Factor analysis of the Zaporizhzhia wind park impact on bats based on the index of their activity and dynamics of species diversity

Based on the studies in the period of 2010-2021, there was obtained some information on the behavior of bats and their species composition at the Zaporizhia wind park in the Zaporizhia Region, Ukraine. In the study area, there was established the presence of 10 species of bats belonging to 7 genera. The most complete results were obtained in 2019, when 4 field research methods were used, 28 expedition trips were carried out and 4177 bat signals were recorded. In total, the information on 43938 bat signals was collected in 2010-2021. The total operating time of the ultrasonic detectors was 4565 hours. There were calculated indices of the voice activity of bats for three research methods by seasons, different functional zones and phases of the annual cycle of animals. Surveys on transects with the geo-position of each recorded signal made it possible to characterize the distribution of bats over the territory of the wind farm, and the use of the Kernel Density tool allowed to differentiate the wind farm site of the Zaporizhia wind park into the zones that are conditionally safe for bats and high-risk zones. There have been determined the periods of the year and time periods of night activity, in which it is advisable to use measures to minimize the negative impact. Using international and own experience, there have been proposed methods for repelling bats away from the wind turbines.


Introduction
The complexity of the instrumental study of bats against the background of the legal requirement to assess the impact of WPP on bats makes the organization and implementation of monitoring of their existence conditions in technogenic areas unconditionally urgent.
National legislation clearly declares the need for such work, but unfortunately, there is no methodological standard for collecting and analyzing information [1]. That is why researchers of bats on the territories of wind parks are forced to rely on international practices [2][3][4][5][6]. This paper proposes some methods for carrying out bat research on the territory of the Zaporizhia wind park and the results of their application.

Survey techniques and standards
The study of bats was carried out on the territory of the Zaporizhia wind park (the Melitopol District, the Zaporizhzhia Region, Ukraine) in 2010-2021 (figure 1).
The main goal of research on the territory of the Zaporizhia wind park was to study the vocal activity of bats in order to find out changes in the ecological state of animal populations, potentially occurring because of the construction and operation of the wind park. The methodology was developped according to the recommendations of "Surveillance and Monitoring Methods for European Bats Guidelines produced by the Agreement on the Conservation of Populations of European Bats (EUROBATS)" [3]. The experience of carrying out similar studies in Germany was also used [4,6].
Three types of ultrasonic detectors were used to record bat signals: Pettersson D240x (2 units), Pettersson D500x (1 unit) and LunaBat DFR-1 Pro (1 unit). Scanning of the territory was carried out using four methods.

Vantage Point Surveys (VP)
In total, 9 vantage points (VP) are located within the wind park (figure 1). A Pettersson D240x detector was used at each vantage point. The operating time of the detector was 10 min. The most complete work was carried out in 2019, when 28 expedition trips to the territory of the wind park were carried out from March 20, 2019 to November 13, 2019. The total time of work at a VP for one trip was 90 minutes, in total for a season -2430 minutes. In total, 454 echolocation calls were recorded by the detectors.

Collection of information on transects
In total, 18 transects have been allocated in the Zaporizhia wind park, which are located along the lines of installation of wind turbines. Figure 1 shows a GPS track that covers all transects. The passage of the transects was carried out at a vehicle speed of 20-25 km/h. This speed allows to reduce the number of extraneous sounds and successfully record the bat echolocation calls. To work on transects, the LunaBat DFR-1 PRO ultrasonic detector with an external microphone (ME-4x in IP65/67 plug; cable length 1.5 m; figure 2) was used. The microphone was installed on a special plate fixed to the roof of the expedition vehicle. In total, 28 expedition trips were carried out in 2019. The total number of bat echolocation calls was 694. The length of individual transects was in the range of 1.6-10.6 km, and their total length was 91.1 km. Operating time for one night was about 6 hours (about 170 hours in 2019).

Static detector surveys at ground level
At the stationary vantage point, there was used the Pettersson D500x detector, which automatically and in a wide range recorded the echolocation calls of bats in the interval 18:00-06:00 (figure 3). The detector worked 28 nights in 2019. The total operating time was 336 hours. A total of 2882 bat signals were recorded.

Static detector surveys at height
Usually, the researcher is at the ground level during surveys, and an assessment must be made for the dangerous heights of the movable wind turbine. Accordingly, studies are needed at heights of more than 30-40 m above the ground where the wind turbine is installed. In 2017, for the first time in Ukraine, a broadband automatic detector Pettersson D500x was installed on the meteorological mast of the Zaporizhia wind park at a height of 42 m (figure 4). The detector setup was the same as during the surveys on the ground. In total, 5 nights of space scanning were carried out during the period of active autumn migration on September 19-25, 2017. There was obtained information about 22 echolocation calls of bats.    In addition to the studies of 2019, we used data obtained in 2010-2021. During this period, ultrasonic detectors of three types recorded a total of 43,938 signals belonged to bats. The total operating time of the detectors was 4565 hours.

