Technical ReportPDF Available

Reducing bird mortality caused by high- and very-high-voltage power lines in Belgium

October 2012

DOI:10.13140/RG.2.2.11066.90562

Authors:

Antoine Derouaux

Natagora

Joris Everaert

Research Institute for Nature and Forest

Gerald Driessens

Natuurpunt

Show all 6 authorsHide

High- and very-high-voltage power lines have been identified as a major human-induced source of mortality for birds. When birds fly in large groups, or in poor visibility conditions, they can collide with these structures, in some instances in sufficiently large numbers to raise concern. Several international agreements on nature conservation acknowledge this important issue. In answer to this concern, Elia, as the power transmission system operator in Belgium, intends to identify its own overhead power lines that present a serious collision risk for birds. The main objective of this project is to map collision-prone power lines in Belgium and to classify existing power lines according to their “dangerousness” in this regard. This report’s priority classification of power lines should ultimately be seen as an answer to the following question: “Given present-day knowledge of bird distribution and relative sensitivity to collision mortality, what are the most dangerous power-line sections, i.e. those where the mitigation of collision risk should be focused as a priority?” It is neither a substitute for a proper Environmental Impact Assessment nor a mitigation plan in itself. As a first step, a list of collision-sensitive bird species in Belgium was compiled on the basis of a review of the literature, casualty records and expert judgment. Not only “collision risk” itself was taken into account, but also the “conservation value” of each species and the probable population impact of additional mortality. In a second step, this list was used to define four coherent bird species groups of interest, in order to facilitate and organize the mapping work. These groups are the waterbirds (for which numerous data are available at site level), rare breeding birds (for which good data are also available in both Wallonia and Flanders), which migrate in large numbers (the mapping exercise was more difficult for them, because migration is to a considerable extent geographically spread across Belgium, with no clear migrant funnel) and widespread breeding bird species (for which relative density maps need to be built because they are not associated with specific sites or locations). In a third step, up-to-date knowledge of bird distribution in Belgium was used to create maps for each of the coherent bird species groups. Numerous sources of data were used: wintering waterbird counts, roost and colony counts, breeding bird atlases and observations from datarecording portals. Spatial modelling was applied to obtain high-resolution maps of widespread species and help define the best staging area for some other species. A “collision risk score” was defined for each of the geographical zones delimited on the maps, or based on the distance from important bird areas. The combination of the maps, with the aid of this scoring system, allows the classification of all power-line masts according to their associated collision risk for birds. Most of the more dangerous power lines are located in areas with major concentrations of waterbirds occur: the polders area, the wetlands in the vicinity of the Port of Antwerp and some river valleys (Yser, Meuse, Haine). About 3.4% of the network can be considered “high priority” for mitigation. A technical chapter describes the available mitigation tools. In the case of existing power lines, a cost-effective manner to decrease collision risk involves placing markers or “diverters” on the lines, in order to make them more visible to birds, even in poor visibility conditions. On the basis of current knowledge, it can be concluded that any large device (increasing the apparent size of the line to at least 20 cm), placed at least every 5-10 m along the line, preferably on the earth wire, is likely to significantly reduce collision risks. In view of the available budget, the placement of diverters can now be planned on Belgian power lines, focusing first on the high priority sections. Of course, other considerations, such as geographical “grouping” of diverters in order to reduce cost and opportunities provided by other planned work on specific power-line sections, may also influence the final planning, which is now in Elia’s hands.

Conceptual framework and work plan for the prioritization of power lines according to the bird collision risk.

…

: Critical areas for wintering waterbirds in Belgium (in violet "Very Important Sites", in red "Important sites" and in orange "Fairly Important Sites").

…

Important night-roosts of waterbirds in Belgium (in red, the "very important" roost locations, in orange the "important" roost locations). Please note that the polygon outlines are larger than scale to make them easier to see. "Fairly important roosts" are not presented on this map.

…

Sites with sensitive rare breeding species, with an indication of the number of rare breeding species.

…

Map showing locations of recorded active migrants of two sensitive species-in green, the Common Crane, in red the Woodpigeon (source: Aves database, observations.be, years 2009 and 2010). Both species can migrate over the whole country, but large groups are more likely to be observed in Eastern Belgium for the Crane and over the Condroz and Famenne ecoregion for the Woodpigeon, especially around the Meuse valley and the Northern ridge of the Ardennes region.

…

Figures - uploaded by Jean-Yves Paquet

Content may be subject to copyright.

Content uploaded by Jean-Yves Paquet

Content may be subject to copyright.

-h

-v

Final Report – 31 October 2012

Authors: Antoine DEROUAUX 1, Joris EVERAERT 2, Nicolas BRACKX 3, Gerald DRIESSENS 4, Alberto

MARTIN GÌL 1, Jean-Yves PAQUET 1

1 Aves-Natagora, Département Études

2 Instituut voor Natuur- en Bosonderzoek (INBO)

3 Vogelbescherming Vlaanderen

4 Natuurpunt Studie

This project was initiated by Elia and coordinated by Aves-Natagora, in collaboration with

Natuurpunt, INBO and Vogelbescherming Vlaanderen.

Recommended citation: DEROUAUX, A., EVERAERT, J., BRACKX, N., DRIESSENS, G.,

MARTIN GÌL, A., PAQUET, J.-Y. (2012): Reducing bird mortality caused by high- and very-high-

voltage power lines in Belgium, final report, Elia and Aves-Natagora, 56 pp.

Reducing bird mortality caused by high- and very-high-voltage power lines

Table of Contents

Executive summary ........................................................................................................................5

I. Introduction and objectives .....................................................................................................7

II. Bird mortality caused by high- and very-high-voltage power lines - an overview ........................9

II.1. Why is reducing collision mortality important?....................................................................9

II.2. Mortality caused by power lines - the nature of the risk ..................................................... 10

II.3. Future focus ...................................................................................................................... 13

III. Step 1: Classifying birds of Belgium according to their sensitivity to collision risk .................. 14

III.1. International significance of Belgian avifauna .................................................................. 14

III.2. Information sources and classification procedure .............................................................. 14

IV. Step 2: Definition of species groups used in the mapping/prioritization procedure ................... 19

V. Step 3: Compiling maps of critical areas for sensitive species ................................................... 20

V.1. Some precautionary remarks ............................................................................................. 20

V.2. Waterbirds ........................................................................................................................ 20

V.3. Rare breeding birds ........................................................................................................... 25

V.4. Migrants in large numbers ................................................................................................. 28

V.5. Local movement of birds ................................................................................................... 30

V.6. Widespread breeding species............................................................................................. 31

V.7. Other important stop-over areas for plovers ....................................................................... 34

VI. Step 4: Using the maps to prioritize power lines ...................................................................... 36

VII. Mitigation techniques to reduce collisions between birds and power lines .............................. 39

VII.1. Burying cables................................................................................................................ 39

VII.2. Removing static wire ...................................................................................................... 39

VII.3. Habitat management ....................................................................................................... 39

VII.4. Marking lines ................................................................................................................. 40

VIII. Conclusions ......................................................................................................................... 43

Annex I. Classification of Belgian birds according to the power-line collision risk ........................ 44

Annex II: Collision victims brought into a recovery centre ............................................................ 50

Annex III: Bibliography ................................................................................................................ 51

Reducing bird mortality caused by high- and very-high-voltage power lines

Executive summary

· High- and very-high-voltage power lines have been identified as a major human-induced

source of mortality for birds. When birds fly in large groups, or in poor visibility conditions,

they can collide with these structures, in some instances in sufficiently large numbers to raise

concern. Several international agreements on nature conservation acknowledge this important

issue. In answer to this concern, Elia, as the power transmission system operator in Belgium,

intends to identify its own overhead power lines that present a serious collision risk for birds.

· The main objective of this project is to map collision-prone power lines in Belgium and to

classify existing power lines according to their “dangerousness” in this regard. This report’s

priority classification of power lines should ultimately be seen as an answer to the following

question: “Given present-day knowledge of bird distribution and relative sensitivity to

collision mortality, what are the most dangerous power-line sections, i.e. those where the

mitigation of collision risk should be focused as a priority?” It is neither a substitute for a

proper Environmental Impact Assessment nor a mitigation plan in itself.

· As a first step, a list of collision-sensitive bird species in Belgium was compiled on the basis

of a review of the literature, casualty records and expert judgment. Not only “collision risk”

itself was taken into account, but also the “conservation value” of each species and the

probable population impact of additional mortality.

· In a second step, this list was used to define four coherent bird species groups of interest, in

order to facilitate and organize the mapping work. These groups are the waterbirds (for which

numerous data are available at site level), rare breeding birds (for which good data are also

available in both Wallonia and Flanders), which migrate in large numbers (the mapping

exercise was more difficult for them, because migration is to a considerable extent

geographically spread across Belgium, with no clear migrant funnel) and widespread breeding

bird species (for which relative density maps need to be built because they are not associated

with specific sites or locations).

· In a third step, up-to-date knowledge of bird distribution in Belgium was used to create maps

for each of the coherent bird species groups. Numerous sources of data were used: wintering

waterbird counts, roost and colony counts, breeding bird atlases and observations from data-

recording portals. Spatial modelling was applied to obtain high-resolution maps of widespread

species and help define the best staging area for some other species.

· A “collision risk score” was defined for each of the geographical zones delimited on the maps,

or based on the distance from important bird areas. The combination of the maps, with the aid

of this scoring system, allows the classification of all power-line masts according to their

associated collision risk for birds.

· Most of the more dangerous power lines are located in areas with major concentrations of

waterbirds occur: the polders area, the wetlands in the vicinity of the Port of Antwerp and

some river valleys (Yser, Meuse, Haine). About 3.4% of the network can be considered “high

priority” for mitigation.

· A technical chapter describes the available mitigation tools. In the case of existing power

lines, a cost-effective manner to decrease collision risk involves placing markers or

“diverters” on the lines, in order to make them more visible to birds, even in poor visibility

conditions. On the basis of current knowledge, it can be concluded that any large device

(increasing the apparent size of the line to at least 20 cm), placed at least every 5-10 m along

the line, preferably on the earth wire, is likely to significantly reduce collision risks.

· In view of the available budget, the placement of diverters can now be planned on Belgian

power lines, focusing first on the high priority sections. Of course, other considerations, such

as geographical “grouping” of diverters in order to reduce cost and opportunities provided by

other planned work on specific power-line sections, may also influence the final planning,

which is now in Elia’s hands.

Reducing bird mortality caused by high- and very-high-voltage power lines

I. Introduction and objectives

Power lines and other wire infrastructures have been identified as a mortality source for birds for more

than 130 years. Back in 1876, an American naturalist found more than a hundred Horned Larks

Eremophila alpestris killed by telegraph wires in Colorado, and he drew this sad conclusion (C

OUES,

1876): "Usually, a remedy has been or may be provided for any unnecessary or undesirable

destruction of birds; but there seems to be none in this instance. Since we cannot conveniently abolish

the telegraph, we must be content with fewer birds”.

There has been a huge increase in the use of wire systems around the world, with these now covering

an estimated 65 million km for medium- to high-voltage lines only (JENKINS ET AL., 2010).

Meanwhile, there has also been a lot of progress made in the understanding and prevention of injury to

birds by power lines. Several international agreements on nature conservation acknowledge this

important issue. Among these, Recommendation No. 110 of the Bern Convention on the Conservation

of European Wildlife and Natural Habitats states that signatories should take cost-effective measures

to reduce bird mortality caused by power lines.

In answer to this recommendation and also with a general purpose of minimizing the impact of grid

development on wildlife, Elia, the power transmission system operator in Belgium, intends to identify

its own overhead power lines that present a serious collision risk for birds. It is very important to

identify those power lines with high collision risks because although solutions do exist to prevent

collision, it is necessary to prioritize their implementation in order to maintain cost-effectiveness

(BEVANGER ET AL., 2009).

The present project, initiated by Elia and coordinated by Aves-Natagora, in collaboration with

Natuurpunt, INBO and Vogelbescherming Vlaanderen, intends to map collision-prone power lines in

Belgium. This project meets recommended action II.4 of the recently published Budapest Declaration

on bird protection and power lines, aiming at prioritizing power lines for mitigation “in accordance

with bird distribution data and in consultation with relevant government, industry, academic and NGO

experts”.

To achieve this goal, a conceptual framework was developed (Figure 1). As a first step, a list of bird

species reported in Belgium and their sensitivity to collision to power lines has been compiled on the

basis of a review of the literature and casualty records. In a second step, based on this list, four bird

species groups of interest were defined in a coherent manner with available distribution data. In a third

step, up-to-date knowledge of bird distribution in Belgium was used to create a map of critical areas

for bird/power line collision risks, using methodologies similar to the recent sensitivity mapping

exercise for wind energy in Flanders (EVERAERT ET AL., 2011). Finally, a scoring system has been

applied to high-voltage masts, according to their relative location on the sensitivity map. This allowed

us to highlight high-risk power-line masts, in order to help Elia to define priority sectors for

mitigation. Finally, technical solutions for mitigation are suggested.

Figure 1: Conceptual framework and work plan for the prioritization of power lines according to the bird

collision risk.

It is very important to note that this report does not intend to propose a mitigation plan in itself. It is

merely an aid to decision-making, offering a framework to classify power-line sections according to

the risk they present. The report also describes state-of-the-art mitigation techniques, but defining a

proper mitigation plan for power lines in Belgium lies beyond the scope of this report.

The sensitivity map developed in the third step of this process is not only a useful decision-making aid

to mitigate present-day collision risks – it could also be used during the planning process of future grid

development. Indeed, even if the power line mortality rarely directly affects population dynamics of

impacted species (see below), the mere presence of industrial components, such as high-voltage power

lines, in the landscape will reduce the available habitat for birds or reduce the quality of the available

habitat (SOSSINKA & BALLASUS, 1997). Among the recognized examples, some meadow birds are

known to avoid surroundings of the lines for breeding (ALTEMÜLLER & REICH, 1997) and wintering

geese use the vicinity of power line rights of way significantly less (KREUTZER, 1997). Minimizing the

development of future power lines in bird-rich areas is thus of prime importance. Where building new

lines is inevitable, preventive measures in the line design (such as undergrounding lines) can be

usefully implemented (JENKINS ET AL., 2010). However, the mapping exercise presented in this report

is not a substitute for a proper Environment Impact Assessment: the local impact of installing new

power lines must be assessed using local data, and the prioritization exercise here, targeting the

country level, may not be relevant in this context.

Before starting to describe the four steps of the work that have been conducted (chapter III to VI), we

start with a general overview of the issue (chapter II). This short review is valuably complemented by

the recent detailed review published for the AEWA and CMS by Bureau Waardenburg (PRINSEN ET

AL., 2011a). Finally, in Chapter VII, we enumerate the possible solutions concerning the identified

“collision-prone” power lines in Belgium.

Review of the literature

Casualty records

Expert

assessment

Species classification (n=450)

(according to (1) collision sensitivity, (2) conservation

relevance and (3) risk of massive mortality)

Definition of four

"species groups of

interest"

Maps of critical areas

(for each group of interest)

Prioritization of power-

line sections

Step 1

Step 2

Availability/relevance

of distribution data

Step 3

Step 4

Scoring

system

Distribution

data

Reducing bird mortality caused by high- and very-high-voltage power lines

II. Bird mortality caused by high- and very-high-voltage

power lines - an overview

II.1. Why is reducing collision mortality important?

Man-made structures create a high collision risk for wild birds all around the world. Mortality due to

these collisions is a well-described phenomenon, at least in North America and Europe (B

EVANGER,

1998). Despite the difficulties inherent to mortality estimations, estimated casualty numbers are very

high. For example, during one year in the USA, communication masts may kill 4 to 50 million birds,

collisions with power lines may kill more than 175 million birds, tens to hundreds of thousands of

birds are electrocuted and wind turbines also kill thousands of birds… (MANVILLE II, 2005). All in all,

an estimated one billion birds are killed annually in the USA by anthropogenic causes, including about

58% from collisions with buildings (mostly window strikes) and about 14% with power lines

(ERICKSON ET AL., 2005b). In France, over 20 years 4,895 casualties (collisions and electrocutions)

have been reported by naturalists, but this figure is a considerable underestimate because it is not

obtained through systematic searches but rather through a compilation of reported incidents; 47% of

the reported cases came from Provence – Alpes – Côte d’Azur (KABOUCHE ET AL., 2006). The LPO

claims 5 to 10 million birds are killed every year by power lines in France, which seems realistic given

the 90,000 km of high- and very-high-voltage power lines in that country (see hereafter for an

estimation of mortality/km/year).

