Data Interpretation

O R I G I N A L P A P E R

What is the impact of wind farms on birds? A case study in southern Spain

M. A. Farfán Æ J. M. Vargas Æ J. Duarte Æ R. Real

Received: 6 September 2007 / Accepted: 18 June 2009 / Published online: 5 July 2009 � Springer Science+Business Media B.V. 2009

Abstract Wind farming is a relatively new form of obtaining energy that does not cause air pollution or other forms of environmental degradation associated with fossil fuel

technologies. However, their use impacts on the environment, and the current rate at which

they are being put into operation, combined with poor understanding of their medium- and

long-term impact, is a cause of concern. Wind farms represent a new source of impact and

disturbance for birds that adds to the long list of disturbance factors caused by human

activity, such as power lines, radio and television towers, highways, glass windows, the

practice of poisoning, illegal hunting and overexploitation. Due to the precarious situation

of several bird species and their decline, any additional cause of mortality may be sig-

nificant and should give rise to increased attention and research. The aim of the present

work is to analyse the effect of the ‘‘Sierra de Aguas’’ wind farm on bird density and

abundance, flight behaviour, and bird mortality. Mortality rates did not increase due to the

presence of the wind turbines. The results suggest that the presence and operation of the

wind turbines did not have a clearly negative effect on passerine birds present in the region

where wind farm is located. However, raptors used the space around the wind farm with

lower frequency than prior to its existence, which represented a displacement of the home

range of these species.

Keywords Birds � Human activity � Renewable energy � Southern Spain � Wind farm

M. A. Farfán � J. Duarte Biogea Consultores, C/Navarro Ledesma no 243, 29010 Málaga, Spain

J. M. Vargas (&) � R. Real Departamento de Biologı́a Animal, Facultad de Ciencias, Universidad de Málaga, 29071 Málaga, Spain e-mail: [email protected]

123

Biodivers Conserv (2009) 18:3743–3758 DOI 10.1007/s10531-009-9677-4

Introduction

A basic aim throughout Europe is to reduce greenhouse gas emissions by the increased use

of energy from renewable sources. As a result, wind energy and other alternative energy

sources that produce ‘‘clean’’ energy have received strong public support (Leddy et al.

1999). This fact, together with the progressive reduction of the costs associated with wind

energy production (Hansen et al. 1992), has led to a considerable proliferation in the

number of wind farms around the world (Osborn et al. 2000).

Wind farms are a relatively new form of obtaining energy that does not cause air

pollution and other forms of environmental degradation associated with fossil fuel tech-

nologies (Nelson and Curry 1995). However, they impact on the environment, and the

current rate at which they are being put into operation, combined with poor understanding

of their medium- and long-term impact, is a cause of concern. There are at least four

aspects to take into account: (1) wind farms are a noise source; (2) they destroy plant cover

(Larsen and Madsen 2000; Fox et al. 2006; Larsen and Guillemette 2007); (3) they have an

aesthetic impact on the landscape; and (4) they have a negative effect on flying fauna,

especially birds (Nelson and Curry 1995). In the latter case we can distinguish two dif-

ferent effects. The first concerns direct mortality due to collision with turbines, and the

second consists in disturbances in the spatial use of territory caused by their presence.

Wind farms represent a new impact and disturbance source for birds (Orloff and

Flannery 1992) that adds to the long list of disturbance factors caused by human activity,

such as power lines (Ferrer et al. 1991; Janss and Ferrer 1998; Penteriani 1998), radio and

television towers (Stahlecker 1979; Smith 1985), highways (Fajardo et al. 1998), glass

windows (Klem 1990a, b), the practice of poisoning (Harmata et al. 1999; Vargas 2002),

illegal hunting (Lucio and Purroy 1992; Villafuerte et al. 1998) and overexploitation

(BirdLife International 2004). Nowadays, there is abundant literature on the high mortality

of birds caused by man-made structures, and the impact of these structures on bird pop-

ulations as well as the development of protective measures (Klem 1990a; Fajardo et al.

1998; Alonso and Alonso 1999). In the United States alone the annual number of birds

killed by collisions with these elements varies between 100 million and one billion (Klem

1990b; Manville 2000). However, as the commercial production of wind energy is recent—

in the United States it started at the beginning of the 1980s (American Wind Energy

Association (AWEA) 1995)—the effects of wind farms on bird populations has been little

studied.

The rapid development of wind farms in recent decades has generated controversy

regarding the negative influence of these structures on bird conservation (Drewitt and

Langston 2006). However, scientific literature on this issue is scarce. Several studies show

that the negative impact of wind farms on bird populations is much less than other sources

of collision (de Lucas et al. 2004, 2005). According to Erickson et al. (2001), the number

of bird deaths due to collisions with turbines represent 0.01–0.02% of the total, which is far

lower than impacts with other structures such as vehicles, buildings, glass windows, power

lines or communication towers. On the hand, other studies report higher mortality levels

and highlight raptors as being the most affected (Orloff and Flannery 1992; Barrios and

Rodrı́guez 2004).

In any case, due to the precarious situation of several bird species and their decline—

especially some raptors and other long-lived birds with low reproduction and slow mat-

uration rates—any cause of additional mortality can be significant and should give rise to

increased attention and research, particularly in cases such as wind farms where there is

cumulative mortality as a consequence of multiple installations (Drewitt and Langston

3744 Biodivers Conserv (2009) 18:3743–3758

123

2006). In this sense, monitoring programs and observing avifauna will provide better

information on the causes of mortality and the factors to take into account when identifying

the best location for wind farms. Since the factors that cause mortality in birds increase as

the human demand for resources grows, any initiative or effort by organisations and groups

involved in addressing the factors mentioned is relevant in reversing the current trend in

bird mortality.