Research area zoning in terms of bat sound density
To obtain more specific information on the territorial distribution of bats in the project area, we used the Kernel Density method for point objects [7,8].
The result of applying this technique is the construction of zones of increased activity of bats, both in certain periods and for all observation seasons, on a cartographic basis. Such zoning allows to identify places of potential risk for bats and to provide appropriate recommendations on environmental management and minimization of impacts.

Methodical substantiation of the Kernel Density technique application
The technique calculates the number of point or polylinear objects per unit area (density) using the function of the kernel to fit the surface which gradually narrows to each point or polyline (figure 5).
Conceptually, a smoothed curved surface is selected (built) for each point. The value of the surface is the maximum at the location of the point and decreases with increasing distance from  Figure 5. Principles of raster image construction for points that have different values [7]. the point, reaching zero at a distance equal to the specified search radius from the point. Only a circle environment can be used. The volume below the surface is equal to the value of the Population field for a point, or a unit, if set to "NONE". The density value in each cell of the original raster is calculated by adding the values of all kernels at those points where they are superimposed on the center of the raster cell. The core function is based on the quartet core function described in B W Silverman [8].
If the value of the Population field differs from the NONE value, the value of each element determines the number of point counts. For example, a value of three causes the point to be counted three times. Values can be integers or numbers with a floating point. To characterize the density of bats in different areas of the wind farm and at different times of the year, we gave 2-kilometer parameters of the radius of impact to each point (signal).
The species identification of bats was carried out based on the recommendations of M Barataud [9].

Determination of bat activity indicators
It is a common international practice to carry out such studies to determine activity indicators in order to assess the possible impact of a project wind farm on bats objectively.
Based on the results, the activity index for all bats is determined using the following formula: Where: I x -activity indicator for the species or group of echolocation "x"; N x -number of bat flights of the species or group of echolocation "x" observed during registration at the point counting / transect (or during all registrations taken into account); T -is the total duration of the analyzed registration at the point counting / transect (or all registrations taken into account), indicated in minutes.
Such calculations make it possible to characterize the periods of minimum and maximum activity of bats for further action and measures for the management of natural environmental complexes.

Results
Sufficiently extensive field research data are needed for a qualitative assessment of the wind farm impact on bat populations. The most complete studies were in 2019, so we will give a brief description of them. Table 1 shows the results of three methods that were applied on the territory of the wind park along the passage of the Power Transmission Lines (PTL), partly in settlements and during the expedition vehicle crossings between transects and vantage points. As you can see, a total of 4177 signals were recorded. It is obvious that their maximum number was received at a stationary vantage point (2882 signals), since the detector worked all night. The study of voice activity on the transects gave us 694, and at the vantage points 454 signals.
To characterize the behaviour of bats, absolute indicators are insufficient, so there was used the activity index calculated for all three methods and shown in table 2 and figure 6. The maximum indicator in the first half of the year was in the middle of June (June 11 -31 passes/h), which falls on the time of departure of young bats after the breeding season. In the second half of the year, the highest activity of bats was recorded at the end of August (August 29 -19.8 passes/h) and indicates the beginning of autumn migration.
Long-term studies (2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017)(2018)(2019)(2020)(2021) show that in the second half of the year in August-September, the activity of bats was relatively higher than in the first half of the year. The differences with the data in 2019 are probably due to the fact that the detectors worked exclusively within the Zaporizhia wind park, which was represented by agricultural land. In previous years, the results of the analysis of bat activity also included data collected in settlements, on the coasts of the Molochnyi Estuary and the Sea of Azov. It is in these three biotopes that the foraging and migratory behaviour of bats has always been more active than in agrocenoses [6,[10][11][12].  An analysis of the bat behaviour in different periods of the life cycle is shown in table 3. As you can see, the maximum activity of animals was recorded in the period of disintegration of colonies, the beginning of autumn migrations and grouping in flocks (the second half of August), which confirms our long-term surveys. Such results will enable us to assess the wind park impact on bat populations and develop measures to minimize the negative impact.

Static detector surveys at height
Research carried out in 2017 showed the following: • bats also hunt (migrate) at a height of 42 m and above; • in total, 22 signals were recorded by detectors; • the proportion of bats using these altitude intervals in different periods of the year ranged from 1.7 to 3.4% of the total number of recorded signals.
The technical complexity of installing and maintaining the detector on the meteorological mast did not allow us to collect more extensive data, but such work is very relevant and it is mandatory for wind parks.  Notes*: N -number of passes (signals), T -total duration of the analyzed registration in minutes, I x -number of passes/h.

Species diversity of bats
6928 signals were analyzed for 5 seasons in 2010-2019, 6636 signals of them were identified to the species (genus). As we can see from the detailed characteristics of the species diversity in table 4, representatives of the Pipistrellus genus dominated in the territory of the Zaporizhia wind park. In total, 4 species of Pipistrellus were recorded, which in total amounted to 3852 signals (55.6%). Such a picture is typical for the entire Azov and Sivash areas, which is confirmed by our preliminary studies in the Azov-Black Sea region. The subdominants were Nyctalus noctula (1108 signals, 15.99%) and Vespertilio murinus (1064 signals, 15.36%). According to our data, these 2 species retain their positions throughout the region. However, they can change places depending on the season of the year and the place of research.