However, the actual impact of these high mortality rates on bird population dynamics is not easily

established (BEVANGER, 1998; BAROV, 2011). The fundamental question here is to determine if

human-induced mortality is offset by better survival and reproduction rates of the survivors

(compensated mortality) or if this mortality is simply adding to the “natural” mortality and not offset

by population dynamics (additive mortality). The second option could of course constitute a

significant problem for nature protection if this concerns endangered species for example. Technically,

this is not a question that is easy to answer because population dynamic parameters (mortality,

fecundity, etc.) are complicated and costly to measure in the field, and even more so if these

parameters need to be compared between different populations, subject to various collision-risk levels

(BEVANGER, 1999).

An indication of the significance of the “population impact” of collision mortality could consist of

comparing it with other mortality causes. This can also give some understanding of how power-line

mortality compares with other human uses of birds, such as hunting. For example, collisions with

power lines kill 2.4 times more Ptarmigan (Lagopus sp.) than hunting, according to a study in

subalpine area of Norway (BEVANGER & BROSETH, 2004). An indirect effect of power-line mortality

has also been demonstrated in a Bonelli’s Eagle Hieraaetus fasciatus population: eagle pairs having

high-risk lines at close range of their nest showed turnover rates twice as high as pairs not in the

vicinity of such lines (MAÑOSA & REAL, 2001). A comparative and very detailed study on the Eagle

Owl in the Italian Alps (focusing on electrocution risks) concluded that, although interaction with

resource availability is complicating the picture, there are significant effects of electrocution mortality

on the population level, including population decrease in high-risk areas (SERGIO ET AL., 2004). In a

very recent review of transmission infrastructures and Natura 2000, some examples of population-

level impact on globally endangered species relating to collision mortality are given. This concerns

endangered geese species (but in that case, the actual population effect is only suspected and not

quantified), the Little Bustard and Great Bustard and the raptor cases explained above (BIO

INTELLIGENCE SERVICE, 2012).

For smaller birds, like passerines, a population impact is less likely, even if mortality figures are

sometimes very high (SCOTT ET AL., 1972), because their population dynamics depend more on a high

fecundity than on long-term survival. However, even if there are no clear population effects, hence no

definite impacts on nature conservation, several arguments can be made to defend a precautionary

approach that intends to minimize collision risks with birds in all cases, and not only where

endangered populations of emblematic species are affected:

· There is always the risk of having very dangerous power lines locally, depleting a

local population and eventually creating a “population sink”.

· There is an ethical and moral obligation to reduce human-induced mortality.

· With the anticipated necessary change to “green energy” production, transmission

grids are expected to constantly grow in the near future (BAROV, 2011; BIO

INTELLIGENCE SERVICE, 2012). Therefore, increasing public support for such a

development, e.g. by minimizing bird fatalities on existing power lines, is necessary.

In the following section, we will detail the nature of power-line collision risks for birds, emphasizing

the factors determining the risk level, to explain how we compiled the sensitive species list for

Belgium.

II.2. Mortality caused by power lines - the nature of the risk

II.2.1. Electrocution versus collision risks

Power lines are divided into four categories, according to their transmission voltage:

· very high voltage (225 to 400 kV);

· high voltage (50 to 225 kV);

· medium voltage (15 to 50 kV);

· low voltage (220 and 380 V)

Medium- and low-voltage lines mostly induce electrocution risks for birds. Several bird species use

the pylons to perch on: as a hunting post, resting place or even for nesting. Most fatalities occur in

long-winged species (storks, raptors, etc.) which can make contact with two phases simultaneously (or

one phase and the earth wire). Some types of pylons, in particular metallic structures, are more

dangerous than others and the electrocution risk increases when the distance between conductors

decreases. Short-circuits caused by birds occasionally create bush fires in dry areas (KABOUCHE ET

AL., 2006).

Although high- to very-high-voltage lines are almost never insulated, the large distances between

wires generally prevent electrocution risk for birds; the main risk is represented by the collision

hazard. During their flight, birds simply collide with the wires and die (factors inducing this risk are

explained below). Other impacts of high-voltage lines on birds, like the putative effect of strong

electric and magnetic fields, have been judged insignificant (SILNY, 1997), but the possible impact on

habitat availability is a growing concern, although little studied (BAROV, 2011).

In Belgium, Elia operates the high- and very-high-voltage grid system; therefore, the rest of this report

will focus on collision risks.

II.2.2. Quantifying the number of killed birds

Several authors have tried to assess the mortality due to power lines by counting dead birds

encountered under the wires. A large number of biases are associated with such counts (ERICKSON ET

AL., 2005b). Size and coloration of the carcass may induce strong variation in detectability rate

between species. Dense vegetation beneath the lines may hamper research and, even more importantly,

carcass disappearance or scavenger removal rates can be very high. Where these removal rates have

been experimentally studied, most carcasses were eliminated in a matter of a few days (SCHICKER,

1997; ERICKSON ET AL., 2005b). Therefore, mortality estimations always have to be made with

caution.

In the Netherlands, with 4,600 km of high-voltage power lines, 750,000 to 1,000,000 birds are

estimated to be killed by collision each year, meaning an average of 163-217 fatalities per km per year

(ERICKSON ET AL., 2005a). There are differences between habitats: on grassland there are 113

collisions/km/year; on agricultural land 58 collisions/km/year, and near river crossings

489 collisions/km/year. Elsewhere in Europe, there are few studies of that kind, but interestingly in

Belgium Elia studied the impact of 3 km of a 70,000-V power line in Western Flanders (MORTIER,

2000). Between 5 March and 1 April 2001, 30 dead birds were found (among 25 birds identified, a

quarter were gulls Larus sp.; other species included two Grey Herons Ardea cinerea and a Mute Swan

Cygnus olor). Using adjustment techniques because of scavengers removing corpses, the author

estimated that 468 birds were killed each year under these 3-km lines. This leads to an estimated

Reducing bird mortality caused by high- and very-high-voltage power lines

collision risk of 157 birds/km/year, a figure comparable with the studies in the Netherlands. However,

the situation in Flanders and the Netherlands may be worse than average because of the abundance of

waterbirds in these regions. In habitats without wetlands, death rates are probably lower, as shown in a

Spanish study, averaging about 30 killed birds/km/year (ALONSO & ALONSO, 1999) and in general can

vary from 0.1 to 80 casualties/km/year (JENKINS ET AL., 2010).

Taking a conservative figure of 30-100 dead birds/km/year in Belgium over the total length of high-

voltage grid of 5,614 km in Belgium (Elia website), we come to an extrapolated range of 170,000 to

500,000 birds killed every year nationwide.

II.2.3. Conditions increasing bird strikes with power lines

This rather high mortality figure is, however, hiding a diversity of situations: not all species of birds

are similarly affected and not all power lines are equally dangerous. In order to focus actions on the

most dangerous sites, it is necessary to understand the factors and conditions that create a higher

collision risk. The following paragraphs first set out the morphological and behavioural characteristics

of birds that are collision-prone and then review the power lines and environmental situations

increasing that risk.

In short, collisions are more frequent when:

· power lines cross wetlands;

· power lines cross forests;

· power lines cross the wintering area of large flocking birds;

· power lines are multistage (two or three wires and/or an earth wire);

· weather conditions are bad (fog, rain, snow, wind, etc.);

· adults feed their young;

· fledglings are learning to fly;

· birds are migrating by night.

Bird morphology

Nearly all bird species are affected by collision (from the small Goldcrest Regulus regulus to big

swans Cygnus sp.), but some are hit more than others. The sensitivity of a particular bird species

depends on its morphology and its flight behaviour. Nocturnal birds, social birds and birds with aerial

courtships are more sensitive to collision (BEVANGER, 1998).

It is possible to define groups of birds based on their morphology and to predict the relative sensitivity

to collision of these groups (BEVANGER, 1998). “Poor fliers” (grouses, rails, pheasants, etc.) have a

high collision susceptibility. These birds spend most of their life on the ground and fly only if

necessary (i.e. disturbance). Their low flight agility is a problem for avoiding wires easily.

The morphological approach has been further tested using casualty records from Spain (JANSS, 2000).

Four morphological parameters (length, weight, wing span and tail length) can be used to classify

birds according to discriminant analyses and to predict their collision sensitivity, electrocution

sensitivity or their “mixed sensitivity”. This model, which was further backed up by Italian data

(RUBOLINI ET AL., 2005) shows that the main morphological determinant of risk exposure is the “wing

loading value”: heavy birds with small wings are more prone to collision with high-voltage power

lines.

Although medium-voltage lines are dangerous for raptors (BAYLE, 1999), this bird category has a high

conservation value and is relatively unaffected by collisions with high- and very high-voltage power

lines, probably because they have the lowest wing loading. They represent only 0.1 – 0.4% of dead

birds found under these wires in the Netherlands, Germany, France and Italy.

Collision also depends on the visual field of the bird. A recent study measured the visual field of the

Kori Bustard Ardeoti kori, the Blue Crane Anthorpoides paradisea and the White Stork Ciconia

ciconia, three species sensitive to collisions with power lines (MARTIN & SHAW, 2010). It appears that

these birds have a narrow and vertically long binocular visual field (they grab food items with their

bill using visual guidance). Blind spots above and below the binocular fields differ between these

species. A bustard looking down for food when flying will have a blind spot moving its head only 25°

down; for a crane this is 35°, and a stork 55°. This means that storks are more likely than bustards and

cranes to detect power lines and other obstacles in front of them in flight. This study suggests that

some birds are very vulnerable and that even collision mitigation with special marks might not be

useful because some birds do not see the obstacle at all.

Bird behaviour

Even if some birds are “anatomically” more prone to collision, as explained above, actual risk

exposures also depend on the usual behaviour of the bird. For example, in Spain, cranes might suffer a

higher rate of mortality than bustards, because they more commonly cross over a power line and in

spite of their better flight capacity (JANSS, 2000).

The behaviour of the birds is in turn influenced by their life cycle. A Spanish report showed there were

few or even no fatalities caused by power lines in the case of resident individuals, as opposed to

migrant birds visiting the same areas (ALONSO & ALONSO, 1999). Migrant and over-wintering birds

are probably less familiar with obstacles and are therefore more vulnerable.

During migration, especially nocturnal migration, even small passerines can be exposed to collision, as

a striking example in Dungeness (a coastal site of South-East England) shows. In the vicinity of this

bird observatory, two 400-kV power lines were erected, and many birds were subsequently found dead

(SCOTT ET AL., 1972). Between January 1964 and November 1970, 1,285 corpses of 74 species were

found under the wires, especially nocturnal migrants (rails, thrushes, warblers, etc.). Diurnal migrants

(swift, hirundines, larks, pipits, wagtails and finches) are apparently less sensitive, except when they

have to fly low because of bad weather conditions.

In Illinois (USA), 200 to 400 individual waterfowl birds were killed each autumn by colliding with

high-voltage power lines that cross a small slag pit (32 ha) of a power plant next to a 872-ha lake

(ANDERSON, 1978). Several risk factors were identified that increase exposure: number of birds

present (large groups increase collision risks), low visibility (waterfowl almost never collided during

daylight when visibility was good), species composition or behaviour of birds, disturbance and

familiarity of birds with the area.

Collision can also be related to reproductive stage. Common Terns Sterna hirundo seem to suffer only

low rates of direct mortality through collision. However, young birds and feeding adults are more

vulnerable because they fly closer to the cables more frequently. When adults raise their pulli, they

tend to take the shortest way between the colony and the feeding areas, so they take more risks

crossing wires or flying near the cables (HENDERSON ET AL., 1996). Young and inexperienced birds are

also more exposed. For example, Grey Herons Ardea cinerea are more likely to collide with power

lines from August to December (71% of the recorded mortality are first-year birds (ROSE & BAILLIE,

1989)).

Finally, disturbance increases collision risk, as birds suddenly have to leave a roosting or feeding site.

In Wallonia, an individual rare Whooper Swan Cygnus cygnus (the only regular wintering group in the

region) was killed when colliding with a medium-voltage line after disturbance of the swan group by a

careless photographer (M. Ittelet, pers. com.)

Local topography and situation of the power lines

Local topography is an important factor determining collision events. Five types of collision-prone

effects are identified by KABOUCHE et al. (2006):

· Sliding effect: flying birds follow the relief and collide with a power line if it crosses their

flyway perpendicularly.

· Barrier effect: birds following a valley collide with a line that crosses it perpendicularly.

· Obstacle effect: when a group of birds, flying in an open area, collide with a line when

landing or when taking off.

· Springboard effect: flying birds avoiding an obstacle (relief, hedge, bridge, first power

line, etc.) can collide with the wires if they are hidden by the first obstacle.

· Top effect: birds flying towards a summit collide with lines along the summit ridge.

Power lines in forests can be particularly dangerous for birds because of the open area that they create.

These places can attract a lot of birds for hunting in an open area (KABOUCHE ET AL., 2006). Forest

fragmentation created by the rights of way can also force forest birds to cross the open stretches to

reach other parts of their territory and by doing so, they are also exposed to collision. However, if the

Reducing bird mortality caused by high- and very-high-voltage power lines

line crosses a forest with a tree height equal to or exceeding power pylon height, the collision risk is

minimal for birds flying over the canopy (JENKINS ET AL., 2010).

Some studies have suggested that if power lines were associated with other man-made visible

structures like highways, the collision risk would decrease because birds are more cautious around

these artefacts: this is apparently the case for power lines positioned alongside roads

(HOERSCHELMANN ET AL., 1997).

Finally, several papers logically point out that if power lines are located in habitats or sites where birds

tend to gather, like wetlands, the risk is significantly increased (ALONSO & ALONSO, 1999;

KABOUCHE ET AL., 2006).

Type of wires / structures

Collisions occur with all kind of power lines, but mainly with the finest cables (medium and low

voltage, lighting protection cables, etc.). Some kinds of cables are very difficult to see for birds:

copper lines 40/10 are green with a small diameter and are very dangerous for them (KABOUCHE ET

AL., 2006).

The collision rate also depends on power-line design. Lines with multiple states are believed to be

more dangerous for birds. In Norway, an experiment showed a significant reduction in the number of

bird collisions (mainly Ptarmigans – Lagopus sp.) after the removal of the earth wire (BEVANGER &

BRØSETH, 2001). However, in Spain, no significant difference in the bird collision rate was observed

when comparing a transmission line of 380 kV (with three cable levels and two overhead static wires),

a 132-kV distribution line (one cable level and no static wire) and a 13-kV distribution line (one cable

level and no static wire) (JANSS & FERRER, 1998). It must be noted that this last study also concerned

three different habitats.

Of course, the presence of visual markers or other devices can prevent some of the collision events; we

will come back to this particular point in the last section of the report.

Other external causes

Weather conditions influence bird flight altitude. Visibility of power lines decreases during bad

weather: wind, fog, rain, snow, and so on (KABOUCHE ET AL., 2006).

II.3. Future focus

To date, the impacts of high-voltage power lines on biodiversity have been studied mostly from a bird

mortality (by collision) point of view, and with the purpose of avoiding those collisions. However, we

know little about the actual impact of power lines at the population level (NEGRO, 1999). However,

this should not hamper actions to reduce mortality in wild birds, especially if action is focused on

highly sensitive sites.

Planning of future grid development with the help of a “bird sensitivity map” may further reduce the

impact on bird habitats, although this aspect of power-line effects has been studied too little (BIO

INTELLIGENCE SERVICE, 2012).

III. Step 1: Classifying birds of Belgium according to

their sensitivity to collision risk

In this chapter, after briefly depicting Belgian avifauna and its relative international importance, we

will explain the procedure used to classify species in relation to their supposed relevance to the issue

of collision with power lines. This classification will form the basis for a power-lines sector

prioritization procedure.

III.1. International significance of Belgian avifauna

Including rare and accidental species, about 450 different bird species have been reported in Belgium.