The aim of the present work is to analyse the effect of the ‘‘Sierra de Aguas’’ wind farm

on bird density and abundance, flight behaviour and bird mortality.

Materials and methods

Study area

This work was performed in Malaga province (southern Spain) from March 2005 to

February 2007 (Fig. 1). This area was chosen because of its proximity to the Strait of

Gibraltar (*100 km away), which is one of the most important migrating routes in the Palaearctic (Bernis 1980; Bildstein and Zalles 2000).

The wind farm, called the ‘‘Sierra de Aguas’’, is located on a mountain ridge running

SW to NE. It consists of 16 wind turbines evenly distributed in two continuous rows

located between 815 and 940 m above sea level. The wind turbines are around 90 m apart

in each row. The two rows are separated by a 500-m corridor and their total length is

1,600 m each.

Each wind turbine generates 850 KW. Thus, the total power of the wind farm is

13.6 MW. The wind turbines have three blades with variable step. They are made of

Fig. 1 Location of the study area

Biodivers Conserv (2009) 18:3743–3758 3745

123

fibreglass and the rotors are orientated windward. Blade length is 25 m. They are white

except for the ends of the blades, which are red. The tower is cylindrical, 3.3 m in diameter

at the base and 44 m tall.

In the south-western row the vegetation is dominated by trees (Pinus pinaster and Pinus halepensis) due to afforestation. On the other hand, scrubland and pasture are more abundant around the north-eastern row. The most representative species in the study area

are Phlomis purpurea, Phlomis lychnitis, Quercus coccifera, Chamaerops humilis, Ros- marinus officinalis, Cistus albidus, Ulex parviflorus and Stipa tenacissima.

Several breeding raptor species can be found in the area, such as griffon vulture (Gyps fulvus), sparrowhawk (Accipiter nisus), golden eagle (Aquila chrysaetos), Bonelli’s eagle (Hieraaetus fasciatus), short-toed eagle (Circaetus gallicus) and common kestrel (Falco tinnunculus). According to the Red List of Andalusian Threatened Vertebrates (Junta de Andalucı́a 2001), golden eagle and Bonelli’s eagle are classified as vulnerable. Passerines

are abundant in the study area, especially thekla lark (Galerida theklae), chaffinch (Fringilla coelebs), rock bunting (Emberiza cia) and stonechat (Saxicola torquata).

Production of energy

Wind farms represent a disturbance factor for birds as well as a new collision risk.

However, both aspects increase when the rotors are working. The ‘‘Sierra de Aguas’’ wind

farm did not produce energy for 1,371 h (Table 1) during the study period, implying that

the wind turbines were in motion during 92.2% of the period.

Parameters analysed

The study covered two annual periods: from March 2005 to February 2006 and from March

2006 to February 2007, and the following seasonal periods: spring (March–May), summer

(June–August), autumn (September–November) and winter (December–February).

We visited the ‘‘Sierra de Aguas’’ wind farm on a weekly basis to analyse three different

parameters as follows:

Abundance of birds and flight behaviour

The flight behaviour of birds in the proximity of wind farm (150 m high and 500 m wide

on each side of the two rows of the wind turbines) was recorded by direct observation from

fixed points. The birds were observed for a total of 209 h during the study period. Each

observation included recording the bird species, number of birds, climatic conditions (wind

direction and speed, both recorded at 9:00 hours), flight altitude (1––below blade; 2––level

with the blade; 3––above the blade), flight class regarding blade (1––within blade radius;

2––outside blade radius), change of flight direction (changes in flight altitude and direc-

tion), activity of wind turbines (1––in motion; 2––still) and flight direction relative to

turbine alignment (1––transversal; 2––parallel).

The data collected were used to determine the monthly number of observations and

individuals, and the monthly rates of observations (number of observations/hour) and

flying (number of birds/hour). In the first case, we recorded the number of observations and

birds each month, whereas in the second case, we divided the number of observations and

birds each month by the number of observation-hours in the corresponding month.

We used the Kruskal–Wallis test to analyse annual and seasonal differences in the number

of observations and individuals, and observations and flying rates (Sokal and Rohlf 1981).

3746 Biodivers Conserv (2009) 18:3743–3758

123

We analysed the distribution of observations according to the following:

1. Altitude. We distinguished three different altitudes: 1––below blade; 2––level with the

blade; 3––above the blade.

2. Wind speed. We distinguished five different wind speeds: 1––very weak; 2––weak;

3––moderate; 4––strong; 5––very strong.

3. Combined altitude and wind speed.

4. Flight direction. We distinguished two different directions relative to wind turbine

alignment: 1––transversal; 2––parallel.

5. Combined altitude and flight direction.

In all cases the v2 (chi-square) test was used to determine if the results obtained regarding the distribution of observations significantly differed from the values expected at

random (Sokal and Rohlf 1981).

Bird density

To determine the effect of the wind farm on the density of resident birds, we selected one

species and estimated its density during spring in 2005 and 2006. We chose the thekla lark

Table 1 Number of hours with- out production of energy in the wind farm ‘‘Sierra de Aguas’’ during the period of study

Year Month Hours without production of energy

Month To origin

2005 March 69 69

April 22 91

May 53 144

June 15 159

July 23 182

August 48 230

September 30 260

October 90 350

November 64 414

December 46 460

2006 January 56 516

February 62 578

March 83 661

April 99 760

May 36 796

June 76 872

July 64 936

August 62 998

September 25 1,023

October 52 1,075

November 56 1,131

December 103 1,234

2007 January 98 1,332

February 39 1,371

Biodivers Conserv (2009) 18:3743–3758 3747

123

(Galerida theklae) because it is abundant and easy to detect. For each survey period (year), line-transect density estimation (Buckland et al. 2001) was obtained using the distance 5.0

Windows-based software package (Thomas et al. 2006), fitting the best detection model

according to Akaike’s information criterion (AIC) (Akaike 1973). The default analysis

consisted in individuals as sighting units. These were grouped into 25-m strips, and

truncated at 75–100 m (when there were no sightings beyond this distance). Each year, two

transects were walked in the morning at a mean velocity of 1.5 km/h.

