MELANCTHON I WIND PLANT POST-CONSTRUCTION BIRD AND BAT MONITORING REPORT: File No Prepared for:

Transcription

1 POST-CONSTRUCTION BIRD AND BAT MONITORING REPORT: 2007 File No Prepared for: Canadian Hydro Developers, Inc. 34 Harvard Road Guelph, ON N1G 4V8 Prepared by: Stantec Consulting Ltd. 361 Southgate Drive Guelph ON N1G 3M5 June, 2008

2 Executive Summary The Melancthon I Wind Plant (the Plant ), a wind plant located in the south-western portion of Melancthon Township, Dufferin County and operated by Canadian Hydro Developers, Inc. ( Canadian Hydro ), began commercial operation on March 4, The results of postconstruction monitoring in 2006 indicated that the operation of the Melancthon I Wind Plant resulted in very low bird and bat mortality in A second year of the post-construction monitoring program was undertaken in 2007, primarily to evaluate bird use and disturbance effects in the study area. Carcass searches for birds and bats were conducted at each of the 45 wind turbine generators during both the spring and fall migration periods in Correction factors for scavenger removal and searcher efficiency were measured and applied to obtain an estimate of actual mortality for these periods. The 2006 and 2007 post-construction bird and bat monitoring program has demonstrated that direct mortality of bird and bats species is low at the Melancthon I Wind Plant. The 2007 avian mortality rates at the Melancthon I Wind Plant were 0.4 birds per wind turbine during spring migration and 1.0 bird per wind turbine during fall migration. The 2007 avian mortality rate (1.4 birds per turbine over a twelve-week period) is at the low end of the range observed at North American facilities (summarized by Arnett et al. [2007]). Additionally, there was no direct evidence of avoidance, no relationship of mortality to turbine lighting or distance from woodlands, and demonstration of limited bird use in the blade sweep area. During the spring migration period, direct mortality to bats was very low. The 2007 bat mortality rates at the Melancthon I Wind Plant were 0.2 bats per wind turbine during spring and 4.2 bats per wind turbine during fall. Bat fatalities were more common earlier in the fall migration period in comparison to the latter stages of migration. The majority of the bat carcasses (approximately 80%) were found in August. The results of the 2007 data for Melancthon I Wind Plant yield a bat mortality rate (4.4 bats per wind turbine across the spring and fall migrations) that is well below that of higher concern wind facilities in North America (summarized by Arnett et al. [2007]). There was no correlation between bat mortality rates and wind turbine lighting or the distance to the nearest woodlot. Thirty-two point counts in woodland and field habitat were surveyed in A t-test or one-way ANOVA analysis was conducted to assess whether the differences in breeding bird density between pre- and post-construction conditions were statistically significant. Post-construction densities were also analyzed by distance from the closest wind turbine generator. E.1

3 Executive Summary June 2008 Population densities have remained generally consistent within the wind plant between the preconstruction and post-construction observations. Further, two grassland species thought to be particularly sensitive to wind turbines (i.e., Bobolink and Savannah Sparrow) have increased in density from the pre-construction data to the post-construction data. These increases may be related to habitat changes, but nonetheless suggest that breeding bird density has not been adversely affected by the construction and operation of the wind plant. Finally, it is noted that the post-construction monitoring data support the level of potential effects predicted in the Environmental Screening Report. Additionally, based upon two years of postconstruction monitoring, no significant unanticipated effects of the wind plant on birds or bats have been observed. E.2 cs w:\active\ \reports\2007 final report\2007 m1 post-construction monitoring report_final_ doc

4 Table of Contents 1.0 INTRODUCTION PROJECT OVERVIEW STUDY OBJECTIVES METHODS MORTALITY MONITORING Scavenger Correction Searcher Efficiency USAGE MONITORING RESULTS DIRECT EFFECTS MORTALITY Scavenger Correction Searcher Efficiency Correction Percentage of Area Searched Direct Impacts Bird Mortality Direct Impacts Bat Mortality INDIRECT EFFECTS BIRD USAGE Species Density Height Analysis DISCUSSION DIRECT EFFECTS BIRDS DIRECT EFFECTS BATS INDIRECT EFFECTS BIRD USAGE CONCLUSIONS REFERENCES i i