Although a small and densely populated country, Belgium hosts no fewer than 94 regular breeding

birds of European Conservation Concern, of which 46 have an unfavourable conservation status in

Europe (HEATH & EVANS, 2000). Altogether, there are 194 breeding species in Belgium (including

exotics): 183 in Flanders (VERMEERSCH ET AL., 2004), 173 in Wallonia (JACOB ET AL., 2010) and 103

in Brussels (WEISERBS & JACOB, 2007). In Wallonia, breeding populations of several raptor species

(including the Honey Buzzard Pernis apivorus, a Bird Directive Annex I species) and forest species

(including the Middle Spotted Woodpecker Dendrocopos medius and the Black Stork Ciconia nigra)

are of special conservation importance, because their density in Wallonia is higher than the European

average (PAQUET & JACOB, 2011). In Flanders, several waterbirds have a major breeding population,

and in particular three species regularly exceed the threshold of 1% of the entire biogeographical

population breeding in the Port of Zeebrugge (EVERAERT & STIENEN, 2007).

During the winter, waterbird populations of international importance (several species of geese and

ducks) are observed in Flanders. Several sites support more than 1% of the biogeographic population

of at least one species. For example, the polder complex north of Bruges is home to more than 60% of

the total biogeographic population of Pink-footed Geese Anser brachyrhynchus. An estimated 350,000

waterbirds winter in Belgium, mostly in Flanders (HEATH & EVANS, 2000).

Due to its central position in Europe and on the southern edge of the North Sea, millions of birds also

travel across Belgium during pre-nuptial and post-nuptial migrations, some of them without a stop or

roosting only for a few hours/days. Counts of migrating birds from fixed observation points reached an

impressive total of 5.4 million individuals in 2010 (see www.trektellen.org). Locally, migration can be

very intense, such as on some of the Ardennes crests, where migration flow amounts to an average of

750 migrating individuals per watching hour (DERMIEN & PAQUAY, 2000). Species like Cranes Grus

grus have quite narrow migration corridors and most of the Western European population of this

species probably flies over Eastern Belgium during each migration trip.

III.2. Information sources and classification procedure

The species list that was used as a basis for our classification was the list of birds observed in Belgium

provided by the Rarity Committees, available on their websites (http://www.la-ch.be/ for the Walloon

committee and http://www.bahc.be/ for the Flemish committee). As the systematic and taxonomic

sciences are currently very dynamic fields in ornithology, there are some open questions concerning

the exact status of some taxa (i.e. the subspecific/specific status of the Bean Geese Anser [fabalis]

fabalis and Anser [fabalis] rossicus) but this has no real relevance for the present work.

We decided not only to focus on sensitivity to collision with power lines, but also to include the

concept of “conservation relevance” of the collision for the species. If Belgium only has a marginal

population of a species, there is no point in considering it in the prioritization procedure. In contrast, a

very common species with no conservation issue must not be included either in a list which intends to

focus on the risk to biodiversity. However, even for very common species, high mortality numbers can

be seen as detrimental from an ethical point of view. Therefore, we also included indicative “group

criteria” to account for species with aggregative social behaviour that can cause the species to suffer

numerous casualties.

Reducing bird mortality caused by high- and very-high-voltage power lines

Please note that the time period relevant for our conservation relevance and sensitivity classification is

2001 to 2011, i.e. the same as the period considered for the mapping exercise (see Chapter V). Future

knowledge about bird sensitivity to collision and evolution of bird conservation status may trigger the

need for an update of this classification.

The Table 1 summarizes the considered criteria and their different levels used in the classification of

species, which are detailed below. We chose to use a simple “classification level” (only three levels in

the sensitivity to collision), because more detailed levels would have been difficult, and also mostly

arbitrary, to apply to a 450-species list. As already stated, the likelihood or lack of likelihood of an

impact on population level is particularly difficult to judge since there is an absence of hard evidence

in most cases. The goal was simply to focus the subsequent mapping procedure on the most relevant

species, and to help define the most significant “group of species” used in the mapping exercise (step 2

and 3).

Table 1: Criteria and classification levels used in the listing procedure.

Sensitivity to collision (COLS)

Null, almost

never cited in mortality studies or in review

Sometimes cited in studies as found injured or dead, impact on

population unlikely

Regularly cited in studies as found injured or dead, possible

impact on population

Presence in Belgium (BEL)

Presen

t all year round

Only present during the breeding season

igrant

, staging or migrating species

Occasional

or very rare

Conservation relevance (CONS)

scarce, population in Belgium not to be regarded as

relevant for conservation of the species or, on the contrary,

very common species with no conservation issue

elevant

for conservation

ighly

relevant for conservation (rare or declining species,

Belgium hosting an important part of the population, and/or

international obligation to protect the species)

Could be present in large group (GROUP)

risk of high mortality due to aggregative behaviour

ome

risk of high mortality for large groups

III.2.1. The collision sensitivity criterion (COLS)

This criterion evaluation was mainly based on a review of the literature (see Chapter II), because

systematic surveys of power line-related bird mortality in Belgium are rare. The Table 2, from the

Bern Convention document, as well as more recent review were used (PRINSEN ET AL., 2011a; BIO

INTELLIGENCE SERVICE, 2012).

Table 2: List of bird families that should tend to be focal species for environmental assessments where they are

at risk as they are considered to be particularly sensitive, or potentially so, to power lines (according to the Bern

Convention). Key: 0 - no casualties reported or likely; I - casualties reported, but no apparent threat to the bird

population; II - regionally or locally high casualties; but with no significant impact on the overall species

population; III - casualties are a major mortality factor; threatening a species with extinction, regionally or on a

larger scale.

(a) due to electrocution (b) due to collisions

Loons (Gaviidae) and Grebes

(Podicipedidae) 0 II

Shearwaters, Petrels (Procellariidae) 0 I - II

Bobbies, Gannets (Sulidae) 0 I - II

Pelicans (Pelicanidae) I II - III

Cormorants (Phalacrocoracidae) I II

Herons, Bitterns (Ardeidae) I II

Storks (Ciconidae) III III

Ibisses (Threskiornithidae) I II

Flamingos (Phoenicopteridae) 0 II

Ducks, Geese, Swans, Mergansers

(Anatidae) 0 II

Raptors (Accipitriformes and

Falconiformes)) II – III I - II

Partridges, Quails, Grouses

(Galliformes) 0 II - III

Rails, Gallinules, Coots (Rallidae) 0 II - III

Cranes (Gruidae) 0 II - III

Bustards (Otidae) 0 III

Shorebirds / Waders (Charadriidae +

Scolopacidae) I II - III

Skuas (Stercorariidae) and Gulls

(Laridae) I II

Terns (Sternidae) 0 – I II

Auks (Alcidae) 0 I

Sandgrouses (Pteroclididae) 0 II

Pigeons, Doves (Columbidae) II II

Cuckoos (Cuculidae) 0 II

Owls (Strigiformes) I – II II - III

Nightjars (Caprimulgidae) and Swifts

(Apodidae) 0 II

Hoopoes (Upudidae) and Kingfishers

(Alcedinidae) I II

Bee-eaters (Meropidae) 0 – I II

Rollers (Coraciidae) and Parrots

(Psittadidae) I II

Woodpeckers (Picidae) I II

Ravens, Crows, Jays (Corvidae) II – III I - II

Medium-sized and small songbirds

(Passeriformes) I II

Additional expert material from this project's partners was then added to the table. Particularly helpful

in this regard was the list of bird victims of collision/electrocution with power lines that was provided

by Vogelbescherming (Annex II). The information was collected from the Recovery Centres by

Vogelbescherming between 1991 and 2010, and the Ligue Royale Belge de Protection des Oiseaux

(1991-2000 period). The LRBPO does not have centralized statistics from their Centres available after

2000. The figures in Annex II only relate to birds that have been found wounded and are checked into

a recovery centre, meaning birds that have been found still alive and brought in for recovery.

Consequently, the overall casualty figure is a lot higher, and impossible to estimate from this data. The

Annex II shows victims from “power lines” in general, but most of these (95%) are brought in with

collision-like wounds, and very few with traces of electrocution. The top 10 species brought in for

collision are common waterbirds (Grey Heron, Black-headed Gull, Herring Gull, Mallard, etc.),

common raptors (Buzzard, Kestrel, Sparrowhawk) and doves (Woodpigeon and Collared Dove). More

surprisingly, the Woodcock is a regularly found victim of collision. This could possibly be explained

by its display behaviour – called “roding” – characterized by the male flying over the canopy or open-

Reducing bird mortality caused by high- and very-high-voltage power lines

areas in the forest at dusk. The sensitivity of the Woodcock to collision was also recently quoted in a

proposed list of priority species for mitigation of power-line impacts (BIO INTELLIGENCE SERVICE,

2012). For the year 2010 and 2011, it was also possible to compare the collision victim numbers with

the overall number of checked-in individuals from each species (Table 3). This confirms the elevated

risk for waterbirds and also for the Woodcock.

In brief, species groups that are considered “sensitive” to collision are:

· Waterbirds (anatidae, shorebirds – including Woodcock – and gulls)

· Grouses, Pheasants and allies

· Great Cormorant and Grey Heron

· A few species of raptors

· Rails & Crane

· Doves

· Owls & allies

· Woodpeckers

· Swift, Kingfisher, Nightjar

· Jackdaw

· Starling

Most passerines are not judged sensitive, and seabirds are insensitive (for their living habitat is devoid

of power lines!). The risk of collision mortality is judged “medium” for most species of raptors, a very

different picture from that if electrocution risks were contemplated (PRINSEN ET AL., 2011a). However,

as pointed out later, we decided to take rare raptor species into consideration, as even a limited

additional mortality can potentially have significant effects on population dynamics.

Table 3: Birds with a high (>2%) proportion of power-line victims, data from 2010 and 2011

(Vogelbescherming VL) – see also Annex 2.

Dutch name Scientific name

Number of

power-line

victims Total number of

victims

Proportion

involving power

lines

Kleine Zwaan Cygnus columbianus 1 1 100.0%

Kolgans Anser albifrons 1 2 50.0%

Tafeleend Aythya ferina 1 2 50.0%

Regenwulp Numenius phaeopus 1 3 33.3%

Roerdomp Botaurus stellaris 2 10 20.0%

Wilde Zwaan Cygnus cygnus 1 8 12.5%

Houtsnip Scolopax rusticola 29 241 12.0%

Rietgans Anser fabalis 1 9 11.1%

Grauwe Gans Anser anser 5 46 10.9%

Grote Mantelmeeuw Larus marinus 2 24 8.3%

Havik Accipiter gentilis 2 26 7.7%

Knobbelzwaan Cygnus olor 9 153 5.9%

Koekoek Cuculus canorus 1 17 5.9%

Sperwer Accipiter nisus 19 412 4.6%

Brandgans Branta Leucopsis 1 22 4.5%

Torenvalk Falco tinnunculus 6 140 4.3%

Boomklever Sitta europaea 1 27 3.7%

Slechtvalk Falco peregrinus 1 27 3.7%

Blauwe Reiger Ardea cinerea 14 394 3.6%

Aalscholver Phalacrocorax carbo 3 90 3.3%

Witte Kwikstaart Motacilla alba 1 33 3.0%

Nijlgans Alopochen aegyptiaca 1 35 2.9%

Kokmeeuw Chroicocephalus ridibundus 10 398 2.5%

Fuut Podiceps cristatus 1 47 2.1%

III.2.2. Conservation relevance criteria (CONS)

Several existing sets of data on the conservation relevance of Belgium bird populations have been

used to objectivize the CONS criteria (see Annex I). These data have been calculated at a regional

level (Flanders or Wallonia), as conservation of nature is a regional competence. Instead of re-

calculating these values at federal level, we decided to consider the “regional relevance” that is

important at national level, as both regions have a responsibility to protect regional populations. If

only one of the two regions hosts a significant population of a certain species, it is logical to consider

this species at national level too, but probably in the subsequent mapping procedure, only the

concerned region will host important sites or areas for the species.

The Red List of endangered species has been published independently in the two regions, using

slightly different methodologies (DEVOS ET AL., 2004; PAQUET & JACOB, 2010). All “red list” species

(excluding the near-endangered or “rare-zeldzaam” categories) in one of the two Red Lists have been

attributed a CONS level of “2”.

In Flanders, several waterbird populations are reaching the threshold of 1% of the biogeographical

population. These species have been given a score of 2 for the CONS criteria. In Wallonia, a few

breeding birds also reached 1% of the European breeding population. If these species are also of

conservation concern (i.e. if they are declining or if they are included in Annex I of the Birds

Directive), their CONS level has been set to “2”.

III.2.3. Group-forming criterion (GROUP)

This criterion was used to highlight species that would be irrelevant in terms of conservation of the

species populations but that could suffer high losses from power-line collisions and hence may benefit

from targeted actions. Logically, sensitive species for GROUP criteria again include waterbirds, and

also thrushes, corvids and some granivorous passerines.

The final classification of species, including their presence type in Belgium, is presented in Annex I.

Reducing bird mortality caused by high- and very-high-voltage power lines

IV. Step 2: Definition of species groups used in the

mapping/prioritization procedure

In the previous chapter, we have classified bird species according to their relevance to the collision

risk problem in the Belgian context. In this second step, we used the classified list presented in

Annex I to define a “coherent group” of species, for the ultimate goal of organizing the mapping

procedure.

We first focus on species that are highly sensitive to collision and highly relevant for conservation

(COLS = 2 and CONS = 2); this accounts for 50 species, mostly waterbirds, some migrants (Crane,

storks, etc.), some nocturnal birds and several woodpeckers. We then included a few supplementary

rare bird species with a lower collision sensitivity but with a high conservation value (i.e. some raptors

such as the Red Kite). The detailed distribution of these birds is generally well known in Belgium.

Finally, we organized this large species set into four coherent species groups, according to their

ecology, their distribution type (i.e. widespread or very localized) and geographical data availability of

the listed species.

· Group 1: Waterbirds. This group incorporates a large number of species linked to special

areas (wetlands). Some of them can also be relatively widespread, so that it is highly probable

that at least some individuals can be seen anywhere on any day (i.e. Black-headed Gull north

of the Meuse valley in the winter). As a rule, these species are numerous and tend to be social,

meaning that they congregate on some discrete sites. Because of this, power-line collision

mortality can potentially affect a large number of individuals, although, for most species, this

mortality is unlikely to strongly impact population dynamics. Information on distributions and

numbers are usually available as waterbirds are traditionally well surveyed by ornithologists.

Common waterbirds can also be seen as an “umbrella” group for other, rarer species linked to

wetland habitats.

· Group 2: Rare breeding birds. Some of the listed sensitive species are restricted to

particular sites or occur in very small numbers. Collision with a power line is by definition a

rare event for these species. Nevertheless, additional mortality may have a major impact on

population dynamics. Rare breeding birds can also act as umbrella species for other birds

linked to “special” habitats. Good data on distribution, sometimes even including nest

location, are also available for most rare breeding bird species.

· Group 3: Migrants in large numbers. These species are potentially affected while they cross

the country during migration, sometimes even in large flocks. However, these flocks generally

use very large corridors. Furthermore, preferential corridors and flight altitude are very much

influenced by local weather conditions. Until now, no real attempt on mapping those corridors

has been made on a Belgian scale. Specific migration counts are undertaken especially during

autumn migration (P

AQUET & DUJARDIN, 2010) but they focus more on describing numbers

and phenology than location of flock corridors.

· Group 4: Widespread species. This group is characterized by widespread distribution

without any pattern of special site occupancy or regular congregation sites. They are territorial

birds either linked to a farmland or woodland habitat. Although a large number of power-line

collisions are possible for these birds, data on the distribution at high resolution are sometimes

not easily available. The risk pattern is, by definition, scattered (meaning low local risk over a

large territory) and so it is difficult to identify the main or priority dangerous lines affecting

these species.

In the following step, we will consider each of the above four groups in a mapping exercise and then

check if the resulting maps can be used in the prioritization process.

V. Step 3: Compiling maps of critical areas for sensitive

species

V.1. Some precautionary remarks

About the regional vs. national criteria

Elia is operating across the whole of Belgium; therefore, it is logical that the most critical bird areas at

country level should be identified. However, bird monitoring and conservation is handled at regional

level in Belgium because it is the responsibility of regional authorities and it is managed by

associations mostly active at regional level. Furthermore, we based a great deal of the present

publication on an existing similar work on sensitivity mapping for birds and wind energy in Flanders

(EVERAERT ET AL., 2011). To maximize coherence between the two approaches (grid and wind energy

sensitivity) and for practical reasons, for Flanders we directly used the critical areas for sensitive

species as defined in Everaert et al. (2011) and for Wallonia we delineated critical areas using similar

approaches, based on data for that part of Belgium (including the Brussels region in some cases). We

think that this is the best way of resolving the dilemma of “nationally” delineating important bird

areas, while taking into account regional nature conservation priorities. This approach may inspire

integrated work at a higher level (multi-country approaches), although additional research and testing

would be necessary.