The Mann–Whitney U test (Sokal and Rohlf 1981) was used to analyse statistically significant differences between the number of individuals observed in the censuses.

Bird mortality

During the study period (from March 2005 to February 2007), we searched for birds which

had been killed by colliding with the wind turbines. A total of 186 h was devoted to this

task. We used a search protocol very similar to that used by other authors (Orloff and

Flanery 1992; Morrison and Sinclair 1998; de Lucas et al. 2004). Thus, we checked 70 m

around all the wind turbines on foot and when a carcass was found we recorded the species,

sex, age, injuries, condition of plumage, date and time of find, distance and orientation to

the closest wind turbine, and estimated time of death. However, we could not take into

account the proportion of corpses being overlooked or removed by scavengers, because the

proposed trial was not approved by the environmental authorities.

Observations of raptor abundance and flight behaviour were compared with data

included in an unpublished report about this group of birds for the period November 2000–

October 2001, which were used as baselines prior to construction of the wind farm. We

used the G-test of independence (Sokal and Rohlf 1981) to analyse the differences in the

abundance of the common kestrel in comparison to other raptor species prior and after the

construction of the wind farm.

Results

Abundance of birds and flight behaviour

In the study period, we recorded 1,031 observations and 1,465 individuals in the proximity

of the wind farm (Table 2). The most abundant birds were Passerines and about eight out

of every ten observations and individuals sighted were species in this group, whereas the

remaining 20% were evenly distributed between raptors and non passerines.

Without taking into account non-identified passerines, the most abundant species were

thekla lark (26.2% of individuals recorded), rock bunting (10.2%), common kestrel (7.6%),

stonechat (6.2%), swift sp. (5.3%), goldfinch (5.1%) and crag martin (5.0%).

The monthly number of observations and individuals is shown in Table 3. No statistical

differences were detected in the annual and seasonal number of observations (annual:

Kruskal–Wallis test, v2 = 1.143, df = 1, nonsignificant; seasonal: Kruskal–Wallis test, v2 = 9.199, df = 7, nonsignificant) and individuals (annual: Kruskal–Wallis test, v2 = 0.368, df = 1, nonsignificant; seasonal: Kruskal–Wallis test, v2 = 6.513, df = 7, nonsignificant).

The monthly observation rate was not constant over the study period. It fluctuated

between a minimum value of 2.4 observations/hour in December 2005 and a maximum

value of 10.7 observations/hour in June 2006 (Table 3), with a mean monthly value of

3748 Biodivers Conserv (2009) 18:3743–3758

123

Table 2 Number of observations and individuals recorded in the wind farm throughout period of study

Species No. of observations % Observations No. of individuals % Individuals

Raptors

Honey buzzard 1 11.3 1 9.4

Griffon vulture 2 2

Short-toed eagle 4 7

Sparrow hawk 5 5

Booted eagle 4 5

Common kestrel 95 111

Bonelli’s eagle 1 2

Golden eagle 1 1

Montagu’s harrier 1 1

Black kite 2 2

Passerines

Crag martin 23 83.8 73 80.1

Meadow pipit 3 9

Black redstart 2 2

Spotless starling 1 1

Thekla lark 348 384

Linnet 32 66

Melodious warbler 2 2

Thrush sp. 1 3

Red-rumped swallow 2 2

House martin 12 22

Nightingale 1 1

Stonechat 68 91

Dartford warbler 15 16

Sardinian warbler 15 22

Spotted flycatcher 2 2

Chaffinch 58 72

Goldfinch 9 75

Crossbill 12 12

Rock bunting 115 150

Non identified 143 169

Non passerines

Alpine swift 1 4.9 2 10.5

Hoopoe 1 1

Woodpigeon 1 1

Bee-eater 5 17

Swift sp. 19 77

Red-legged partridge 24 56

Total 1,031 1,465

Biodivers Conserv (2009) 18:3743–3758 3749

123

4.9 ± 2.0 observations/hour. In general, the observation rates were higher in the second

year than in the first, although no statistically significant differences were found (Kruskal–

Wallis test, v2 = 0.653, df = 1, nonsignificant). The highest observation rates occurred in spring, but there were no statistically significant seasonal differences (Kruskal–Wallis test,

v2 = 9.653, df = 7, nonsignificant). The total monthly flying rate showed considerable variations too and varied between a

maximum value of 20.6 birds/h in June 2006 and a minimum value of 3.5 birds/hour in

August of the same year (Table 3) with a mean monthly value of 7.0 ± 3.5 birds/h. The

highest total flying rates occurred during the second year in spring although no statistical

differences were detected (annual: Kruskal–Wallis test, v2 = 1.080, df = 1, nonsignifi- cant; seasonal: Kruskal–Wallis test, v2 = 8.200, df = 7, nonsignificant).