5 Table of Contents List of Appendices Appendix A Figures Appendix B Tables Appendix C Direct Mortality Results Appendix D Breeding Bird Point Count Results Appendix E Breeding Bird Height Analysis Results List of Figures Appendix A Figure 1.1 Location of Study Area Figure 2.1 Point Count Locations List of Tables Appendix B Table 2.1 Summary of Carcass Search Dates, Times, and Weather Conditions Table 2.2 Summary of Searcher Efficiency Trials Table 2.3 Point Count Station Details Table 3.1 Results of Spring and Fall Scavenger Trials Table 3.2 List of Bird and Bat Fatalities Melancthon I Wind Plant Spring 2007 Table 3.3 List of Bird and Bat Fatalities Melancthon I Wind Plant Fall 2007 Table 3.4 Melancthon I Wind Plant Bird Density per 10 ha in Field and Crop Habitat Table 3.5 Melancthon I Wind Plant Bird Density per 10 ha in Forest Habitat Table 3.6 Melancthon I Wind Plant Species Density per 10 ha in Forest Habitat by Distance from Turbine Table 3.7 Melancthon I Wind Plant Flight Height Observations ii cs w:\active\ \reports\2007 final report\2007 m1 post-construction monitoring report_final_ doc

6 1.0 Introduction 1.1 PROJECT OVERVIEW The Melancthon I Wind Plant (the Plant ), a wind plant located in the south-western portion of Melancthon Township, Dufferin County and operated by Canadian Hydro Developers, Inc. ( Canadian Hydro ), began commercial operation on March 4, The Plant is comprised of 45 General Electric 1.5 MW, model sle wind turbine generators strategically placed over approximately 3,500 hectares in Melancthon Township (Figure 1.1, Appendix A). A pre-construction monitoring program, consisting of baseline bird and bat studies, was conducted within the Plant s siting area in The results of these baseline studies (Stantec, 2005a; 2005b; 2005c), indicated that bats and most groups of birds were not expected to experience any potential negative effects as a result of the Plant. The siting area is not in the path of a major migratory flyway, nor does it contain any topographical or other physical features that would concentrate birds or significantly elevate the wind turbine generators. Additionally, as noted in the baseline studies, breeding grassland bird species are present in the siting area. Grassland breeding birds were identified as being potentially affected through both direct mortality from the wind turbines (particularly for those with aerial flight displays) and behavioural disturbance (particularly in the form of avoidance of habitat in the vicinity of operational turbines). Therefore, as noted in Section of the Environmental Screening Report (Stantec, 2005d), a commitment was made to monitor the presence of bird carcasses at the base of the turbines for a period of one year after the turbines become operational. In fulfillment of this commitment, Canadian Hydro retained Stantec Consulting Ltd. ( Stantec ) to complete the one year post-construction monitoring works. This monitoring work was carried out in 2006, following the commercial operation date of the wind plant. The results of this monitoring program indicated that the operation of the Melancthon I Wind Plant resulted in very low bird and bat mortality in 2006 (Stantec, 2007). As part of the 2006 post-construction monitoring works, and building upon discussions with Environment Canada and Canadian Wildlife Service ( EC ), additional breeding bird point counts were undertaken in woodlands and wetlands adjacent to the wind turbine generators to supplement the 2004 pre-construction grassland bird baseline information. Although the Plant was operational in March 2006, these point counts were considered pre-construction data as it is recognized that disturbance effects to breeding birds may not become fully evident until the second year of operation (Environment Canada, 2007b). As noted in the 2006 post-construction monitoring report (Stantec, 2007), a key contributing factor to Canadian Hydro adding a second year to the post-construction monitoring plan was to further assess any potential disturbance effects. 1.1