About the time reference of data used

The mapping exercise presented below is based on the most up-to-date bird distribution data available,

which is from the period 2001 to 2011. This is important to keep in mind because bird ranges are in

constant evolution (most species are constantly expanding or decreasing their range), and also because

knowledge of these distributions as well as the techniques to record them are developing all the time.

More details of the data “time-stamp” are presented here.

The importance of waterbird sites was evaluated using data from the entire reference period (annual

counts for the period 2001-2011). Colony locations and rare bird breeding sites are primarily described

in the regional breeding bird atlases (Flanders: 2000-2002, Brussels: 2001-2004, Wallonia: 2001-

2007) but have been updated using data from long-term monitoring schemes (VERMEERSCH &

ANSELIN, 2009) or from web-based recording portals (waarnemingen.be / observations.be). Migration

corridors have been delineated with knowledge gained in recent years (mostly after 2007) thanks to

dedicated web-based recording tools (i.e. trektellen.nl), as well as from general knowledge from

birdwatchers in Belgium. Plover stop-over sites have been modelled from observation data concerning

the period 2008-2011. Common bird maps were based on regional atlas data (Flanders: 2000-2002,

Brussels: 2001-2004, Wallonia: 2001-2007).

V.2. Waterbirds

Belgium, and Flanders in particular, is home to large numbers of waterbirds (see III.1). “Waterbirds”

is a multi-family group of species sharing a strong ecological link with wetlands: it includes wildfowl,

herons, rails, waders and gulls, but traditionally excludes passerines also linked to water (like the

Dipper Cinclus cinclus). Waterbird species are often regarded as particularly sensitive to power-line

collision, with this sensitivity being enhanced by their social behaviour (most species are highly social

during migration and in winter, and many breed in colonies, i.e. are social during reproduction). Large

groups of waterbirds can easily be disturbed by humans and disturbance often leads to a higher risk of

collision, caused by panic. Colonial and communal roosting behaviour implies regular commuting

trips between roosts/colonies and foraging sites. Note that some waterbirds (e.g. the Common Snipe

Gallinago gallinago) may also sometimes use dispersed habitats (wet meadows and pastures in the

Ardennes).

Waterbirds are also probably the best monitored species group in the world, with mid-winter counts

organised throughout Europe for more than 40 years (GILISSEN ET AL., 2002; DELANY & SCOTT,

Reducing bird mortality caused by high- and very-high-voltage power lines

2006). In Belgium, every winter, mid-monthly counts of waterbirds are organised in the three regions

(with the help of hundreds of volunteers, coordinated by INBO for Flanders, and Aves for Wallonia

and Brussels). Most breeding colonies are also well monitored (gulls, the Great Cormorant, the Grey

Heron), although not annually (VERMEERSCH ET AL., 2006; JACOB ET AL., 2010). For some species (the

Great Cormorant and gulls), results of counts at roost sites in winter are available (DEROUAUX ET AL.,

2010; DEVOS, 2011).

The critical areas for waterbirds have been defined using information from these three data sources:

mid-winter counts (aiding identification of important wintering waterbird sites), colony censuses and

roost counts.

Wintering waterbird sites

In the wind energy sensitivity map for Flanders, critical areas for waterbirds were identified using a set

of numerical criteria applied to the mid-winter censuses (EVERAERT ET AL., 2011). For consistency and

practical reasons, we decided to stick with the original delineation of critical areas for waterbirds in

Flanders. For the rest of the country, we applied similar criteria to identify critical waterbird areas,

although some minor adaptations were necessary (for example, only mid-January regional counts exist

for Wallonia and Brussels, as opposed to the six-monthly counts of waterbirds in Flanders).

Mid-January census data (also called “International Waterfowl counts – IWC”) from 2001 to 2010

were considered in the analysis. Two threshold criteria (2 and 15% of the regional population) were

calculated for each species by estimating the total wintering population of the considered species as

the average total count in mid-January during the decade from 2001 to 2010. For Flanders, instead of

the mid-January counts, winter maximums were used. Table 4 shows how the criteria were applied.

Table 4: Criteria applied to define the importance of waterbird sites in the mapping process. In this table,

“Regularly” means at least 50% of the considered counts. Note that exotic species (Canada Goose) and gulls

were not considered here.

Importance for the waterbirds

Criteria

Fairly Important Site

Regularly 100

–

1000 waterbirds

Important Site

Regularly mor

e than 1000 waterbirds

Regularly at least 2% of the regional wintering population of

at least one species

Very Important Site

Regularly at least 15% of the regional wintering population of

at least one species

The resulting map of critical areas for waterbirds in Belgium is shown in Figure 2. As explained

above, numerical criteria were applied on a regional basis for estimating the threshold. Nevertheless, a

logical image is emerging from the analysis, with higher numbers of important waterbird sites in

Flanders, especially in the polders and in the Province of Antwerp. Because of the higher numbers of

waterbirds in Flanders, the number of “very important” and “important” sites is higher than in

Wallonia and the Brussels region.

Brussels was considered together with Wallonia (as it is for the organization of the winter census); this

approach was necessary because if Brussels had been considered on its own, many more sites for very

small waterbirds would have been selected under the criteria of 2% of the regional population,

although Brussels typically has less than 2% of all the waterbirds wintering in Belgium.

We think that taking this “regional approach” for the criteria threshold, but a common, national, set of

criteria is a valid compromise. This prevents regionally important sites being overlooked and, at the

same time, suggests a clear priority for action for nationally important sites (in the case of waterbirds,

most of which are located in Flanders).

Figure 2 : Critical areas for wintering waterbirds in Belgium (in violet “Very Important Sites”, in red

“Important sites” and in orange “Fairly Important Sites”).

Colony locations

In the wind energy sensitivity map for Flanders, important waterbird colonies were identified using a

set of numerical criteria applied to nest counts in these colonies (EVERAERT ET AL., 2011). We applied

the same criteria to data from Wallonia and Brussels to obtain a Belgian map of important waterbird

colonies. Colony data from INBO (Flanders) or from Aves (Wallonia and Brussels) were selected

from the period 2000 to 2010. Note that for some colonies, only one count is available for the period.

Species taken into account were the Grey Heron, the Little Egret, the Eurasian Spoonbill, the White

Stork, the Great Cormorant, the Black-headed, Mediterranean, Herring and Lesser Black-backed

Gulls, and the Common, Little and Sandwich Terns. Colonies were selected according to the following

criteria: “Important Colony Sites” are sites where 10 to 100 breeding pairs are regularly (i.e. at least

50% of the available counts) counted, and “Very Important Colony Sites” are sites where more than

100 breeding pairs are regularly recorded, or at least 2% (> 10 breeding pairs) of the regional breeding

population.

Figure 3 shows the locations of the identified important colony sites. Again, most sites are located in

Flanders or in Northern Wallonia (lowland Belgium).

Reducing bird mortality caused by high- and very-high-voltage power lines

Figure 3: Important breeding colonies of waterbirds in Belgium (in red, “very important” colonies, in orange

“important” colonies). Please note that the polygon outlines are larger than scale to make them easier to see.

Roost sites

Outside the breeding season, many bird species have the habit of congregating in large numbers (up to

millions of individuals for some species) at specific, traditional places to spend the night; those places

are named “communal night roosts”. Several ethological explanations exist for this spectacular

behaviour: safety against predator attacks, temperature regulation and information exchange

facilitation. As regards power-line collision risks, those roosting places are significant because they

correspond to daily movements of flocks up and down the daily feeding grounds. Moreover, collision

risks may also increase for roosting birds (more than for colonial breeding birds) because they are

migrants that are not necessarily familiar with the local environment (PAQUET ET AL., 2003).

Data used for assessing roost location importance come from:

· organized coordinated counts of the Great Cormorant and gulls (INBO and Aves data);

· other counts (e.g. roost monitoring of the Goosander in the Ardennes and the Great White

Egret in some areas).

Again, for Flanders, important roosting places were identified in the wind energy sensitivity mapping

exercise, and the same set of criteria was used in Wallonia. Table 5 details the criteria and thresholds

used to define the relative importance of roosting places.

Table 5: Criteria applied to define the importance of waterbird roosting sites at the regional level. For example,

the threshold of 2% was reached when a roost of more than five Great White Egrets or 80 Great Cormorants

was counted.

Importance for waterbirds

Criteria / Threshold

Fairly

mportant

ite

Less than 100 individual

regularly counted

Important

ite

Between 100 and 1000 individuals regu

larly

counted

Very Impo

rtant Site

More than 1000 individuals or at least 2% of the

regional population regularly counted.

The resulting map is shown in Figure 4. Communal night roosts were located and identified for the

following waterbird species: gulls (all species of Larus sp.), the Great Cormorant, geese (several

species of Anser sp.), the Eurasian Curlew, the Great White Egret and the Goosander.

Figure 4: Important night-roosts of waterbirds in Belgium (in red, the “very important” roost locations, in

orange the “important” roost locations). Please note that the polygon outlines are larger than scale to make

them easier to see. “Fairly important roosts” are not presented on this map.

Local movements of waterbirds

In the Flemish wind energy risk map for birds (EVERAERT ET AL., 2011), known local movements of

some wintering waterbird species groups are also mapped. Originally, there were two maps. The first

map shows daily local migration – mainly of wintering ducks, geese and waders – from resting or

roosting areas to foraging areas and back. Many waterbird species like ducks rest in the daytime on

open water and wetlands, and during dusk they fly to their foraging areas such as grasslands,

agricultural fields and other wetlands. Additionally, waterbirds can also be forced to adopt a tidal

rhythm for foraging. The second map shows daily local migration – mainly of wintering gulls, geese,

waders (mostly curlew) and cormorants – from foraging areas to roosting areas. These movements

take place around sunset.

Both maps were created using the available information, based on censuses and current knowledge of

local birdwatchers. The migration routes are displayed as areas (migration corridors) where most birds

can be expected to fly. In many cases the width of the corridor varies between 1000 m and 2000 m,

although in certain cases a different width was used because of the combination with other nearby

corridors, spatial properties of the surroundings (e.g. along canals), distance between feeding and

roosting areas, etc.

The data are of course incomplete. Most of the information does not come from censuses with set

criteria and it is possible that some data do not reflect the actual situation.

Table 6: Criteria applied to define the importance of local migration corridors of wintering ducks, geese and

waders towards their foraging areas.

Importance of local migration corridor

Criteria / Threshold

Important: to foraging place and back

Betwee

n 100 and 2000 individuals per 24

Important: to roosting place

Between 100 and 2000 individuals per evening

Very important: to foraging place and back

More than 2000 crossing individuals per 24

Very important: to roosting place

More than 2000 cross

ing individuals per evening

Reducing bird mortality caused by high- and very-high-voltage power lines

V.3. Rare breeding birds

The list of species identified as sensitive to collision with power lines includes several rare breeding

bird species, of which very small populations are often concentrated at well-known sites. For some

species, nest locations are even monitored from year to year. Most of these species present a particular

legal status (Annex I of the Birds Directive) or are cited in the regional Red Lists of Endangered

Species (see Table 7). Identifying dangerous power lines is particularly important for these species, as

their populations are generally under high pressure from other factors already. However, a few of

these species (Black Stork and Eagle Owl) recently increased their populations and are now quite

widespread in Wallonia. It is nevertheless important to minimize collision risks as they are still

vulnerable to additional mortality.

Table 7: Legal and biological status of the rare breeding species that are sensitive to collision with power lines

in Belgium (V

ERMEERSCH

NSELIN

, 2009; J

ACOB ET AL

., 2010). Species with an asterisk are considered of

intermediate sensitivity to collision in Annex I but are considered in the rare breeding bird map because of their

high conservation value and potential sensitivity to supplementary mortality.

Species

Annex I Red List (VL) Red List (WAL) Population

estimate (VL) Population

estimate

(WAL)

Hazel Grouse

Bonasa bonasia

X - Critically

Endangered - 48-140 pairs

Black Grouse

Tetrao tetrix

X Extinct Critically

Endangered

0 13-27 hens

Great Bittern*

Botaurus

stellaris X Critically

Endangered Critically

Endangered 20 pairs 0-1 pair

Little Bittern*

Ixobrychus

minutus X Critically

Endangered Critically

Endangered 21 pairs 1-4 pairs

Black Stork

Ciconia nigra

X - Vulnerable - 75-95 pairs

White Stork

Ciconia ciconia

X - - 55-60 pairs 0-1 pair

Red Kite*

Milvus milvus

X Irregular Vulnerable - 150-180 pairs

Marsh Harrier*

Circus

aeruginosus X Least Concern Endangered 90 pairs 1-7 pairs

Hen Harrier*

Circus cyaneus

X Irregular Endangered 0-1 pairs 1-5 pairs

Montagu’s Harrier*

Circus pygargus

X Critically

Endangered Endangered 0-1 pairs 2-13 pairs

Peregrine Falcon*

Falco

peregrinus X - Vulnerable 27 pairs >25 pairs

Spotted Crake

Porzana

porzana X Endangered - 5-15 pairs irregular

Corn Crake

Crex crex

X Critically

Endangered Critically

Endangered 5-6 males 1-12 males

Kentish Plover

Charadrius

alexandrinus X Critically

Endangered - 23-31 pairs -

Pied Avocet

ecurvirostra

avocetta X Vulnerable Vulnerable 410-425 pairs 1-3 pairs

Eurasian Eagle Owl

Bubo bubo

X - Vulnerable - 80-85 pairs

Tengmalm's Owl

Aegolius

funereus X - Vulnerable - 21-100 pairs

European Nightjar

Caprimulgus

europaeus X Vulnerable

Endangered 500-550 pairs 50-60 pairs

Eurasian Wryneck

Jynx torquilla

Critically

Endangered Endangered Irregular

breeding 45-58 pairs

Grey-headed

Woodpecker

Picus canus

X - Endangered - 33-40 pairs

The critical areas for these species have been mapped using two different approaches.

· The site-based approach is similar to the waterbird sites approach developed earlier. The

mapping consists of identifying concrete sites where the breeding population, or the core

breeding population, is located.

· For a few rare breeding species at risk of collision, the availability of detailed information

allowed for a more accurate definition of the “risk-area” than the site-based approach: accurate

breeding territory locations and estimation of flight-range from the nest during the breeding

season. Collision-sensitive species for which this information is available are the following:

the White Stork, the Black Stork, the Eagle Owl (only present in Wallonia) and the Nightjar

(in Wallonia, where it is rarer than in Flanders).

Below, we describe how these two approaches have been applied to map critical areas for these

species.

Part 1: Site-based species

For Flanders, we used the “Bijzondere Broedvogelgebieden” thematic map developed for the wind

energy sensitivity mapping (EVERAERT ET AL., 2011). This map has been developed according to data

gathered for the “Bijzondere Broedvogels Vlaanderen” scheme developed by INBO and other data

collected up to 2009. Species that were already taken into consideration in other thematic maps

(colonial waterbirds) were excluded here. Only sites containing more than 2% of the estimated

Flemish breeding population were retained. The species used in this thematic map are the Great

Bittern (2% = 1 pair), the Little Bittern (2% = 1 pair), the Marsh Harrier (2% = 1 pair), the Hen

Harrier (2% = 1 pair), the Montagu’s Harrier (2% =1 pair), the Spotted Crake (2% = 1 pair), the Pied

Avocet (2% = 8 pairs), the Kentish Plover (2% = 1 pair), and the European Nightjar (2% = 10 pairs).

For Wallonia, we used the following data:

· detailed data of specific bird inventories in the Natura 2000 sites, conducted since

2006 under an agreement with the Walloon Region - DEMNA (SPW);

· detailed locations of the known breeding pairs of some rare species, as collected

during the compilation of the Breeding Bird Atlas of Wallonia (during the period

2001-2007);

· detailed observation data of rare species, pooled by Centrale Ornithologique Aves;

these data are mostly collected by amateur birdwatchers through observation data

portals and validated by Aves;

· for some declining species (e.g. the Hazel Grouse), we used the distribution map

produced for the 2001-2007 Atlas to delineate the present-day range. Natura 2000

sites that were designated for a species but are now outside the species range are

excluded here.

Table 8 summarizes how the species sites were selected.

Table 8: Site selection criteria used for the rare breeders in Wallonia.