Annual flying rate of raptors prior to the construction of the wind farm was 1.1 raptors/

hour. This rate diminished slightly during the first year after the construction of the wind

farm (0.93 raptors/h; Kruskal–Wallis test previous study vs first year of wind farm,

v2 = 0.429, df = 1, P = 0.513) and more sharply and significantly during the second year (0.39 raptors/h; Kruskal–Wallis test previous study vs second year of wind farm:

v2 = 3.857, df = 1, P \ 0.05). During the two years of existence of the wind farm raptors contributed very little to total flying rate values. More specifically, the maximum value for

Table 3 Monthly abundance of observations and individuals and monthly variation of the observation and flying rates

Year Month Observations Individuals Observations/h Total birds/h Raptors/h Non raptors/h

2005 March 41 50 4.3 5.3 0.3 4.9

April 86 96 8.1 9.0 0.0 9.0

May 54 69 5.9 7.6 0.3 7.3

June 46 91 4.2 8.3 1.1 7.2

July 25 47 2.8 5.3 3.1 2.1

August 41 55 5.0 6.7 2.8 3.9

September 43 46 4.1 4.4 1.6 2.8

October 17 18 4.1 4.3 1.7 2.6

November 26 43 2.9 4.9 0.1 4.8

December 21 36 2.4 4.2 0.0 4.2

2006 January 27 45 4.2 7.0 0.0 7.0

February 47 54 5.9 6.8 0.1 6.6

March 58 68 7.0 8.2 0.2 7.9

April 74 100 8.5 11.4 0.7 10.7

May 64 75 6.3 7.4 0.0 7.4

June 67 129 10.7 20.6 0.2 20.5

July 32 36 4.0 4.5 0.9 3.6

August 32 42 2.7 3.5 0.4 3.1

September 34 40 4.3 5.0 0.9 4.1

October 55 101 4.6 8.4 0.8 7.7

November 21 28 3.5 4.7 0.0 4.7

December 58 82 5.8 8.2 0.6 7.6

2007 January 20 34 3.3 5.7 0.0 5.7

February 42 80 4.2 8.0 0.0 8.0

3750 Biodivers Conserv (2009) 18:3743–3758

123

this group was 3.1 raptors/hour in July 2005, with a mean monthly value of 0.7 ± 0.9

raptors/h. The highest values were detected in summer and the lowest in winter (Table 3)

but, despite this, there were no statistically significant seasonal differences (Kruskal–

Wallis test, v2 = 12.463, df = 7, nonsignificant). The monthly flying rates of nonraptors fluctuated between 20.5 birds/h in June 2006 and

2.1 birds/h in July 2005, with a monthly mean value of 6.4 ± 3.8 birds/hour. The lowest

flying rates occurred during the first year in spring-summer (Table 3), although the dif-

ferences were not statistically significant (annual: Kruskal–Wallis test, v2 = 2.803, df = 1, nonsignificant; seasonal: Kruskal–Wallis test, v2 = 10.667, df = 7, nonsignificant).

Table 4 shows the distribution of the observations grouped by raptors, passerines and

nonpasserines according to altitude. In all cases the most frequent observations occurred

below the blades (*80%). The second-most frequent observations involved raptors, although at lower percentages, and took place at the level of the blades (12.1%). These

results differ from those found prior to the existence of the wind farm, when the most

frequent observations for raptors occurred at the level of the blades (61.2%). For passerines

and nonpasserines the proportions of observations at the level of the blades and above

blades was very similar. We found statistically significant differences between the distri-

bution of observations according to altitude versus the values expected at random (raptors:

v2 = 128.34; df = 2; P \ 0.01; passerines: v2 = 959.47; df = 2; P \ 0.01; nonpasse- rines: v2 = 46.71; df = 2; P \ 0.01).

Most observations of raptors occurred in strong winds (43.1%) and there were fewer in

very weak winds (34.5%). However, the situation was the reverse in passerines and

nonpasserines. Thus, observations were more frequent in very weak winds, 66.8 and 66.7,

respectively, and less frequent in strong winds, 17.9 and 13.9%, respectively (Table 5). In

all cases, there were fewer observations when there was a very strong wind. In all three

bird groups the distribution of the observations according to wind speed significantly

differed from the values expected at random (raptors: v2 = 73.05; df = 4; P \ 0.01; passerines: v2 = 1,248.25; df = 4; P \ 0.01; nonpasserines: v2 = 70.67; df = 4; P \ 0.01). Prior to the construction of the wind farm flight rate was reported to be related to wind speed only in winter.

Table 4 Distribution of the observations according to altitude. 1––below blades; 2––same level that blades; 3––beyond blades

Raptors Passerines Non passerines

1 2 3 1 2 3 1 2 3

N 96 14 6 717 85 62 40 5 6

% 82.8 12.1 5.2 83.0 9.8 7.2 78.4 9.8 11.8

Table 5 Distribution of the observations according to wind speed. 1––very weak; 2––weak; 3––moderate; 4––strong; 5––very strong

Raptors Passerines Non passerines

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

N 40 15 6 50 5 577 81 16 155 35 34 4 4 7 2

% 34.5 12.9 5.2 43.1 4.3 66.8 9.4 1.9 17.9 4.1 66.7 7.8 7.8 13.7 3.9

Biodivers Conserv (2009) 18:3743–3758 3751

123

Independently of wind speed, the most frequent observations occurred below the blades

(Table 6); 82.8% for raptors, 83.0% for passerines and 78.4% for nonpasserines. In all

three cases there were no statistically significant differences in the distribution of the

observations versus the values expected at random (raptors: v2 = 3.38; df = 15; nonsig- nificant; passerines: v2 = 13.33; df = 7; nonsignificant; nonpasserines: v2 = 7.01; df = 7; nonsignificant).