Discussion
Given the rather large area of the Zaporizhia wind park, it is expected that bats will use it unevenly. Thus, there is a need for zoning the entire territory and identifying places of animal concentration. To do this, we used the "Kernel Density" method. Thanks to the technical capabilities of the LunaBat DFR-1 Pro detector, each recorded signal of a bat had its own geographical coordinate. A total of 694 signals were documented. In more detail, by seasons, we gave such a characteristic in a previous article [10]. Figure 7 shows the results of such zoning for 2019. Taking into account the numerical values and the 2-kilometer radius of influence of individual points, we divided the obtained indicators into 6 gradations, which were respectively characterized by low, medium and high signal density (two gradations each).
The northern sites of the wind park fell into the high-density zone, where according to the plan, the wind turbines No. 155, 157, 158, 163, 164, 165, 166    area and the best-preserved forest belts of high bonitet are located here. For such locations, there is a high probability of bat collisions with towers and blades of WTGs, PTL, as well as receiving barotrauma [13]. Medium signal density levels were found in the following four zones. The other areas of the Zaporizhia wind park are unattractive for bats, since the level of signal density is characterized by the lowest indicators. Thus, by applying the Kernel Density method, we were able to identify zones and compile a list of the wind turbines for which seasonal environmental management efforts are needed to minimize the possible negative impact on bat populations.

Dynamics of the bat species diversity
Indirect evidence of the stability of local bat populations is the dynamics of their species composition. Table 4 shows the ratio of bat species for 5 seasons in the period 2010-2019. In all years, representatives of the Pipistrellus genus occupied a dominant position. Their total part has always been more than 50% of the number of all other species. Understanding the complexity of identification to a species in the Pipistrellus kuhlii + Pipistrellus nathusii complex, we still tend to believe that Pipistrellus kuhlii is the most numerous. In the period 2014-2017, the subdominant was Nyctalus noctula, the share of which was 12.2-24.3%. In 2010, the second most important species was Eptesicus serotinus (14.3%), and in 2019, Vespertilio murinus (21.2%). Thus, over the entire period of research, the species diversity of the Zaporizhia wind park chiropterocomplex is represented by at least 10 species. According to the Eurobat commission, [14] Pipistrellus kuhlii is in the group with a low risk of collisions with wind turbine generators, while Nyctalus noctula (16% of the whole bat complex ), has a high risk of falling under a moving rotor. A fairly effective method to reduce fatality is to change the rotation speed of wind turbines [15,16].
Most species of bats marked by the detector within the wind farm and adjacent areas are widespread and with a stable population status (table 4).
In the project area of the wind farm there are no natural formations that could be used by bats for daytime, breeding and wintering. The main places of concentration of bats are the settlements, the population of which depends on the presence of hollow trees, convenient structures for habitation, in particular, high-rise buildings, and fresh water bodies [6,10,11].
The species composition of bats in the study area is poorer than in other natural areas. Analysis of literature data and recording of signals with the help of an ultrasound detector suggests that the fauna of bats here is close to that in the North Priazovia, where Pipistrellus dominate, in particular, Pipistrellus kuhlii ) [12,17].
The species spectrum of bats is represented by widespread species that have a fairly wide area of existence. No endemic species have been registered. There are no territories unique for bats too.

Environmental management of the Zaporizhia wind park
In accordance with the calculations of the activity index of bats, we state that the most dangerous period for them is August and the first half of September (tables 2 and 3). It is during this period that measures are needed to minimize the negative impact of wind turbines on bats, as also pointed out by L Rodrigues et al. [18].
As an environmental management of the Zaporizhia wind park area, we suggest using the experience of colleagues when the start-up rotor speed (define a threshold wind speed, at which the rotors start to operate) was increased from 3.5 to 5-6 m/s [19,20]. This recommendation is based on the foraging behaviour of bats, which typically prey in wind speeds up to 3.5 m/s. At these speeds, bats are very sensitive to collisions and barotrauma (figure 8; [21]). The financial IOP Publishing doi:10.1088/1755-1315/1049/1/012069 13 losses of wind farms when the start-up speed increases to 5 m/s are minimal and are in balance with the task of conserving bats [22]. Active methods to deter bats from wind turbines may also be applicable. The Ultrasonic Acoustic Bat Deterrent System for the Wind Industry, developed by the NRG Systems Company, has shown a 32-67% reduction in bat fatality from 15 wind turbines in Illinois (USA) [23].

Conclusions
Thus, a comparison of the data obtained in 2010-2021 concerning the study of foraging behaviour and vocal activity of bats allows us to draw the following conclusions: Monitoring the state of bats in the project area both during the construction of the wind farm and its operation is mandatory for the development and implementation of measures to minimize the possible negative impact on bats.