Species Site selection method

Hazel Grouse (

Bonasa bonasia

) Natura 2000 sites designated for the species, excluding the sites falling outside

the 2001-2007 range + Natura 2000 sites where recent sightings are recorded

Great Bittern (

Botaurus stellaris

) Natura 2000 sites where breeding was recorded after 2001

Little Bittern (

Ixobrychus minutus

) Natura 2000 sites where breeding was recorded after 2001

Marsh Harrier (

Circus aeruginosus

) Natura 2000 sites where breeding was recorded after 2001

Hen Harrier (

Circus cyaneus

) Hand-drawn perimeters where breeding is regularly recorded; those perimeters

are used to conduct Agri-Environment Schemes favourable to this species

Montagu’s Harrier (

Circus pygargus

) Hand-drawn perimeters where breeding is regularly recorded; those perimeters

are used to conduct Agri-Environment Schemes favourable to this species

Tengmalm’s Owl (

Aegolius funereus

) Natura 2000 sites designated for the species + Natura 2000 sites containing at

least one breeding territory during the compilation of the 2001-2007 Atlas,

except where its presence is known to be irregular (Western Ardennes)

Eurasian Wryneck (

Jynx torquilla

) Natura 2000 sites where breeding has been recorded since 2001 (excluding

isolated or unconfirmed cases)

Grey-headed Woodpecker (

Picus canus

) Natura 2000 sites where observations suggest breeding after 2001 (excluding

isolated or unconfirmed cases)

Black Grouse (

Tetrao tetrix

) Natura 2000 sites where the remaining population is still observed

Corncrake (

Crex crex

) Hand-drawn perimeters where the species is regularly recorded, limited to the

Fagne-Famenne region, the stronghold of the species in Belgium

Reducing bird mortality caused by high- and very-high-voltage power lines

Part 2: Buffer-based species

The Black Stork Ciconia nigra breeds in Wallonia and is only observed as a migrant in the two other

regions. Available information includes all known breeding locations and territory centroids collected

during the compilation of the 2001-2007 Atlas (sources: SOLON asbl – SPW – Aves) and a satellite-

tracking study in France indicating that breeding birds are using a range extending easily up to 20 km

from the nest (VILLARUBIAS, 2003; JIGUET & VILLARUBIAS, 2004). If a 20-km radius is drawn around

all known nests, the entire region south of the Sambre and Meuse valleys is included in this buffer

zone. This does not adequately represent variation in the actual presence of the species, which is

clearly more at risk around the main forested zones than in the agricultural landscape of the Condroz.

Therefore, we decided to arbitrarily concentrate the risk-area on a 5-km radius around all known nests.

An analysis with casual, occurrence data of the Black Stork shows that this 5-km buffer covers most

(>50%) of all the Black-Stork occurrence data in Wallonia (from the Aves occurrence database).

The European Nightjar Caprimulgus europaeus breeds in both Flanders and Wallonia. For Flanders,

the sites provided for the wind energy sensitivity map were used (EVERAERT ET AL., 2011). For

Wallonia, we selected the breeding territories recorded after 2001, excluding isolated breeding cases

(source: Centrale Ornithologique Aves and the 2001-2007 Atlas). This targets the four remaining

major breeding areas of the species: Lagland military camp, the Croix-Scaille forest, the Cul-des-Sarts

area and the Spa-La Reid area. A 1500-m radius buffer was drawn around these territories. Although

radio telemetry has shown that a foraging distance of up to 3 km can be recorded (ALEXANDER &

CRESSWELL, 1990), we chose to focus the risk area on a smaller radius, closer to the usual territory

size (1500 m).

The Eagle Owl Bubo bubo is only a regular breeder in Wallonia, and the species is well monitored.

We use the known breeding location (Atlas data, IRSNB data, COA) and draw a 1500 m buffer around

all these locations.

The same 1500-m buffer was used around all the known nest location of the Peregrine Falco

peregrinus from the Flemish wind energy map (EVERAERT ET AL., 2011) or from the Aves database for

Wallonia and Brussels.

Critical areas for the Red Kite Milvus milvus were defined by Aves in an unpublished work concerning

a wind energy sensitivity map for Wallonia, based on spatial modelling of the probability of the

presence of nest locations.

The resulting thematic map

The general thematic map for rare breeding birds is shown in Figure 5. Most of the critical areas for

this theme are located south of the Sambre and the Meuse valley. Most of the critical areas contain

only one rare species, but highly critical sites are again the Eastern polders and some high plateaus of

the Ardennes (Hautes-Fagnes, Saint-Hubert, Croix-Scaille) as well as the Anlier forest.

Figure 5: Sites with sensitive rare breeding species, with an indication of the number of rare breeding species.

V.4. Migrants in large numbers

Mapping specific corridors for seasonal bird migration is especially difficult in a low-lying country.

While in mountainous areas or regions with special coastal geography, clear migrant funnels can be

observed, Belgium lacks such geographical bottlenecks. As a result, millions of migrant birds cross the

country over a broad front every year. This creates a dilution of the collision risk over the whole

country, even where minor concentration effects and no uniform risk exists for certain species. In the

recent windmill risk atlas for the Netherlands (AARTS & BRUINZEEL, 2009) and Flanders (EVERAERT

ET AL., 2011), migration corridors have been tentatively defined, mainly on the basis of expert

judgements, and radar data and visual counts of active migrants by amateur birdwatchers (see

www.trektellen.org).

Air Force and meteorological radars allow the detection of migration flocks. This “by-product” of

radar surveillance has been widely used to study different aspects of bird migration (BERTHOLD,

2001). In Belgium, many observations are recorded by military radars but this information remains to

be analysed in a systematic manner to define migration corridors (S. Sorbi – Belgian Air Force, pers.

com.). However, it is likely that this analysis will reveal varying migration intensity spread across the

country, like in the Netherlands (AARTS & BRUINZEEL, 2009), rather than selective, narrow corridors.

Hundreds of birdwatchers are counting migrating or passing birds from well-defined posts every year,

especially in the autumn (PAQUET & DUJARDIN, 2010). Most of these counts are collected on the

dedicated www.trektellen.org website. While these data prove to be very useful in terms of describing

migration intensity and phenology, on a geographical basis they are supposed to reflect much more

active birdwatcher locations rather than migration funnels (although on the basis of experience,

birdwatchers probably select strategic counting posts with high potentials for condensed bird

migration). Using data from trektellen.org together with unstandardized records of migrating birds

from the Aves bird observation database, we tentatively plotted the migration map for two collision-

sensitive species in 2009 and 2010: the Woodpigeon and the Common Crane (Figure 6).

Reducing bird mortality caused by high- and very-high-voltage power lines

Figure 6: Map showing locations of recorded active migrants of two sensitive species – in green, the Common

Crane, in red the Woodpigeon (source: Aves database, observations.be, years 2009 and 2010). Both species can

migrate over the whole country, but large groups are more likely to be observed in Eastern Belgium for the

Crane and over the Condroz and Famenne ecoregion for the Woodpigeon, especially around the Meuse valley

and the Northern ridge of the Ardennes region.

This map, together with expert knowledge, suggests three main migration corridors for Wallonia, as

well as the ones already defined for Flanders (EVERAERT ET AL., 2011): one following the Meuse

valley (Woodpigeon corridor 1), another one following the Northern ridge of the Ardennes

(Calestienne – Woodpigeon corridor 2) and a third one, which is broader to take into account the

major Crane corridor. In Flanders, migration corridors generally focus more on migrant waterbirds, the

Eastern corridor also reflecting the greater passage of Common Crane and Woodpigeon (EVERAERT ET

AL., 2011). The final migration map for Belgium is presented in Figure 7.

It is important to stress the fact that these corridors are defined at macro-scale and that meso- and

micro-scales (i.e. the main local migration axis through a given landscape) are not shown here.

Due to the diluted risk for migrants in terms of collision with power lines (diluted in time as migration

only concerns a narrow time window and diluted in space as these corridors are over-simplifying a

broad front migration pattern), we think that these migration corridors can only be considered a

secondary criterion in defining the most dangerous power lines.

Figure 7: Proposed migration corridor map for Belgium. In violet, the major corridor of the migrants following

the coastline and in pink, major inland migration corridors defined by Everaert et al. (2011) and this

publication.

V.5. Local movement of birds

Some species of birds, especially waterbirds, commute every day at dawn between feeding areas and

roosting places. These movements often concerned fast-moving large groups of birds and are

especially risky because power lines are less visible at this moment of the day. These regularly-used

daily corridors have been mapped in Flanders (EVERAERT ET AL., 2011) and used in this study (Figure

Figure 8 : Proposed daily corridor map for Belgium. In light blue, moderately used corridors; in dark blue:

intensively used corridors, as defined by Everaert et al. (2011).

Reducing bird mortality caused by high- and very-high-voltage power lines

V.6. Widespread breeding species

Among the targeted species for the mapping, the fourth group (widespread species) consists of species

not linked to well-defined sites or areas. Their population is rather widespread in Belgium, present in

variable densities in a range of habitats. This is typically the case for some forest birds (woodpeckers)

or farmland birds (Grey Partridge). For these species, which are sensitive to collision, the site-based

approach for the sensitivity mapping (as used for waterbirds) is not useful. It is often not possible to

identify concrete sites where it would be critical for the species to address the power-line issue.

Therefore, for widespread species, we explored the use of relative density maps as a better approach to

sensitivity mapping. These high-resolution maps will not only delineate the areas where the species is

present or absent, they will also classify the relative density of the species over the whole country.

From these maps, we will be able to select the “high-density” sectors, and consequently sensitive areas

for power lines. High-resolution density mapping is generally possible through distribution modelling,

using known observations of a species and statistical relationships between these data and

environmental descriptors to predict the distribution (or abundance) of the species over a whole study

area (FRANKLIN, 2009). It is necessary to use distribution modelling rather than “true observation

data” or “real counted birds” as it is generally not possible to have an indication of the abundance of

every species at a very fine-grained resolution (like 1-km square) for the entire country. Distribution

modelling has already been used in sensitivity mapping for windmills (PEARCE-HIGGINS ET AL., 2008)

and also recently applied in a multi-specific approach with bird atlas data to aid nationwide spatial

planning (HERRANDO ET AL., 2010).

Model-building procedure

A complete explanation of how we developed high-resolution relative density maps is available in the

methodological chapter of the Breeding Bird Atlas of Wallonia (PAQUET ET AL., 2010). In brief, we

used the 1x1-km bird sampling from each of the three regional Atlases. This is a standardized protocol

to sample birds present in squares of 1 km² (two one-hour visits during one reproductive period). The

detailed methods vary slightly between the three regional atlases but as these variations mainly

concern bird quantification, we simplified the data by only using presence/absence data. The grid used

(although composed of 1x1-km squares) differs between the atlases (in Flanders, the grid has 1-km

UTM squares while in Wallonia, the grid is based on a National Geographic Institute map). We

circumvented this problem by using the appropriate grid for each region at the same time for the data,

the calculated variable and the model projection. The model is thus calculated for the whole area but

each variable is calculated in the correct grid. Finally, one important difference between the three

sampling schemes lies in the coverage of the sampling; this coverage (meaning the proportion of

sampled squares related to the total surface) is about 17% for Wallonia, 100% for Brussels (every

square of this small region was sampled) and 30% for Flanders. These coverage rates are well above

the minimal sampling effort necessary to obtain a high-quality spatial distribution model, which is

estimated at around 4-5% (AIZPURUA ET AL., 2012). In order to avoid a bias due to over-sampling in

Flanders and Brussels, we randomly selected a number of sampled squares to make them equate with

the coverage rate of Wallonia (17%).

To model the distribution of the species considered, we used “presence” data provided by the sampling

procedure on the one hand, and 20 variables concerning the environment in every square on the other

hand. These variables describe the land-use (deduced from the CORINE land cover map, version

2006, published by the European Topic Centre on Land Use and Spatial Information) and bioclimatic

variables calculated from the WordClim dataset (HIJMANS ET AL., 2005). The modelling method is

MaxEnt, a presence-only method that is now widely used in distribution work (PHILLIPS ET AL., 2006)

and that was successfully used in the Breeding Bird Atlas of Wallonia (PAQUET ET AL., 2010). The

projected result of the modelling procedure is a map estimating the probability of occurrence (ranging

from 0 to 1) for every 1x1 km for that species. The probability of occurrence can be related to the

“habitat suitability” of the square for the species. A bootstrap procedure, leaving out 30% of all the

presence data to train the model, is used to validate the model. The final model is the average of

10 models.

When producing the map, a colour gradient is applied to symbolize the increasing probability of

occurrence (or “habitat suitability”), with white representing a very high likelihood of absence of the

species. This visually important threshold was determined for each species with a different cut-off

analysis proposed by MaxEnt, and a “Balance training omission, predicted area and threshold value”

was chosen.

Modelling results

Among the species identified as “sensitive to power-line collision”, given the available data and the

prevalence of the species (the proportion of sampled data where the species were detected), the

following breeding birds can be mapped using MaxEnt habitat suitability modelling: the Grey

Partridge, the Northern Lapwing, the European Turtle Dove, the Green Woodpecker, the Black

Woodpecker and the Middle Spotted Woodpecker. A proposed high-resolution distribution model is

proposed for these species in Figure 9. Note that this is, to our knowledge, the first attempt at drawing

a Belgium-wide high-resolution habitat suitability map for breeding birds. The distribution of the Barn

Owl could not be modelled by this method because the available sampling data concerns only diurnal

birds.

Comparing the resulting maps side by side in Figure 9 suggests that it will be difficult to use them to

prioritize hazardous power lines for these sensitive species as the high-density areas for these species

are quite extended, but above all because the “best areas” for these six species are not congruent. Even

for the three woodpecker species, we see that the southern broadleaved forests are best for the Middle-

Spotted Woodpecker, while the Black Woodpecker has its higher densities in coniferous woodland in

the Ardennes and in the Campine, and the Green Woodpecker prefers mixed farmland areas, outside

the dense woodlands. Therefore, adding the known distribution of these species to the criteria used to

prioritize power lines may result in a dilution of the priority settings. Moreover, as these birds are

territorial, the risk of collision mortality is further diluted, even in territories crossed by power lines.

We therefore suggest using these distributions only to refine a prioritization exercise mainly based on

waterbird congregation criteria and rare-bird areas.

Reducing bird mortality caused by high- and very-high-voltage power lines

Grey Partridge Northern Lapwing

European Turtle Dove Middle-spotted Woodpecker

Black Woodpecker Green Woodpecker

Figure 9: Habitat suitability maps for six rather widespread breeding species in Belgium (1-km resolution). The

maps show the variation of probability of presence calculated by the MaxEnt model using sampling data from

the three regional atlases. White indicates an area where the species is most probably absent. A high probability

of presence is linked to habitat suitability.

V.7. Other important stop-over areas for plovers

Two Birds Directive Annex I species of “waterbirds”, the Eurasian Dotterel Charadrius morinellus

and the Golden Plover Pluvialis apricaria are both recognized as “sensitive to collision” and “of

conservation relevance” (see Annex I), although their main staging areas are not well covered by the

critical area delineation for Group 1 (waterbirds). Their habitat is more “terrestrial” than other

shorebirds. Dotterels are usually seen in very open farmland habitats during migration. Formerly

thought to be a very rare bird in Belgium, it recently appeared to be a regular staging post-nuptial

migrant in some farmland areas in central Belgium, sometimes in relatively large groups (ROUSSEAU-

PIOT, 1995; MARIAGE & FARINELLE, 2005). This new status was linked to better knowledge and a

systematic survey of this charismatic bird by birdwatchers. The Golden Plover is an abundant

migrating and wintering bird in meadows and open land, mostly in the western part of the country and

sometimes also on interior farmland areas. The use of intensive agricultural habitats is not anecdotal

for this species, as open fields can be a very profitable staging habitat (LINDSTRÖM ET AL., 2010).

In order to identify the most critical areas for these species, we used a similar spatial modelling

approach as for the widespread breeding species (see page 31). Instead of atlas sampling, we used

observation data of both species, extracted from the www.waarnemingen/observations.be database.

From this dataset, we selected the locations of resting groups of more than 100 Golden Plovers or

resting Dotterels during the period 2008-2011 (records of flying birds were excluded to focus on

staging areas). For each species, only one observation per year inside a given 1x1-km square was

retained (this was to avoid a long-staying group of Dotterels, recorded sometimes by dozens of

different birders, creating a bias in the modelling procedure). From this data, we proceeded as

explained for breeding birds in Chapter V.5. The result is a map showing the probability of occurrence

(from 0 to 1) of staging Golden Plovers (Figure 10) or Dotterels (Figure 11).