When we analyzed the distribution of the observations relative to flight direction

(transversal and parallel), statistically significant differences were found (v2 = 46.52; df = 1; P \ 0.01). The most frequent observations occurred in the transversal direction relative to the wind turbines (60.6%) (Table 7). This tendency is mainly due to passerines,

as the most frequent observations for this group occurred in the transversal direction

(66.9%) and the differences were statistically significant versus the values expected at

random (v2 = 98.69; df = 1; P \ 0.01). However, for raptors and nonpasserines, the most

Table 6 Distributions of the observations according to altitude and wind speed. 1––below blades; 2––same level that blades; 3––beyond blades

Altitude Wind speed Raptors Passerines Non passerines

N % N % N %

1 Very weak 33 28.4 470 54.4 23 45.1

Weak 11 9.5 67 7.8 4 7.8

Moderate 6 5.2 15 1.7 4 7.8

Strong 41 35.3 137 15.9 7 13.7

Very strong 5 4.3 28 3.2 2 3.9

2 Very weak 5 4.3 61 7.1 5 9.8

Weak 3 2.6 6 0.7 0 0.0

Moderate 0 0.0 1 0.1 0 0.0

Strong 6 5.2 10 1.2 0 0.0

Very strong 0 0.0 7 0.8 0 0.0

3 Very weak 2 1.7 46 5.3 6 11.8

Weak 1 0.9 8 0.9 0 0.0

Moderate 0 0.0 0 0.0 0 0.0

Strong 3 2.6 8 0.9 0 0.0

Very strong 0 0.0 0 0.0 0 0.0

Table 7 Distribution of the observations according to altitude and flight direction regarding alignments of wind turbines. 1––below blades; 2––same level that blades; 3––beyond blades

Altitude Direction Raptors Passerins Non passerines

N % N % N %

1 Transversal 18 15.5 500 57.9 17 33.3

Parallel 78 67.2 217 25.1 23 45.1

2 Transversal 4 3.4 46 5.3 1 2.0

Parallel 10 8.6 39 4.5 4 7.8

3 Transversal 4 3.4 32 3.7 3 5.9

Parallel 2 1.7 30 3.5 3 5.9

3752 Biodivers Conserv (2009) 18:3743–3758

123

numerous observations occurred parallel to wind turbine alignment (77.6% and 58.8%,

respectively), although statistically significant selection of flight direction was found for

the first group only (raptors: v2 = 35.31; df = 1; P \ 0.01; nonpasserines: v2 = 1.59; df = 1; nonsignificant). In contrast, prior to the building of the wind farm 64.5% of raptor flights were reported to occur in transversal direction.

Although most frequent raptor and nonpasserines flights occurred both parallel to the

wind turbines and below the blades (Table 7), differences were not significant (raptors:

v2 = 7.80; df = 5; nonsignificant; nonpasserines: v2 = 1.15; df = 5; nonsignificant). For passerines, the most frequent observations were recorded in the transversal direction and

below the blades (57.9%). The differences detected between the observed values and those

expected at random were statistically significant (v2 = 15.42; df = 5; P \ 0.01). During the study period, we recorded only seven observations where birds avoided

crossing the wind turbine rows. In all cases, these were passerines. In the same period, we

recorded 51 observations in the transversal direction and at the same altitude as blades.

Only 27.5% of these occurred within the blade radius. Therefore, and taking into account

that we recorded 1,031 observations during the 2-year study period, the observations with a

high risk of collision represented a very small percentage (1.4%), that is, about one in

every hundred.

When we analyze the results for raptors only, the percentage of observations with a high

risk of collision is very similar to that obtained for all birds. Thus, we recorded 116

observations of raptors and only two occurred in the transversal direction and at the same

altitude as the blades (1.7%).

The particular case of raptors

The number of raptors recorded decreased 38.2 and 71.7%, respectively, during the first

and second year after the construction of the wind farm (previous study vs first year of

wind farm: v2 = 13.67, df = 1, P \ 0.01; previous study vs second year of wind farm: v2 = 30.46, df = 1, P \ 0.01). Differences in the number of raptors between first and second year of wind farm operation were statistically significant too (first year of wind

farm vs second year of wind farm: v2 = 18.99, df = 1, P \ 0.01). When we exclude the common kestrel from the reckoning, the number of raptors

observed during the first and second year of the wind farm decreased more strongly, 86.4%

and 88.3, respectively, with respect to the previous report (previous study vs first year of

wind farm: v2 = 67.70, df = 1, P \ 0.01; previous study vs second year of wind farm: v2 = 72.00, df = 1, P \ 0.01). The first and second year the number of raptors remained similarly low (first year of wind farm vs second year of wind farm: v2 = 0.15, df = 1, nonsignificant).

Before the construction of the wind farm the common kestrel was the most frequent

raptor observed. Out of 152 raptors observed, 49 individuals (32.2%) belonged to this

species. In the present study the common kestrel was the most frequent raptor observed

too, but the proportion of this species with respect to the total number of raptors registered

was higher than that reported in the previous study. To be exact, 85.1 and 75.1% of the

raptors observed were common kestrels in the first and second year, respectively. Number

of common kestrel observed increased significantly during first year of the wind farm (80

individuals; previous study vs first year of wind farm: v2 = 7.45, df = 1, P \ 0.01). On the contrary, the second year of the wind farm the number of common kestrel decreased

significantly with respect to the previous report (32 individuals; previous study vs second

year of wind farm: v2 = 4.05, df = 1, P \ 0.05). The seasonal distribution of the

Biodivers Conserv (2009) 18:3743–3758 3753

123

observations of common kestrel during the study period showed a decreasing trend

(Fig. 2). Differences between both years were statistically significant (first year of wind

farm vs second year of wind farm: v2 = 21.63, df = 1, P \ 0.01). Nevertheless, and despite its constant presence, mainly during the first year of the wind farm, there were very

few observations with a high risk of collision. Only 17.9% of the observations of this

species occurred in the transversal direction and 82.4% of these occurred across the cor-

ridor between the two rows. In only three cases, common kestrels passed through the rows

at the same altitude as the blades, and in one of them the kestrels flew outside blade radius.