7 Introduction June STUDY OBJECTIVES In February 2007, EC released the following two documents related to avian monitoring protocols at wind plants: Wind Turbines and Birds: A Guidance Document for Environmental Assessment (Environment Canada, 2007a) Recommended Protocols for Monitoring Impacts of Wind Turbines on Birds (Environment Canada, 2007b). Building upon these documents, the following objectives for the post-construction monitoring program were identified: to evaluate the predictions made during the Environmental Screening Process to allow for improved predictions in the future to determine whether any mitigation measures that had been proposed and implemented were effective to determine whether any significant unanticipated effects are occurring and, if so, to identify any possible mitigation measures to obtain quantitative information on the effects of the plant that can be used to understand cumulative environmental effects more accurately to inform future decisions about development or placement of wind turbine generators. In meeting these objectives, two types of post-construction monitoring studies were implemented, consistent with the EC guidance documents and plant-specific recommendations: 1. effects of wind turbine generators on bird usage in the siting area (i.e., disturbance effects) 2. bird and bat collision rates with wind turbines (i.e., direct mortality). The second year of the post construction monitoring program was undertaken in 2007 to evaluate bird use of the study area. As noted above, this was a supplementary year of postconstruction monitoring data for birds and bats beyond the monitoring commitment Canadian Hydro made in the Environmental Screening Report. The monitoring protocols used in 2007 once again followed EC s post-construction monitoring recommendations (2007b). Environment Canada also provided comments specific to the proposed 2007 post-construction work program for the Melancthon I Wind Plant. These were received after the 2007 monitoring program had started (June 15, 2007) and thus integrated to the extent practicable into the program following their receipt. 1.2 cs w:\active\ \reports\2007 final report\2007 m1 post-construction monitoring report_final_ doc

8 2.0 Methods 2.1 MORTALITY MONITORING The following terms are utilized throughout this report: fatality refers to an event resulting in a bird or bat death carcass refers to the dead body of a bird or bat mortality refers to the rate of death attributable to a certain cause, which in this case, is a wind turbine generator(s). Carcass searches for birds and bats were conducted at each of the 45 wind turbine generators during both the spring and fall migration periods. During spring migration, carcass searches were conducted at six-day intervals for a four week period; specific dates were May 10, 11, 16, 22, 28 and June 3, In the fall, carcass search frequency was increased based on the results of the spring scavenger trials. Fall carcass searches were conducted at three- or fourday intervals for an eight week period; specific dates were August 2, 9, 13, 16, 20, 23, 24, 27, 30 and September 3, 6, 10, 13,, 20, 24, and 27, The fall carcass searches were extended through the month of August to assess effects to early migrant bats. A complete summary of survey dates, times, and weather conditions is provided in Table 2.1 (Appendix B). The carcass search area at every wind turbine generator was standardized to facilitate the application of correction factors. The carcass searches consisted of one surveyor covering a 35 m 1 radius area under each turbine, walking in concentric circles spaced at five metre intervals and searching for bat and bird carcasses. If a bird or bat carcass was discovered the following information was recorded: date and time it was found state of decomposition estimated number of days since death injury sustained (or best estimate if the carcass was in poor condition) species (or best estimate if the carcass was in poor condition) distance and direction from the nearest wind turbine generator substrate where the carcass was found. 1 The 35 m radius search area was again employed in the 2007 fieldwork in order that the results could be directly compared to the 2006 fieldwork, which also used a 35 m radius search area. 2.1