Figure 10: Distribution map (by probability of occurrence) of a large group (>100 individuals) of staging

Golden Plovers Pluvialis apricaria in Belgium (2008-2011). Large groups of the Golden Plover are rare outside

the Polders.

Reducing bird mortality caused by high- and very-high-voltage power lines

Figure 11: Distribution map (by probability of occurrence) of staging Dotterels Charadrius morinellus in

Belgium (2008-2001). Very open, steppe-like agricultural areas are selected as stop-over sites by this species.

These maps will be used in the scoring procedure as a secondary criterion, as for the widespread

breeding birds. The cut-off of occurrence probability above which a square is declared to be part of the

“areas” of the species is, as for the breeding birds, the cut-off calculated as the “Balance training

omission, predicted area and threshold value” proposed by MaxEnt.

VI. Step 4: Using the maps to prioritize power lines

In sensitivity mapping, the goal is to constitute a map of added constraints and identify high- to low-

risk areas at the best possible resolution. This is a powerful planning tool when a decision has to be

made at a regional scale as to where to build potentially harmful devices like windmills (BIRDLIFE

EUROPE, 2011). This can also be used when considering grid development, but in the present

publication the goal was slightly different: it was to identify and rank, within the existing high-voltage

power grid, the line segments that present a high risk of causing additional mortality to birds, in order

to take mitigating actions, like installing deterring devices (see page 39) in the most suitable order of

priority. Therefore, we had to combine the different maps detailed in Chapter 5 in a way allowing us to

“score” the existing line pylons according to their relative collision threat.

To do this, we combined the different maps in a single “polygon” assembly (i.e. a large shapefile of

polygonal units). Polygons can be derived from the 1x1-km grid used to model widespread species, or

the waterbird areas including various buffer drawn around the site itself, or the migration corridors,

and the combination of all these polygonal units. The resulting shapefile contains more than

129,000 spatial units, each of them thus containing specific information about the presence of sensitive

bird species (waterbird colony, number of rare bird species, migration corridors, etc.).

The next step was to create a scoring system, mostly based on expert judgement, to characterize all

these spatial units in relation to the collision risk. Table 9 shows the applied scoring system. As

explained earlier, we hypothesized that the most detrimental power-line effects would be close to

critical waterbird areas, especially roost sites and colonies, as they involve regular movements of large

numbers of birds entering and leaving these areas. We also hypothesize that focusing on mitigation

efforts for lines crossing sensitive rare-bird areas would be relevant, as it makes sense in terms of

focusing on conservation efforts: regional authorities as well as nature conservation organizations are

often already investing in these areas to protect a target species. Other sensitive species, like

widespread breeding species and migrating birds in certain corridors, are also present around some

power lines but because power lines probably represent a “diluted” risk for these species, we advocate

handling these factors only as a secondary-priority criterion.

Table 9: Priority scoring system for the spatial units in the final map.

Distance

from

the

ritical

area

Critical area

considered

Inside

Less than 1 km

Between 1 and

3 km

Between 3 and

5 km

Over 5 km

Waterbird roost

If very

important =25;

important=20

Waterbird colonies

If very

important =25;

important=20

Important

waterbird site

If very

important =30;

important=25;

fairly

important=20

Daily corridor

4 if important, 6 if very important

Rare

ird Area

10 points for area with 1 rare species, 20 for area with 2

3 rare species

25 for area with

4-5 rare species and 30 for area with more than 5 species

Migration corridor

points if power line pylon is inside

, 12 for the coastal corridor

Plover staging a

rea

5 point

for each of the 2 species

, when presence cut

off is reached *

Widespread

breeding bird

point

by species

, when presence cut

off of the species is reached

* For scoring purposes, the cut-off value of the probability index that was used to determine the “presence” of a species is

the cut-off calculated by MaxEnt where the test omission rate is around 20%, meaning that the cut-off is slightly more

stringent than the one used in Figure 9.

Reducing bird mortality caused by high- and very-high-voltage power lines

The scoring system applied to our final spatial map allowed us to draw a kind of “collision risk

landscape” for Belgium, presented in Figure 12.

Figure 12: The final collision risk map for Belgium, showing on a gradient from green (low-risk areas) to red

(very critical areas for bird collision risks) the status of bird collision risk with high-voltage power lines.

The map in Figure 12 clearly emphasizes regions of high importance for waterbirds like the polders

and the Port of Antwerp, or in Wallonia the Haine valley marshes or the Meuse river valley. Of

course, the scoring system is based on expert judgements. However, we tried different combinations of

the scoring system (like increasing the relative score for rare birds, etc.), and the risk map turned out to

be rather robust.

Finally, for each of the 22,780 pylons in the Elia grid we calculated a risk score simply according to

the spatial unit score where the pylon is standing. It is important to note here that we used the pylon

location for practical GIS reasons, but of course the dangerous part of the grid, in terms of collision, is

the line between the pylons, not the pylons themselves. So once a pylon has been identified as

dangerous according to our scoring system, work should focus on the lines attached to the pylon.

This step allows us to make a classification of the entire Elia grid according to bird risks as illustrated

in Table 10 and Figure 13. About 3.4% of the pylons are considered as “high priority” (score higher

than 50) and they are located mostly around important waterbird areas like coastal polders and the

lower Meuse valley. The highest priority should also be given to the longest segment in critical areas,

so there is no doubt that the initial efforts should focus on the Polders and Yser valley areas, then on

the area around the Port of Antwerp and more sporadically the areas around some river basins in

Wallonia (Meuse, Ourthe, Haine).

Table 10: Frequency distribution of Elia pylons according to their dangerousness for birds.

Priority score class

Number of

pylons

% of the

grid

0 to 10 3015 13.2%

11 to 20 6608 29.0%

21 to 30 7066 31.0%

31 to 40 3751 16.5%

41 to 50 1567 6.9%

51 to 60 494 2.2%

more than 60 265 1.2 %

Figure 13: Map of the current Elia network of power lines. Pylons are coloured according to their priority

scores. Most of the high-priority lines are close to waterbird sites, with a higher risk clearly existing in the

Polders area and the Yser valley, and also around the Harchies marshes and the lower Meuse valley.

It should be emphasized that this priority score probably needs to be regularly updated. Knowledge of

bird distribution is developing all the time (for example, in future years we may be able to obtain

detailed distribution maps for breeding Woodcocks, which we are unable to draw now). Bird ranges

themselves are also always changing (for example, the Red Kite and the Peregrine are expanding their

breeding range). The priority classification of “dangerous” power lines presented here should thus be

seen as an answer to the following question: “Given present-day knowledge of bird distribution and

relative sensitivity to collision mortality, what are the most dangerous power-line sections, i.e. those

where the mitigation of collision risk should be focused as a priority?” The next chapter looks at

mitigation methods.

Reducing bird mortality caused by high- and very-high-voltage power lines

VII. Mitigation techniques to reduce collisions between

birds and power lines

This last chapter intends to review the possible technical methods for mitigating the risk once a

harmful power line has been identified (according to chapter IV). Addressing all the technical issues

related to the application of the described mitigation techniques (like cost of stopping power-line

operation in order to install diverting devices, and cost of the risk to employees of setting up devices)

is beyond our competence in the domain. Consequently, although a tentative cost/efficiency

comparison is included, we focus here on comparing the efficiency of existing techniques according to

the literature.

We also focus below only on existing power lines. Of course, when planning a new power line, the top

priority is to avoid, wherever possible, placing it in or near a critical bird area. The approach

developed in the previous chapter of this publication could also be used for grid development planning

(as a “sensitivity mapping” service (B

IO INTELLIGENCE SERVICE, 2012)), but here we will not expand

on the measures to be taken to decrease the collision risk at the time of planning a new line (see

PRINSEN ET AL., 2011a for more details).

A complete review of applied mitigation measures in Europe has recently become available (PRINSEN

ET AL., 2011a).

VII.1. Burying cables

The ultimate solution to avoid collision is to bury the problematic wires. This solution is very

expensive (a British study estimates lifetime costs of £10.2-24.1 million/km for a buried line, as

opposed to £2.2-4.2 million/km for an overhead lines (cited in BIO INTELLIGENCE SERVICE, 2012))

and difficult to implement in the case of operational high-voltage power lines. A buried high-voltage

line also involves the potential risk of emitting heat and electro-magnetic fields, but this is very

unlikely to have any environmental impact (BIO INTELLIGENCE SERVICE, 2012). Burying the line

could however be preferable when planning a new line crossing known very critical bird areas, if a

deviation of the power line is impossible.

VII.2. Removing static wire

Observations show that the static wire (“Earth Wire”) is the most dangerous cable for birds in a

transmission overhead line. When flying, even if birds can generally see bigger conductor cables, they

more frequently fail to see the static wires and collide with them. This is because earth wires are

generally smaller in diameter and thus less visible than conductors. In two studies where collision

events were observed, collision with the earth wire accounted for 82% and 93% of all events (FAANES,

1987; SAVERENO ET AL., 1996). Thus, one way to reduce collision rates would be simply to remove the

static wire (JENKINS ET AL., 2010): this solution would be relatively costly and might lead to power

outages caused by lightning. As a result, this is unlikely to be considered as a viable solution until new

anti-lightning solutions are developed (PRINSEN ET AL., 2011b).

VII.3. Habitat management

Another possibility would be to take habitat management measures around existing dangerous lines.

The idea here would be to keep birds away from power lines by adjusting the “attractiveness” of the

nearby habitat (PRINSEN ET AL., 2011b). If a power line is located between a roost and a feeding place,

a theoretical possibility would be to create a new roosting or feeding place on the same side of the line

(BROWN & DREWIEN, 1995). However, no actual practical examples of such large-scale mitigation

measures are currently available in the literature (PRINSEN ET AL., 2011a). This is probably because

this type of management would involve taking action on neighbouring land and would require a

significant level of habitat modification, not to mention high environmental and financial costs

(PRINSEN ET AL., 2011b). Most of the time, this is not a realistic approach on a large-scale basis.

VII.4. Marking lines

Overview

The most frequently cited and realistic option for reducing the collision risk is to make power lines

more visible for the birds. Different kinds of devices have been tested; they are collectively referred to

as “diverters”. These accessories make the lines more visible for birds and also for other sky users

(parachutists, paragliders, microlights, helicopters, etc.). Some studies have shown that the frequency

of collisions is reduced by 25 to 95% when diverters are fitted (BROWN & DREWIEN, 1995; SAVERENO

ET AL., 1996; KABOUCHE ET AL., 2006; YEE, 2008). Others have shown a low impact for some species

(JENKINS ET AL., 2010). Recently, a meta-analysis based on 21 different studies and 52 wire-marking

experiments around the world showed that the mortality rate was on average 78% lower where there

were marked lines (BARRIENTOS ET AL., 2011): for unmarked lines, the collision rate was 0.21/1,000

bird crossings, while it was reduced to 0.05/1,000 bird crossings in the case of marked lines.

Interestingly, when marked and unmarked lines are present, the decrease in collision risk in the

marked sectors did not result in an increase of the risk in the nearby unmarked sectors (BARRIENTOS

ET AL., 2011). This is because birds avoid the power lines they notice, by flying over, rather than by

passing the detected line alongside.

The review of the literature did not manage to discriminate relative effectiveness of the variety of the

device available on the market. Effectiveness of visual markers may depend on the landscape and the

targeted bird species. The question of the optimal density of visual markers has not yet been examined.

Consequently, there is a lot of room for improvement, with in particular investigations of non-visual

devices being needed (BARRIENTOS ET AL., 2011).

On the basis of current knowledge, it can be concluded that any large device (increasing the apparent

size of the line to at least 20 cm) placed at least every 5-10 m along the line, preferably on the earth

wire, is likely to significantly reduce collision risks (JENKINS ET AL., 2010; PRINSEN ET AL., 2011b).

The marking colour seems to be an important factor to consider. Yellow appears to be one of the most

effective colours for increasing visibility and reducing mortality in all weather conditions, but yellow

dampers cause UV-sensitivity problems (BROWN & DREWIEN, 1995). Red is more visible during clear

weather with blue sky and even in fog. White (or other pale colours) are more visible during the night

(JANSS & FERRER, 1998). Special colours reflecting light and/or UV are also effective (BROWN &

DREWIEN, 1995). Although this has not been specifically tested, we can thus recommend that when the

critical line concerns for example an important gull night-roost, white is to be preferred, while for a

line crossing a river valley subject to fog, red is the colour that should be chosen.

A brief review of some available diverters

Aviation balls

Aviation balls limit the collision risk but are less effective than dampers or fibreglass plates (BROWN

& DREWIEN, 1995). However, it is likely that lines already equipped with aviation balls (or even

“spacers” sometimes used to keep cables away from each other) are sufficiently visible by birds, given

that the balls or spacers have the characteristics mentioned above.

Spirals

When red and white spirals were used in Southern France, where they were placed at intervals of 30 m

on high-voltage power lines, they reduced bird mortality by up to between 89 and 95% (KABOUCHE ET

AL., 2006). Their high visibility and a certain amount of noise (created by the wind blowing against the

spiral) alerted the (flying) birds to danger. Spirals can be placed on the high-voltage power lines by

helicopter. Red spirals are believed to be more effective in case of frequent fog, while white spirals are

believed to be more effective at night.

In Spain, orange polypropylene spirals (30 x 100 cm, made by APRESA) were placed on static wires

above distribution lines with a diameter of more than 2 cm. If there was only one static wire, one spiral

was placed every 5-10 m along the wire; if there were two wires, spirals were fitted every 10-20 m in

an alternate manner. Some spirals broke because of the effects of wind and the sun (GOBIERNO DE

ARAGON, 2007). Another study in Spain showed that white spirals on distribution lines reduced

Reducing bird mortality caused by high- and very-high-voltage power lines

mortality of local birds (including the Common Crane, the Great Bustard, the Little Bustard, the

Wildfowl, the Lapwing and Storks) by 81%: Orange spirals also seem to reduce risks of collisions but

the effectiveness seems to be lower for nocturnal birds, like the Eagle Owl (JANSS & FERRER, 1998).

In Colorado (USA), yellow PVC spiral vibration dampers (1.27 cm in diameter, 112-125 cm in length)

significantly reduced mortality for cranes, raptors and waterfowl colliding with distribution and

transmission lines (BROWN & DREWIEN, 1995).

The “Firefly”

The “Firefly”, developed by HAMMARPRODUKTER, is a rotating hard plastic disc or rectangular plate

with fluorescent parts that scares birds. These devices reflect sunlight during the daylight hours and

emit luminescent light at twilight and at night. Studies show that it is effective in reducing crane

collision in California (YEE, 2008). HAMMARPRODUKTER specifies that this diverter is easy to fit.

Because of its relatively small size and light weight, distances between individual Firefly BDFs can be

small (often within the quoted 5-10 m distance).

Bird flappers

In Germany, new marking devices consisting of 50-cm-long hard plastic black and white strips fitted

to an aluminium clamp are now extensively placed on transmission lines. The company RWE (see

below) also developed a helicopter allowing rapid installation without impairing the power supply.

These systems have proven effective (HARTMAN ET AL., 2010; PRINSEN ET AL., 2011b), especially in

night conditions (waterfowl) and close to a large gull roost. The reduction however was negligible for

some species (Coot). The German bird flappers are large and exhibit a high level of contrast and the

strips move individually in the wind, resulting in a blinking effect that probably makes them clearly

visible for approaching birds at twilight or even in darkness. Because of their weight, it is not possible

to install a large number on a single span of an earth wire; instead, there are typically several dozen

metres between individual flappers (PRINSEN ET AL., 2011b).

Other types of diverters

Another kind of diverter was developed by SAPREM. They are placed with a specialized machine that

runs along the cables. These are composed of two neoprene black bands hanging from the wire like a

clothespin. In addition to the form, these tags have photoluminescent properties. These tags have been

used in Aragon on distribution lines for wires with a diameter of less than 2 cm. Two years after

installation, tags were still in good shape and had retained their colour. Only in the case of significant

slopes did some tags glide along the wire (GOBIERNO DE ARAGON, 2007). A Spanish study showed

that these markers reduced the mortality rate by 76% on a distribution line (JANSS & FERRER, 1998).