Differences in the abundance of the common kestrel in comparison to other raptor

species between before and after construction of wind farm were statistically significant

(previous study vs first year of wind farm: independence G-test, v2 = 65.09, df = 1, P \ 0.001; previous study vs second year of wind farm: independence G-test, v2 = 22.01, df = 1, P \ 0.001).

Bird density

The results of the censuses of thekla lark in each year of the study period demonstrated

7.43 individuals/10 ha and 7.30 individuals/10 ha, respectively (Table 8). The densities

recorded in spring 2005 are slightly higher than these obtained in spring 2006. However,

no statistically significant differences were detected (Mann–Whitney U4 = 0.500; nonsignificant).

Bird mortality

During the entire study period (24 months), we found the remains of one bird only: some

feathers of an adult male common kestrel. There were obvious signs that it had been eaten

by a mammal. In addition, we found carnivore excrement next to the feathers. The remains

were located near a wind turbine and we think that its death was due to collision with a

wind turbine. However, we can be certain of nothing in this respect. If we assume that

collision was the cause of death, then the mortality rate in the wind farm would be 0.03

birds/turbine/year.

Fig. 2 Distribution of observations of common kestrel (Falco tinnunculus) in the wind farm ‘‘Sierra de Aguas’’ along period of study

3754 Biodivers Conserv (2009) 18:3743–3758

123

The low number of dead birds prevented us from comparing different areas in the wind

farm or establishing a chronological pattern between collisions and species phenology.

Discussion

The effects of a wind farm on birds depend on a wide variety of factors such as the

characteristics and location of the wind farm, topography, habitat of the surrounding land,

and species present (Percival 2000; Barrios and Rodrı́guez 2004; de Lucas et al. 2004,

2008; Hoover and Morrison 2005). Based on the results of the present study, the presence

and operation of wind turbines have a clearly negative effect on the raptors present in the

region. The effect is different on the common kestrel, which decline in number during the

second year of operation, than on the other raptor species, which decline sharply the first

year and remain in low numbers during the second one.

In general, most research, shows that a negligible number of birds are disturbed by wind

farms (de Lucas et al. 2004, 2005; Walker et al. 2005; Madders and Whitfield 2006 and

references therein). Only our results obtained for Passerines are in accordance with that

assertion. The monthly number of Passerine observations and individuals does not sig-

nificantly differ by year and season. Similarly, neither did the observation and flying rates

show significant variations over the study period. We did not observe passerine birds

avoiding flying close to the wind farm. In the case of the thekla lark, the densities in the

wind farm were similar to those detected in other habitats and regions within its range in

the Iberian Peninsula (Dı́az 2003), although in the second year of the study period the

values recorded were slightly lower than those of the first year.

There is some evidence in the scientific literature that the longer the period the turbines

are in operation, the greater the decrease in bird abundance (Stewart et al. 2004). In

contrast, Smallwood and Thelander (2004) observed increasing raptor presence at a wind

farm over the long term. This fact suggests that any earlier displacement effect no longer

occurs and that the birds have probably become used to the wind farms. In our case, only

the common kestrel decreased significantly during the two years of farm operation, while

the other raptor and Passerine species remained steady. However, 2 years is a very short

period to obtain accurate data on the effects of wind farms on bird abundance. Like

Madders and Whitfield (2006), we think that post-construction impact studies should be

conducted over a longer period.

We verified that the most frequent observations, independently of wind speed, occur

below the blades and, therefore, outside their radius. Nevertheless, for raptors we observed

an important change in this aspect probably associated to the presence of wind turbines.

The fact is that the most frequent observations for this group in the post-construction

period occur below the blades while in the previous study occurred at the level of the

blades. This fact shows that birds detect the presence of wind turbines and regulate its fly to

Table 8 Densities of thecla lark (Galerida theklae) in the wind farm ‘‘Sierra de Aguas’’ along 2 years of study

Year Density (individuals/10 ha) AIC LC UC CV

2005 7.43 20.24 0.252 2.187 0.502

2006 7.30 10.37 0.288 1.849 0.408

AIC: Akaike’s information criterion. LC and UC: Confidence intervals. CV: coefficient of variation

Biodivers Conserv (2009) 18:3743–3758 3755

123

the new conditions (de Lucas et al. 2007). Thus, the observations with a risk of collision

and collision rates are very few. Specifically, the mortality rate in the ‘‘Sierra de Aguas’’

wind farm is 0.03 birds/turbine/year. This result is in agreement with most studies on

collisions caused by wind turbines, which have recorded low levels of mortality (Percival

2000; Erickson et al. 2001; de Lucas et al. 2004). However, there are some other issues to

address in this respect. First, in some cases, as in this study, mortality rates may be

underestimated because the adjustments due to corpses being overlooked or removed by

scavengers cannot be made (Langston and Pullan 2003).

Second, the percentage of dead birds resulting from collisions with wind turbines is

much smaller than that due to other human structures, although we have to take into

account that, currently, there are far fewer wind farms than these other disturbance factors

caused by human activity. Such low levels of additional mortality may in fact be significant

for certain species, such as raptors, with low reproduction and slow maturation rates,

particularly in situations where cumulative mortality occurs due to the considerable pro-

liferation in the number of wind farms.

Third, although the mortality rate per turbine was low, overall collision rates may be

very high when the number of wind turbines involved is large (Orloff and Flannery 1992).

