerøbugten i
relation til menneskelige forstyrrelser. – Videnskabelig rapport
fra DCE nr. 239.
Pineda, I. 2018: Oshore Wind in Europe. Key trends and statistics
2017. – Wind Europe, Brussels, Belgium.
Pineda, I. & P. Tardieu (eds) 2018: Wind in power 2017. Annual com-
bined onshore and oshore energy statistics. – Wind Europe,
Brussels, Belgium.
Piper, W., G. Kulik, J. Durink, H. Skov & S.B. Leonhard 2008: Horns
Rev II Oshore Wind Farm Monitoring of Migrating Waterbirds
baseline Studies 2007-08. – Report by Orbicon A/S and DHI A/S
to Dong Energy, Copenhagen, Denmark.
Plonczkier, P. & I.C. Simms 2012: Radar monitoring of migrating
pink‐footed geese: behavioural responses to oshore wind
farm development. – J. Appl. Ecol. 49: 1187-1194.
Poot, M.J.M., P.W. van Horssen, M.P. Collier, R. Lensink & S. Dirk-
sen 2011: Eect studies Oshore Wind Egmond aan Zee:
cumulative eects on seabirds. – NoordzeeWind Report
OWEZ_R_212_20111021_Cumulative_Eects. Bureau
Waardenburg, Culemborg, The Netherlands.
Schwemmer, P. & S. Garthe 2006: Spatial patterns in at-sea be-
haviour during spring migration by little gulls (Larus minutus)
in the southeastern North Sea. – J. Orn. 147: 354-366.
Skov, H. & E. Prins 2001: Impact of estuarine fronts on the dispersal
of piscivorous birds in the German Bight. – Mar. Ecol. Prog. Ser.
214: 279-287.
Skov, H., S. Heinänen, R. Žydelis, J. Bellebaum, S. Bzoma et al. 2011:
Waterbird Populations and Pressures in the Baltic Sea. – Nordic
Council of Ministers, Copenhagen, Denmark.
Skov, H., S. Heinänen, T. Norman, R. Wad, S. Méndez-Roldán & I. Ellis
2018: ORJIP Bird Collision and Avoidance Study. Final Report –
April 2018. The Carbon Trust, UK.
Stone, C.J., A. Webb, C. Barton, N. Ratclie, T.C. Reed et al. 1995: An
atlas of seabird distribution in north-west European waters. –
Joint Nature Conservation Committee, Peterborough UK.
UNECE 1991: Convention on Environmental Impact Assessment
in a Transboundary Context. – United Nations Economic Com-
mission for Europe, Geneva, Switzerland.
Author’s addresses
Anthony D. Fox (tfo@bios.au.dk) & Ib Krag Petersen, Depart-
ment of Bioscience, Aarhus University, Kalø, Grenåvej 14, DK-
8410 Rønde