However, this experiment concerned collision risks of Great Bustards with distribution lines, and this

type of markers may be too small for transmission lines (except for the earth wire). The tags may

prove to be more effective than coloured spirals but less durable (BARRIENTOS ET AL., 2011).

In southern Spain, three black stripes of 70 x 0.8 cm hanging on a distribution line did not reduce the

mortality of bustards, cranes and lapwings (JANSS & FERRER, 1998).

Cost Diverters - cost and practical information

Spiral: €150/spiral, including helicopter - €80/spiral without helicopter costs (KABOUCHE ET AL.,

2006). Individual spirals must be placed at an interval of 30 m on the two external wires, in an

alternate manner (thus resulting in a visible spiral every 15 m). The total cost could thus amount to

€5,000 to €8,500/km.

Spirals are available from PLP-APRESA in Spain > http://www.plp-spain.com/en/presentacion.aspx

Bird flappers: information available here > www.rwerheinruhrnetzservice.com

Neopren X-Band: information available here > http://www.saprem.com/emenu.html (this company

also provides a robot system to install the device on transmission lines)

Firefly: information available here >

http://www.hammarprodukter.com/659.php?itemgroup=107

From the distributor in Belgium (info@foreestgroenconsult.nl), we learned that the cost of a single

Firefly is €28.70 to €31.30 per piece, depending on the ordered quantity. For materials only, the cost is

thus around €3,000 to equip one wire with a Firefly every 10 m over a distance of 1 km (this does not

take into account the installation cost).

The price of placing a marker will of course depend on several factors, especially the total number of

diverters fitted, but apparently, a cost of as low as €2,500/km can be obtained in some circumstances

(PRINSEN ET AL., 2011a). However, Elia has learned that it would cost as much as €25,000 per line

segment to adjust diverters if the line section has to be shut down during installation (J. Mortier, pers.

com.). On the basis of the available literature, this price seems very high. An investigation is necessary

and direct contact with the producer and/or potential installer of diverters by Elia would be welcome to

define the exact budget. Also the fact that priority segments are scattered around the country could

increase the relative price (because of the need for dispersed interventions).

If we take as “high priority” a pylon score of above 60 (from Figure 13), this would cover 1.2% of the

total Elia network, so approximately 72 km. Based on an estimated price of €10,000/km (quite high on

the basis of the available literature, but to be checked with the producers), an estimated budget of

€720,000 would be needed to fit the most dangerous power lines in Belgium with diverters.

Reducing bird mortality caused by high- and very-high-voltage power lines

VIII. Conclusions

We believe that we, through this publication, have provided Elia with an objective and documented

basis for decision-making concerning the mitigation of bird collisions with existing power lines.

First, a Belgian list of bird species sensitive to collision was constructed, combining a review of the

literature, casualty reports and local expert judgements.

In the second step, using the best available data, high-resolution distribution maps for these sensitive

species were obtained. It must be stressed that some of these maps represent a breakthrough: relative

density maps for common breeding bird species, using combined data from three different breeding

bird atlases and state-of-the-art spatial modelling techniques, were never compiled before for Belgium.

This exercise also demonstrates that bird data collected independently in the three regions of this

federal country can be combined in a meaningful way to produce nationwide results. It also shows that

a data portal like waarnemingen.be/observations.be can really provide a useful, and in some cases

even indispensable, data source for such applications as high-resolution mapping.

Finally, these maps were combined, using a scoring system designed to emphasize areas where

damage associated with power-line collisions is believed to be most significant. By doing so, power-

line sections are classified according to the risk they present for birds. Most of the power lines are not

to be considered “dangerous” for birds, but in some areas the situation is critical, especially close to

where major waterbird concentrations occur.

There are several effective technical measures that have been designed to prevent power line collision:

diverters can be fitted directly on the line so that the power line is more visible for birds, even in poor

light conditions.

Depending on the available budget, diverter placement can now be planned on Belgian power lines,

focusing first on the high-priority sections. Of course, other considerations like geographical

“grouping” of diverter placement in order to reduce cost and opportunities offered by other planned

work on specific power-line sections may also influence the final planning that lies now in the hands

of Elia.