9 Methods June 2008 Carcasses that were found in reasonable condition were collected for confirmation of species, if necessary, or for later use in searcher efficiency trials. Considering searcher efficiency and scavenger rates, it is likely that some bird and bat carcasses are missed during the searches. Birds and bats that strike a moving wind turbine blade may be flung away from the turbine search area, removed by scavengers, or carcasses may be overlooked due to local conditions (e.g., tall grass). Therefore, information to calculate various corrective factors for searcher efficiency and scavenger rates was also collected during the field work. There are numerous published and unpublished approaches to incorporating these corrective factors into an overall assessment of total bird and bat mortality. Currently, EC recommends the following correction formula (Environment Canada, 2008a): C = c / (Se x Sc x Ps), where C is the corrected number of bird or bat fatalities c is the number of carcasses found Se is the proportion of carcasses expected to be found by searchers (searcher efficiency) Sc is the proportion of carcasses not removed by scavengers over the search period Ps is the percent of the area searched. Although not prescribed in any guideline, EC has indicated that most birds and bats will fall within 50 m of the wind turbine generator base. This value was used to determine the percent of area searched (Ps). Methods regarding scavenger removal and searcher efficiency trials are discussed in Sections and 2.1.2, respectively. Mortality was compared between lit and unlit wind turbine generators to investigate any correlation related to aviation safety lighting. The relationship between distance to woodlots and mortality was also investigated. Consistent with the recommendations of EC (2007b), this report also presents the raw data to permit further calculations at a later date, or simple re-analysis should better statistical techniques become available in the future Scavenger Correction Scavenger trials are designed to correct for carcasses that are removed by predators before the search period. These trials involve the distribution of carcasses in known locations at each wind turbine generator, followed by periodic checking to determine the rate of removal. 2.2 cs w:\active\ \reports\2007 final report\2007 m1 post-construction monitoring report_final_ doc

10 Methods June 2008 As recommended by Environment Canada (2006b), two dead, dark coloured chicks were placed in two locations within a 35 m radius around each of the 45 wind turbine generators 2. If there was more than one substrate type (e.g., gravel, hay, crop, ploughed soil, etc.), the two chicks were placed on different substrates. UTM coordinates were taken at each chick location and the distance and direction from the wind turbine generator was measured. The location of each chick was also drawn on a turbine map with substrate compositions. During the spring trial the chicks were placed on May 10 and 11, 2007 with their presence or absence recorded at six-day intervals on May 16, 22, and 28. In the fall trial, carcass search and scavenger trial intervals were increased based on the results of the spring scavenger trial. Chicks were placed on August 9, 2007 and their presence or absence recorded at three- and four-day intervals on August 13, 16, and 20, Data from the scavenger trials were subsequently used to calculate the proportion of carcasses expected to remain over the search period. The proportion of carcasses placed at the beginning of each search period was compared to those remaining at the end. Separate scavenger corrections were calculated for spring and fall. Proportions of carcasses remaining after each 6-day interval were pooled for the spring, and each 3-day or 4-day interval for the fall, to calculate the overall scavenger correction factors: Sc = n visit1 + n visit2 + n visit3,, where n visit0 + n visit1 + n visit2 Sc is the proportion of carcasses not removed by scavengers over the search period n visit0 is the total number of carcasses placed n visit1 n visit3 are the numbers of carcasses remaining on visits 1 through Searcher Efficiency Searcher efficiency is designed to correct for carcasses that may be overlooked by surveyors during the survey periods. It is noted that search efficiency will vary for each individual based upon their own unique characteristics. Searcher efficiency trials involve a tester that places bird and bat carcasses under wind turbine generators prior to the standard carcass searches to test the searcher s detection rate. Environment Canada (2007b) provides detailed recommendations on determining searcher efficiency, expressed as a proportion of carcasses expected to be found by individual searchers. Three searchers conducted the 2007 carcass searches at the Melancthon I Wind Plant. For the purposes of this report the searchers are referred to as DA, LE, and TO. Searcher efficiency results for each searcher were pooled from both the Melancthon I site and a second wind power project in southern Ontario, as the three searchers undertook the field work at both sites and 2 Wind turbine generator number T14 was omitted during the spring scavenger trials due to the on-site presence of maintenance crews 2.3