Fourth, the low levels of mortality caused by wind turbines reported by most studies on

collisions may reflect the fact that wind farms are located to avoid areas where bird

concentrations are high (Drewitt and Langston 2006). Thus, when wind farms are located

in areas used by birds for hunting, roosting or breeding, the mortality rates increase. This is

the case reported by Hunt (1999) for golden eagle in the Altamont Pass where a minimum

number of 75 individuals are killed annually. This is why positioning new wind farms in

locations with low densities of birds is the best way to minimize mortality rates due to

collisions (Erickson et al. 2001; de Lucas et al. 2004).

Fifth, the low levels of mortality caused by wind turbines in the present study can be

due, at least partly, to the decline in the number of raptors flying around the wind farm.

Presence of wind farm did not cause an increase in raptor mortality rate because raptors

used the space around the wind farm with lower frequency. This fact represents a loss of

habitat for these species.

As Devereux et al. (2008), the results of the present study show that the ‘‘Sierra de

Aguas’’ wind farm does not act as a barrier to passerines (Table 7). However, this is less

evident in the case of raptors and nonpasserines. Thus, the most numerous observations for

both groups occur in the direction parallel to wind turbine alignment. Furthermore, in the

case of raptors the change is obvious because in the previous period the most frequent

observations occurred in transversal direction. According to Drewitt and Langston (2006),

this may be due to the short distance between wind turbines (90 m). In any case, the

potential barrier effect may have been lessened due to the rows being short (Dirksen et al.

1998) and due to the corridor between them. Longer wind turbine rows could effectively

block a flight line used between hunting areas or hunting and nesting areas. The barrier

effect could also increase in areas where several wind farms cumulatively interact to create

an extensive barrier. In both cases, there is an increase in energy expenditure when birds

have to fly further to avoid wind turbines and due to potential disruptions of foraging,

roosting and breeding areas. In any case, as Drewitt and Langston (2006) suggest addi-

tional research before and after the construction of wind farms is needed to better

understand the effects of such structures on bird populations.

Acknowledgments We thank Gonzalo Zubieta, the property developer of the ‘‘Sierra de Aguas’’ wind farm, for allowing us to use and publish the information contained in the present study. This work was

3756 Biodivers Conserv (2009) 18:3743–3758

123

conducted as part of the project CGL2006-09567/BOS (I ? D project) funded by the Ministerio de Edu- cación y Ciencia (Spain), financed jointly by the FEDER; and the project P05-RNM-00935, funded by the Consejerı́a de Innovación, Ciencia y Empresa (Junta de Andalucı́a, Spain).

References

Akaike H (1973) Information theory and an extension of the maximum likelihood principle. In: Petrov BN, Csaki F (eds) Proceedings of the second international symposium on information theory. Akademiai Kiado, Budapest, Hungary, pp 267–281

Alonso JA, Alonso JC (1999) Mitigation of bird collisions with transmisión lines through groundwire marking. In: Ferrer M, Janss GFE (eds) Birds and power lines: collision electrocution and breeding. Quercus, Madrid, pp 121–132

American Wind Energy Association (AWEA) (1995) Avian interactions with wind energy facilities: a summary. Prepared by Colson & Associates for AWEA, Washington

Barrios L, Rodrı́guez A (2004) Behavioural and environmental correlates of soaring-bird mortality at on- shore wind turbines. J Appl Ecol 41:72–81

Bernis F (1980) La migración de las aves en el Estrecho de Gibraltar. Universidad Complutense de Madrid, Madrid

Bildstein K, Zalles JL (eds) (2000) Raptor watch: a global directory of raptor migration sites. Birdlife Conservation, London

Buckland ST, Anderson DR, Burnham KP, Laake JL, Borchers DL, Thomas L (2001) Introduction to distance sampling: estimating abundance of biological populations. Oxford University Press, Oxford, p 432

de Lucas M, Janss G, Ferrer M (2004) The effects of a wind farm on birds in a migration point: the Strait of Gibraltar. Biodivers Conserv 13:395–407

de Lucas M, Janss G, Ferrer M (2005) A bird and small mammal BACI and IG design studies in a wind farm in Malpica (Spain). Biodivers Conserv 14:3289–3303

de Lucas M, Janss G, Ferrer M (2007) Wind farm effects on birds in the Strait of Gibraltar. In: de Lucas M, Janss GFE, Ferrer M (eds) Birds and wind farms: risk assessment and mitigation. Quercus, Madrid, pp 219–227

de Lucas M, Janss G, Whitfield DP, Ferrer M (2008) Collision fatality of raptors in wind farms does not depend on raptor abundance. J Appl Ecol 45:1695–1703

Devereux CL, Denny MJH, Whittingham MJ (2008) Minimal effects of wind turbines on the distribution of wintering farmland birds. J Appl Ecol 45:1698–1694

Dı́az M (2003) Cogujada montesina, Galerida theklae. En: Atlas de las Aves Reproductoras de España. In: Martı́ R, del Moral JC (eds) Dirección General de Conservación de la Naturaleza-Sociedad Española de Ornitologı́a. Madrid, pp 374–375

Dirksen S, Winden JVD, Spaans AL (1998) Nocturnal collisions risks of birds with wind turbines in tidal and semi-offshore areas. In: Ratto CF, Solari G (eds) Wind energy and landscape. Balkema, Rotter- dam, pp 99–107

Drewitt AL, Langston RHW (2006) Assessing the impacts of wind farms on birds. Ibis 148:29–42 Erickson WP, Johnson GD, Stricland MD, Young DP, Sernka KJ, Good RE (2001) Avian collisions with

wind turbines: a summary of existing studies and comparisons to other sources of avian collisions mortality in the united states. Resource Document, National Wind Coordinating Committee, Washington

Fajardo I, Pividal V, Trigo M, Jiménez M (1998) Habitat selection, activity peaks and strategies to avoid road mortality by the little owl Athene noctua. A new methodology on owls research. Alauda 66:49–60