Annex I. Classification of Belgian birds according to the

power-line collision risk

English name Scientific name Sensitivity to collision Conservation relevance

Mute Swan Cygnus olor 2 0

Bewick's Swan Cygnus columbianus

bewickii 2 2

Whooper Swan Cygnus cygnus 2 2

Taiga Bean Goose Anser fabalis 2 1

Tundra Bean Goose Anser rossicus 2 2

Pink-footed Goose Anser brachyrhynchus 2 2

White-fronted Goose Anser albifrons 2 2

Lesser White-fronted Goose Anser erythropus 2 1

Greylag Goose Anser anser 2 2

Lesser Snow Goose Anser caerulescens 2 0

Barnacle Goose Branta leucopsis 2 1

Dark-bellied Brent Goose Branta bernicla 2 1

Pale-bellied Brent Goose Branta hrota 2 0

Black Brant Branta nigricans 2 0

Red-breasted Goose Branta ruficollis 2 0

Ruddy Shelduck Tadorna ferruginea 2 0

Common Shelduck Tadorna tadorna 2 2

Eurasian Wigeon Anas penelope 2 2

American Wigeon Anas americana 2 0

Gadwall Anas strepera 2 2

Baikal Teal Anas formosa 2 0

Eurasian Teal Anas crecca 2 2

Green-winged Teal Anas carolinensis 2 0

Mallard Anas platyrhynchos 2 2

Northern Pintail Anas acuta 2 2

Garganey Anas querquedula 2 2

Blue-winged Teal Anas discors 2 0

Northern Shoveler Anas clypeata 2 2

Red-crested Pochard Netta rufina 2 1

Common Pochard Aythya ferina 2 2

Ring-necked Duck Aythya collaris 2 0

Ferruginous Duck Aythya nyroca 2 1

Tufted Duck Aythya fuligula 2 2

Greater Scaup Aythya marila 2 1

Lesser Scaup Aythya affinis 2 0

Common Eider Somateria mollissima 2 0

King Eider Somateria spectabilis 2 0

Long-tailed Duck Clangula hyemalis 2 0

Common Scoter Melanitta nigra 2 0

Surf Scoter Melanitta perspicillata 2 0

Velvet Scoter Melanitta fusca 2 0

Common Goldeneye Bucephala clangula 2 0

Smew Mergus albellus 2 2

Red-breasted Merganser Mergus serrator 2 0

Goosander Mergus merganser 2 2

White-headed Duck Oxyura leucocephala 2 0

Hazel Grouse Bonasa bonasia 2 2

Black Grouse Tetrao tetrix 2 2

Western Capercaillie Tetrao urogallus 2 0

Grey Partridge Perdix perdix 2 2

Common Quail Coturnix coturnix 2 1

Red-throated Diver Gavia stellata 0 1

Black-throated Diver Gavia arctica 0 1

Great Northern Diver Gavia immer 0 0

White-billed Diver Gavia adamsii 0 0

Little Grebe Tachybaptus ruficollis 1 0

Great Crested Grebe Podiceps cristatus 1 0

Red-necked Grebe Podiceps grisegena 1 0

Slavonian Grebe Podiceps auritus 1 0

Black-necked Grebe Podiceps nigricollis 1 1

Northern Fulmar Fulmarus glacialis 0 0

Cory's Shearwater Calonectris diomedea 0 0

Great Shearwater Puffinus gravis 0 0

Reducing bird mortality caused by high- and very-high-voltage power lines

Sooty Shearwater Puffinus griseus 0 0

Manx Shearwater Puffinus puffinus 0 0

Balearic Shearwater Puffinus mauretanicus 0 0

Little Shearwater Puffinus assimilis 0 0

European Storm-petrel Hydrobates pelagicus 0 0

Leach's Storm-petrel Oceanodroma leucorhoa 0 0

Northern Gannet Morus bassanus 1 0

Great Cormorant Phalacrocorax carbo 2 2

European Shag Phalacrocorax aristotelis 1 0

Pygmy Cormorant Phalacrocorax pygmeus 1 0

Great Bittern Botaurus stellaris 1 2

Little Bittern Ixobrychus minutus 1 2

Black-crowned Night Heron Nycticorax nycticorax 1 1

Squacco Heron Ardeola ralloides 1 0

Cattle Egret Bubulcus ibis 1 1

Little Egret Egretta garzetta 1 1

Great Egret Casmerodius albus 1 1

Grey Heron Ardea cinerea 2 0

Purple Heron Ardea purpurea 1 1

Black Stork Ciconia nigra 2 2

White Stork Ciconia ciconia 2 2

Glossy Ibis Plegadis falcinellus 1 0

Eurasian Spoonbill Platalea leucorodia 1 1

European Honey-buzzard Pernis apivorus 1 2

Black-winged Kite Elanus caeruleus 1 0

Black Kite Milvus migrans 1 2

Red Kite Milvus milvus 1 2

White-tailed Eagle Haliaeetus albicilla 1 0

Egyptian Vulture Neophron percnopterus 1 0

Eurasian Griffon Gyps fulvus 1 0

Short-toed Eagle Circaetus gallicus 1 0

Marsh Harrier Circus aeruginosus 1 2

Hen Harrier Circus cyaneus 1 2

Pallid Harrier Circus macrourus 1 1

Montagu's Harrier Circus pygargus 1 2

Northern Goshawk Accipiter gentilis 1 1

Eurasian Sparrowhawk Accipiter nisus 2 0

Common Buzzard Buteo buteo 2 1

Rough-legged Buzzard Buteo lagopus 1 0

Lesser Spotted Eagle Aquila pomarina 1 0

Greater Spotted Eagle Aquila clanga 1 0

Steppe Eagle Aquila nipalensis 1 0

Golden Eagle Aquila chrysaetos 1 0

Booted Eagle Hieraaetus pennatus 1 0

Bonelli's Eagle Hieraaetus fasciatus 1 0

Osprey Pandion haliaetus 1 1

Lesser Kestrel Falco naumanni 1 0

Common Kestrel Falco tinnunculus 1 1

Red-footed Falcon Falco vespertinus 1 0

European Merlin Falco columbarius 1 1

Eurasian Hobby Falco subbuteo 1 1

Gyrfalcon Falco rusticolus 1 0

Peregrine Falcon Falco peregrinus 1 2

Water Rail Rallus aquaticus 2 2

Spotted Crake Porzana porzana 2 2

Little Crake Porzana parva 2 1

Baillon's Crake Porzana pusilla 2 1

Corn Crake Crex crex 2 2

Common Moorhen Gallinula chloropus 2 1

Common Coot Fulica atra atra 2 1

Common Crane Grus grus 2 2

Little Bustard Tetrax tetrax 2 0

Macqueen's Bustard Chlamydotis macqueenii 2 0

Great Bustard Otis tarda 2 0

Eurasian Oystercatcher Haematopus ostralegus 2 0

Black-winged Stilt Himantopus himantopus 2 1

Pied Avocet Recurvirostra avosetta 2 2

Stone-curlew Burhinus oedicnemus 2 0

Cream-coloured Courser Cursorius cursor 2 0

Collared Pratincole Glareola pratincola 2 0

Black-winged Pratincole Glareola nordmanni 2 0

Little Ringed Plover Charadrius dubius 2 1

Ringed Plover Charadrius hiaticula 2 1

Kentish Plover Charadrius alexandrinus 2 2

Lesser Sand Plover Charadrius mongolus 2 0

Greater Sand Plover Charadrius leschenaultii 2 0

Eurasian Dotterel Eudromias morinellus 2 2

Pacific Golden Plover Pluvialis fulva 2 0

Golden Plover Pluvialis apricaria 2 2

Grey Plover Pluvialis squatarola 2 0

Sociable Lapwing Vanellus gregarius 2 0

White-tailed Lapwing Vanellus leucurus 2 0

Northern Lapwing Vanellus vanellus 2 2

Red Knot Calidris canutus 2 0

Sanderling Calidris alba 2 0

Red-necked Stint Calidris ruficollis 2 0

Little Stint Calidris minuta 2 0

Temminck's Stint Calidris temminckii 2 0

Least Sandpiper Calidris minutilla 2 0

Bonaparte's Sandpiper Calidris fuscicollis 2 0

Baird's Sandpiper Calidris bairdii 2 0

Pectoral Sandpiper Calidris melanotos 2 0

Sharp-tailed Sandpiper Calidris acuminata 2 0

Curlew Sandpiper Calidris ferruginea 2 0

Purple Sandpiper Calidris maritima 2 0

Dunlin Calidris alpina 2 0

Broad-billed Sandpiper Limicola falcinellus 2 0

Buff-breasted Sandpiper Tryngites subruficollis 2 0

Ruff Philomachus pugnax 2 0

Jack Snipe Lymnocryptes minimus 2 1

Common Snipe Gallinago gallinago 2 2

Great Snipe Gallinago media 2 0

Long-billed Dowitcher Limnodromus scolopaceus 2 0

Eurasian Woodcock Scolopax rusticola 2 1

Black-tailed Godwit Limosa limosa 2 2

Bar-tailed Godwit Limosa lapponica 2 0

Whimbrel Numenius phaeopus 2 0

Slender-billed Curlew Numenius tenuirostris 2 0

Eurasian Curlew Numenius arquata 2 2

Spotted Redshank Tringa erythropus 2 0

Common Redshank Tringa totanus 2 2

Marsh Sandpiper Tringa stagnatilis 2 0

Common Greenshank Tringa nebularia 2 0

Greater Yellowlegs Tringa melanoleuca 2 0

Lesser Yellowlegs Tringa flavipes 2 0

Green Sandpiper Tringa ochropus 2 0

Wood Sandpiper Tringa glareola 2 0

Terek Sandpiper Xenus cinereus 2 0

Common Sandpiper Actitis hypoleucos 2 0

Spotted Sandpiper Actitis macularius 2 0

Ruddy Turnstone Arenaria interpres 2 0

Wilson's Phalarope Phalaropus tricolor 2 0

Red-necked Phalarope Phalaropus lobatus 2 0

Red Phalarope Phalaropus fulicarius 2 0

Pomarine Skua Stercorarius pomarinus 1 0

Arctic Skua Stercorarius parasiticus 1 0

Long-tailed Skua Stercorarius longicaudus 1 0

Great Skua Stercorarius skua 1 0

Great Black-headed Gull Larus ichtyaetus 2 0

Mediterranean Gull Larus melanocephalus 2 2

Laughing Gull Larus atricilla 2 0

Franklin's Gull Larus pipixcan 2 0

Little Gull Larus minutus 2 0

Sabine's Gull Larus sabini 2 0

Bonaparte's Gull Larus philadelphia 2 0

Black-headed Gull Larus ridibundus 2 2

Ring-billed Gull Larus delawarensis 2 0

Common Gull Larus canus 2 2

Lesser Black-backed Gull Larus fuscus 2 2

Herring Gull Larus argentatus 2 2

Caspian Gull Larus cachinnans 2 0

Yellow-legged Gull Larus michahellis 2 0

Iceland Gull Larus glaucoides 2 0

Glaucous Gull Larus hyperboreus 2 0

Great Black-backed Gull Larus marinus 2 0

Black-legged Kittiwake Rissa tridactyla 2 0

Gull-billed Tern Sterna nilotica 1 0

Caspian Tern Sterna caspia 1 0

Sandwich Tern Sterna sandvicensis 1 2

Roseate Tern Sterna dougallii 1 0

Common Tern Sterna hirundo 1 2

Arctic Tern Sterna paradisaea 1 1

Reducing bird mortality caused by high- and very-high-voltage power lines

Forster's Tern Sterna forsteri 1 0

Bridled Tern Sterna anaethetus 1 0

Little Tern Sterna albifrons 1 2

Whiskered Tern Chlidonias hybrida 1 0

Black Tern Chlidonias niger 1 1

White-winged Black Tern Chlidonias leucopterus 1 0

Common Guillemot Uria aalge 0 0

Brünnich's Guillemot Uria lomvia 0 0

Razorbill Alca torda 0 0

Black Guillemot Cepphus grylle 0 0

Little Auk Alle alle 0 0

Atlantic Puffin Fratercula arctica 0 0

Black-bellied Sandgrouse Pterocles orientalis 1 0

Pallas's sandgrouse Syrrhaptes paradoxus 1 0

Stock Dove Columba oenas 2 1

Common Wood Pigeon Columba palumbus 2 1

Eurasian Collared Dove Streptopelia decaocto 2 0

European Turtle Dove Streptopelia turtur 2 2

Great Spotted Cuckoo Clamator glandarius 2 0

Common Cuckoo Cuculus canorus 1 2

Yellow-billed Cuckoo Coccyzus americanus 2 0

Western Barn Owl Tyto alba 2 2

Eurasian Scops Owl Otus scops 2 0

Eurasian Eagle Owl Bubo bubo 2 2

Snowy Owl Bubo scandiacus 2 0

Hawk Owl Surnia ulula 2 0

Eurasian Pygmy Owl Glaucidium passerinum 2 0

Little Owl Athene noctua 2 1

Tawny Owl Strix aluco 2 1

Long-eared Owl Asio otus 2 1

Short-eared Owl Asio flammeus 2 1

Tengmalm's Owl Aegolius funereus 2 2

European Nightjar Caprimulgus europaeus 2 2

Common Swift Apus apus 2 1

Pallid Swift Apus pallidus 2 0

Alpine Swift Apus melba 2 0

Common Kingfisher Alcedo atthis 2 2

European Bee-eater Merops apiaster 2 0

European Roller Coracias garrulus 2 0

Eurasian Hoopoe Upupa epops 2 0

Eurasian Wryneck Jynx torquilla 2 2

Grey-headed Woodpecker Picus canus 2 2

Green Woodpecker Picus viridis 2 2

Black Woodpecker Dryocopus martius 2 2

Great Spotted Woodpecker Dendrocopos major 2 0

Middle Spotted Woodpecker

Dendrocopos medius 2 2

White-backed Woodpecker Dendrocopos leucotos 2 0

Lesser Spotted Woodpecker Dendrocopos minor 2 1

Calandra Lark Melanocorypha calandra 1 0

Greater Short-toed Lark Calandrella brachydactyla 1 0

Lesser Short-toed Lark Calandrella rufescens 1 0

Crested Lark Galerida cristata 1 2

Wood Lark Lullula arborea 1 2

Sky Lark Alauda arvensis 1 2

Horned Lark Eremophila alpestris 1 0

Sand Martin Riparia riparia 1 1

Eurasian Crag Martin Ptyonoprogne rupestris 1 0

Barn Swallow Hirundo rustica 1 0

Red-rumped Swallow Hirundo daurica 1 0

House Martin Delichon urbicum 1 2

Richard's Pipit Anthus richardi 1 0

Blyth's Pipit Anthus godlewskii 1 0

Tawny Pipit Anthus campestris 1 1

Olive-backed Pipit Anthus hodgsoni 1 0

Tree Pipit Anthus trivialis 1 2

Meadow Pipit Anthus pratensis 1 2

Red-throated Pipit Anthus cervinus 1 0

Water Pipit Anthus spinoletta 1 0

Rock Pipit Anthus petrosus 1 0

Blue-headed Wagtail Motacilla flava 1 1

Grey-headed Wagtail Motacilla flava thunbergi 1 0

Black-headed Wagtail Motacilla flava feldegg 1 0

British Yellow Wagtail Motacilla flavissima 1 0

Citrine Wagtail Motacilla citreola 1 0

Grey Wagtail Motacilla cinerea 1 0

White Wagtail Motacilla alba 1 0

Pied Wagtail Motacilla yarrellii 1 0

Bohemian Waxwing Bombycilla garrulus 1 0

Dipper Cinclus cinclus 1 0

Winter Wren Troglodytes troglodytes 1 0

Dunnock Prunella modularis 1 0

Alpine Accentor Prunella collaris 1 0

European Robin Erithacus rubecula 1 0

Thrush Nightingale Luscinia luscinia 1 0

Common Nightingale Luscinia megarhynchos 1 2

Bluethroat Luscinia svecica 1 0

Red-flanked Bluetail Tarsiger cyanurus 1 0

Black Redstart Phoenicurus ochruros 1 0

Common Redstart Phoenicurus phoenicurus 1 2

Whinchat Saxicola rubetra 1 2

Stonechat Saxicola torquatus 1 0

Siberian Stonechat Saxicola maurus 1 0

Isabelline Wheatear Oenanthe isabellina 1 0

Northern Wheatear Oenanthe oenanthe 1 2

Black-eared Wheatear Oenanthe hispanica 1 0

Desert Wheatear Oenanthe deserti 1 0

Rufous-tailed Rock Thrush Monticola saxatilis 1 0

Blue Rock Thrush Monticola solitarius 1 0

White's Thrush Zoothera dauma 1 0

Siberian Thrush Zoothera sibirica 1 0

Swainson's Thrush Catharus ustulatus 1 0

Scandinavian Ring Ouzel Turdus torquatus torquatus 1 0

Alpine Ring Ouzel Turdus torquatus alpestris 1 2

Common Blackbird Turdus merula 1 0

Eyebrowed Thrush Turdus obscurus 1 0

Naumann's Thrush Turdus naumanni 1 0

Dusky Thrush Turdus eunomus 1 0

Black-throated Thrush Turdus atrogularis 1 0

Fieldfare Turdus pilaris 1 2

Song Thrush Turdus philomelos 1 0

Redwing Turdus iliacus 1 0

Mistle Thrush Turdus viscivorus 1 0

American Robin Turdus migratorius 1 0

Cetti's Warbler Cettia cetti 1 1

Zitting Cisticola Cisticola juncidis 1 1

Pallas's Grasshopper

Warbler Locustella certhiola 1 0

Lanceolated Warbler Locustella lanceolata 1 0

Grasshopper Warbler Locustella naevia 1 1

River Warbler Locustella fluviatilis 1 0

Savi's Warbler Locustella luscinioides 1 2

Aquatic Warbler Acrocephalus paludicola 1 2

Sedge Warbler Acrocephalus

schoenobaenus 1 2

Eurasian Reed Warbler Acrocephalus scirpaceus 1 0

Marsh Warbler Acrocephalus palustris 1 0

Blyth's Reed Warbler Acrocephalus dumetorum 1 0

Paddyfield Warbler Acrocephalus agricola 1 0

Great Reed Warbler Acrocephalus arundinaceus 1 2

Booted Warbler Hippolais caligata 1 0

Icterine Warbler Hippolais icterina 1 1

Melodious Warbler Hippolais polyglotta 1 0

Blackcap Sylvia atricapilla 1 0

Garden Warbler Sylvia borin 1 0

Barred Warbler Sylvia nisoria 1 0

Lesser Whitethroat Sylvia curruca 1 1

Orphean Warbler Sylvia hortensis 1 0

Common Whitethroat Sylvia communis 1 0

Spectacled Warbler Sylvia conspicillata 1 0

Dartford Warbler Sylvia undata 1 0

Marmora's Warbler Sylvia sarda 1 0

Subalpine Warbler Sylvia cantillans 1 0

Sardinian Warbler Sylvia melanocephala 1 0

Greenish Warbler Phylloscopus trochiloides 1 0

Arctic Warbler Phylloscopus borealis 1 0

Pallas's Warbler Phylloscopus proregulus 1 0

Yellow-browed Warbler Phylloscopus inornatus 1 0

Hume's Warbler Phylloscopus humei 1 0

Radde's Warbler Phylloscopus schwarzi 1 0

Dusky Warbler Phylloscopus fuscatus 1 0

Bonelli's Warbler Phylloscopus bonelli 1 0

Wood Warbler Phylloscopus sibilatrix 1 0

Reducing bird mortality caused by high- and very-high-voltage power lines

Chiffchaff Phylloscopus collybita 1 0

Iberian Chiffchaff Phylloscopus ibericus 1 0

Willow Warbler Phylloscopus trochilus 1 0

Goldcrest Regulus regulus 1 0

Firecrest Regulus ignicapilla 1 0

Spotted Flycatcher Muscicapa striata 1 0

Red-breasted Flycatcher Ficedula parva 1 0

Collared Flycatcher Ficedula albicollis 1 0

Pied Flycatcher Ficedula hypoleuca 1 0

Bearded Tit Panurus biarmicus 1 1

Long-tailed Tit Aegithalos caudatus 1 0

Marsh Tit Parus palustris 1 0

Willow Tit Parus montanus 1 2

Crested Tit Parus cristatus 1 0

Coal Tit Parus ater 1 0

Blue Tit Parus caeruleus 1 0

Great Tit Parus major 1 0

Wood Nuthatch Sitta europaea 1 0

Wallcreeper Tichodroma muraria 1 0

Treecreeper Certhia familiaris 1 0

Short-toed Treecreeper Certhia brachydactyla 1 0

Penduline Tit Remiz pendulinus 1 1

Golden Oriole Oriolus oriolus 1 2

Isabelline Shrike Lanius isabellinus 1 0

Red-backed Shrike Lanius collurio 1 2

Lesser Grey Shrike Lanius minor 1 0

Great Grey Shrike Lanius excubitor 1 2

Woodchat Shrike Lanius senator 1 0

Eurasian Jay Garrulus glandarius 1 0

Magpie Pica pica 1 0

Spotted Nutcracker Nucifraga caryocatactes 1 1

Jackdaw Corvus monedula 2 0

Rook Corvus frugilegus 1 0

Carrion Crow Corvus corone 1 0

Hooded Crow Corvus cornix 1 0

Common Raven Corvus corax 1 2

Common Starling Sturnus vulgaris 2 0

Rose-coloured Starling Sturnus roseus 2 0

House Sparrow Passer domesticus 1 0

Tree Sparrow Passer montanus 1 1

Rock Sparrow Petronia petronia 1 0

White-winged Snowfinch Montifringilla nivalis 1 0

Red-eyed Vireo Vireo olivaceus 1 0

Chaffinch Fringilla coelebs 1 0

Brambling Fringilla montifringilla 1 0

European Serin Serinus serinus 1 2

Citril Finch Serinus citrinella 1 0

European Greenfinch Carduelis chloris 1 0

Goldfinch Carduelis carduelis 1 0

Eurasian Siskin Carduelis spinus 1 0

Common Linnet Carduelis cannabina 1 0

Twite Carduelis flavirostris 1 0

Mealy Redpoll Carduelis flammea 1 0

Lesser Redpoll Carduelis cabaret 1 2

Arctic Redpoll Carduelis hornemanni 1 0

Two-barred Crossbill Loxia leucoptera 1 0

Common Crossbill Loxia curvirostra 1 0

Parrot Crossbill Loxia pytyopsittacus 1 0

Common Rosefinch Carpodacus erythrinus 1 0

Bullfinch Pyrrhula pyrrhula 1 2

Hawfinch Coccothraustes

coccothraustes 1 0

Lapland Bunting Calcarius lapponicus 1 0

Snow Bunting Plectrophenax nivalis 1 0

Pine Bunting Emberiza leucocephalos 1 0

Yellowhammer Emberiza citrinella 1 2

Cirl Bunting Emberiza cirlus 1 0

Rock Bunting Emberiza cia 1 0

Ortolan Bunting Emberiza hortulana 1 1

Yellow-browed Bunting Emberiza chrysophrys 1 0

Rustic Bunting Emberiza rustica 1 0

Little Bunting Emberiza pusilla 1 0

Yellow-breasted Bunting Emberiza aureola 1 0

Reed Bunting Emberiza schoeniclus 1 2

Corn Bunting Emberiza calandra 1 2

Annex II: Collision victims brought into a recovery centre

Data from Vogelbescherming (1990-2010) and the Ligue Royale Belge pour la Protection des Oiseaux

(1990-2000).

Dutch Name Scientific Name 1991 1992 1993 1994 1995 1 996 1997 1998 1999 200 0 2001 2002 2003 20 04 2005 20 06 200 7 20 08 2 009 2010 2 011 Total

Aalscholver Phalacrocorax carbo 1 1 1 1 4 1 2 3 1 2 17

Bergeend Tadorna tadorna 1 1

Blauwe Kiekendie f Circus cyaneus 1 1 2

Blauwe Reiger Ardea ci ne rea 9 5 9 7 5 5 4 12 3 7 5 6 2 8 12 10 6 8 12 10 4 1 49

Boerenzw aluw Hirund o rustica 2 1 1 4

Bokje Lymnocryptes minimus 1 1

Boomklever Sit ta europaea 1 1 2

Boomvalk Falco subbuteo 1 1 1 1 1 1 6

Bosuil Strix al uco 1 1 1 1 1 8 2 4 6 2 27

Brandgans Brant a Leucops is 1 1 1 1 1 2 1 8

Buizerd Buteo buteo 4 1 5 3 6 1 3 2 5 1 4 1 4 3 2 4 2 1 4 4 60

Canadese Gans Branta canadensis 1 3 1 1 2 1 4 3 1 3 4 2 26

Ekster Pica pica 1 1 1 2 1 2 1 9

Fazant Phasianus colchicus 1 1 1 1 1 2 2 9

Fuut Po diceps cristatus 2 1 1 1 1 1 1 1 9

Gaai Garrul us gl andarius 1 1 1 1 4

Geoorde Fuut Podiceps nigric ollis 1 1

Gierzwal uw Apus apus 1 3 6 1 4 7 3 3 1 3 2 1 35

Grauwe Gans Anser anse r 1 1 3 1 5 11

Groene Specht Picus viridis 1 1 1 2 4 1 4 1 15

Groenling Chloris chloris 1 1

Grote Bonte Specht Dendroco pos major 2 2 2 6

Grote Burgemeester Larus h ype rboreus 1 1

Grote Lijster Turdus viscivorus 1 1 2

Grote Mantelmeeuw Larus marinus 1 1 1 1 2 6

Grutto Limosa limosa 1 1

Havik Accipiter gentilis 1 2 1 2 6

Holendui f Columba oenas 1 1 1 1 4

Houtduif Columba palu mbus 3 5 2 2 3 4 3 3 1 4 3 1 11 8 3 15 9 10 12 7 1 09

Houtsnip Scolopax rusticola 2 11 3 3 1 2 26 3 51

Huismus Passer domesticus 2 1 4 2 9

Huiszwa lu w Delichon urbicon 1 2 3

Kauw Corvus monedula 1 1 3 2 2 4 3 3 2 2 23

Kerkuil Tyto alba 2 3 1 2 1 2 1 6 3 1 4 1 27

Kievit Vanellus vanellus 2 2 1 1 1 2 1 1 1 12

Kleine Ma ntelmee uw Larus fuscus 1 1 2 1 1 2 8

Kleine Zwaa n Cygnus columbianus 1 1 2

Knobbelzwaa n Cygnus olor 1 3 4 3 1 3 1 2 2 2 2 6 3 8 1 42

Koekoek Cuculus canorus 1 1 2

Kokmeeuw Chroico cephalus ridi bundu s 20 7 6 8 11 8 3 2 6 8 4 6 9 1 3 4 4 6 3 5 5 1 29

Kolgans Anser albifrons 1 1

Koolmees Parus major 1 1 2 4

Kraai Corvus corone 1 1 2 1 2 1 2 1 2 13

Kraanvogel Grus grus 1 1

Lepelaar Pla talea leu corodia 1 1

Meerkoet Fulica at ra 1 1 2

Merel Turdus merula 1 1 1 1 3 2 5 4 1 6 7 32

Mori ne lplevier Charadrius morinellu s 1 1

Nijlga ns Alo pochen aegyptiaca 1 1 1 1 1 5

Oehoe Bubo bu bo 1 1 1 1 1 5

Ooievaar Ciconia ci conia 4 1 1 1 2 9

Patrijs Perdix perdix 1 1

Purperreig er Ardea purpurea 1 1 2

Pijl sta art Anas acuta 1 1

Raaf Corvus corax 1 1

Ransuil Asio otus 1 3 1 1 6

Regenwulp Numenius phae opu s 1 2 2 1 1 7

Rietgans Anser fabalis 1 1

Roek Corvus frugilegus 1 1

Roerdomp Botaurus stellaris 1 1 2 4

Roodborst Erithacus rubecula 1 2 2 5

Scholekster Haematopus ostralegus 1 1 2

Slechtvalk Falco peregr inu s 2 1 1 1 1 6

Smelleke n Falco columbarius 1 1

Sperwer Accipiter nisus 1 1 1 2 1 1 14 1 1 1 4 1 3 1 4 5 14 56

Spreeuw Sturnus vulgaris 1 1 2 2 1 2 9

Reducing bird mortality caused by high- and very-high-voltage power lines

Annex III: Bibliography

ARTS, B. & BRUINZEEL, L. (2009): De nationale windmolenrisicokaart voor

Reducing bird mortality caused by high- and very-high-voltage power lines in Belgium

Abstract and Figures