Ferrer M, Riva MDL, Castroviejo J (1991) Electrocution of raptors on power lines in southwestern Spain. J Field Ornithol 62:181–190

Fox AD, Desholm M, Kahlert J, Christensen TK, Krag Petersen IB (2006) Information needs to support environmental impact assessments of the effects of European marine offshore wind farms on birds In: Wind, fire and water: renewable energy and birds. Ibis 148(Suppl. 1):129–144

Hansen P, Grand B, Lange N (1992) Power to spare in the upper Midwest. Izaak Walton League of America, Minneapolis, p 30

Harmata AR, Montopoli GJ, Oakleaf B, Harmata PJ, Restan M (1999) Movements and survival of Bald Eagles banded in the Greater Yellowstone ecosystem. J Wildl Manage 63:781–793

Hoover SI, Morrison ML (2005) Behavior of Red-tailed Hawks in a wind turbine development. J Wildl Manage 69:150–159

Biodivers Conserv (2009) 18:3743–3758 3757

123

International BirdLife (2004) State of the World’s birds indicators for our changing world. BirdLife International, Cambridge

Janss GFE, Ferrer M (1998) Rate of bird collision with power lines: effects of conductor-marking and static wire-marking. J Field Orniythol 69:8–17

Junta de Andalucı́a (2001) Libro rojo de los vertebrados amenazados de Andalucı́a. Consejerı́a de Medio Ambiente, Sevilla. 336 pp

Klem D (1990a) Bird injuries, cause of death, and recuperation from collision with windows. J Field Ornithol 61:115–119

Klem D (1990b) Collisions between birds and windows: mortality and prevention. J Field Ornithol 61: 120–128

Langston RHW, Pullan JD (2003) Wind farms and birds: an analysis of the effects of wind farm on birds, and guidance on environmental assessment criteria and site selection issues. Report written by Birdlife International on behalf of the Bern Convention. Council Europe Report T, PVS/inf

Larsen JK, Guillemette M (2007) Effects of wind turbines on flight behaviour of wintering common eiders: implications for habitat use and collision risk. J Appl Ecol 44:516–522

Larsen JK, Madsen J (2000) Effects of wind turbines and other physical elements on field utilization by pink-footed geese (Anser brachyrhynchus): A landscape perspective. Landscape Ecol 15:755–764

Leddy KL, Higgins KF, Naugle DE (1999) Effects of wind turbine on uplan nesting birds in conservation reserve program grasslands. Wilson Bull 111:100–104

Lucio A, Purroy FJ (1992) Caza and conservación de aves en España. Ardeola 39(2):85–98 Madders M, Whitfield DP (2006) Upland raptors and the assessment of wind farm impacts. Ibis 148:43–56 Manville A (2000) Briefing statement published on the World Wide Web dated 04/07/00. Fish and Wildlife

Service, US Morrison ML, Sinclair K (1998) Avian risk and fatality protocol national renewable energy laboratory

(NREL). Golden, Colorado Nelson HK, Curry RC (1995) Assessing avian interactions with windplant development and operation.

Translations 60th North American Wildlife and Nature Resource Conferences, pp 266–277 Orloff S, Flannery A (1992) Wind turbine effects on avian activity habitat use, and mortality in the altamont

pass and solano county wind resource areas. California Energy Commission (CEC), Sacramento, California

Osborn RG, Higgins KF, Usgaard RE, Dieter CD, Neiger RD (2000) Bird mortality associated with wind turbines at the Buffalo Ridge Wind Resource Area, Minnesota. Am Midl Nat 143:41–52

Penteriani V (1998) Birds and wind power lines. W.W.F, Tuscany Region Percival SM (2000) Birds and wind turbines in Britain. Br Wildl 12:8–15 Smallwood KS, Thelander CG (2004) Developing methods to reduce bird mortality in the almont pass wind

resource area. Final report by to the California energy comission. Public interest energy research- environmental area, Contract no. 500-01-019. Bio Resource Consultants, California

Smith JC (1985) Perching and roosting patterns of raptors on power transmission towers in Southeast Idaho and Southwest Wyoming. Raptor Res 19:135–138

Sokal RR, Rohlf FJ (1981) Biometry. The principles and practice of statistics in biological research. Freeman, New York

Stahlecker DL (1979) Raptor use of nest boxes and platforms on transmission towers. Wildl Soc Bull 7: 59–62

Stewart GB, Coles CF, Pullin AS (2004) Effects of wind turbines on bird abundance. Systematic review no. 4. Centre for Evidence-based Conservation, Birmingham

Tellerı́a JL (1986) Manual para el censo de los vertebrados terrestres. Raı́ces, Madrid, p 278 Thomas L, Laake JL, Strindberg S, Marques FFC, Buckland ST, Borchers DL, Anderson DR, Burnham KP,

Hedley SL, Pollard JH, Bishop JRB, Marques TA (2006) Distance 5.0. Release 2. Research Unit for Wildlife Population Assessment, University of St. Andrews, UK. http://www.ruwpa.st-and.ac. uk/distance/

Vargas JM (2002) Alerta cinegética. Reflexiones sobre el futuro de la caza en España. Editorial Otero, Madrid

Villafuerte RJ, Viñuela J, Blanco JC (1998) Extensive predator persecution caused by population crash in a game species: the case of red kites and rabbits in Spain. Biol Conserv 84:181–188

Walker D, McGrady M, McCluskie A, Madders M, McLeod DRA (2005) Resident Golden Eagle ranging behaviour before and after construction of a wind farm in Argyll. Scot. Birds 25:24–40

3758 Biodivers Conserv (2009) 18:3743–3758

123

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.