11 Methods June 2008 both study areas had similar land-use patterns and habitat features. The searcher efficiency trials involved between carcasses for each searcher, the majority of which were bats (which are harder to detect than birds due to their physical size) (Table 2.2, Appendix B). Both studies were completed in The trials occurred randomly over the course of the carcass searches. Searchers were unaware when they were being tested in order to provide a realistic representation of their overall efficiency. The tester placed carcasses of birds and/or bats within the 35 m radius search area at least one hour prior to carcass searches. Typically one and occasionally two specimens per wind turbine generator were placed at randomly selected turbines. For each specimen that was placed, the species, time of placement, UTM coordinates, and substrates were recorded. The UTM and species were used to distinguish fatalities from carcasses placed for the searcher efficiency trials. Immediately following each testing event, the tester retrieved unfound carcasses. If the carcass was absent, but had not been found during the carcass searches, it was considered to be scavenged and was not included in searcher efficiency calculations. Searcher efficiency (Se) was calculated for each searcher as follows: Se = number of test carcasses found number of test carcasses placed 2.2 USAGE MONITORING Pre-construction surveys included point counts during the 2004 and 2006 breeding bird seasons to determine breeding bird density and provide a baseline for post-construction monitoring comparisons (Stantec, 2005a). In 2004, seventeen point counts were conducted in hayfields (planted in hay for later harvesting - 8 points), cropland (planted in row crops - 5 points), bare fields (no crops planted - 2 points), pasture (grazed by cattle - 1 point), and meadow marsh (wet area dominated by cattails - 1 point). To complement the 2004 point counts, located primarily in agricultural habitats where the wind turbine generators are located, 21 additional point counts were conducted in As requested by EC, these additional point counts focused on the patches of forest and swamp habitat within the wind plant area. To ensure an appropriate weighting of habitat types, post-construction surveys conducted in 2007 entailed repeating 11 of the point count locations surveyed in 2004 and all of those surveyed in 2006, for a total of 32 point counts. The point count survey locations for 2007 are illustrated in Figure 2.1 (Appendix A). Points were surveyed on June 12 and June 26, Handheld GPS units were used to ensure the accuracy of the locations from year to year. 2.4 cs w:\active\ \reports\2007 final report\2007 m1 post-construction monitoring report_final_ doc

12 Methods June 2008 As in the pre-construction monitoring, some of the natural communities were not wide enough to accommodate a 100 m-radius circle as recommended by EC. In such cases, the point count was located so the different habitats would evenly split the point count in two halves, to allow for simpler density calculations. In addition, birds observed in each habitat were recorded separately such that the density calculations could be made independently for both habitat types within the same point count station. The habitat at each point count and the distance from each station to the nearest wind turbine generator are detailed in Table 2.3 (Appendix B). To allow comparison of standardized species density numbers, densities per 10 ha were calculated for each point count. The average density for each species and for each species by community was also calculated. All density calculations were corrected for point counts that contained two different habitat types. The maximum single-day count of each species observed at a point count station was used in the analysis, to avoid double counting of territorial birds recorded on both counts. A t-test or one-way ANOVA analysis was conducted to assess whether the differences between pre- and post-construction monitoring results are statistically significant. Post-construction results were also analyzed by distance from the closest wind turbine generator. Species densities were calculated by habitat type for all point count stations located 150 m from a wind turbine generator and compared to densities calculated for point counts located >150 m from a wind turbine generator. This distance was selected to obtain a sufficient sample size for analysis, and was partly based on the results of Leddy et al. (1999) that indicated disturbance effects on grassland breeding birds extend to at least 80 m from turbine base. No point counts were located within 80 m of a turbine base. Currently accepted survey protocols had not yet been developed by the agencies at the time of the 2004 field season. As such, the protocols used during the 2004 field surveys differed slightly from those in 2006 and For example, during the 2006 and 2007 point counts the height of bird flight was recorded, whereas this was not included in the 2004 data. Four height regimes corresponding to the approximate dimensions of the wind turbine generators were used: on ground or below blade sweep (i.e., 0 to 35 m), at blade sweep (i.e., 35 to 125 m), above blade sweep (i.e., 125 to 200 m), and well above blade sweep (i.e., over 200 m). Additionally, the flight height analysis includes data from all 21 point counts surveyed in 2006 and 24 of the point counts completed in

13 Methods June 2008 This page left blank intentionally. 2.6 cs w:\active\ \reports\2007 final report\2007 m1 post-construction monitoring report_final_ doc

14 3.0 Results 3.1 BACKGROUND A review of recent avian mortality rates from 14 facilities across North America with modern turbines was conducted by Arnett et al. (2007). Results from these facilities were based upon standardized mortality monitoring using a systematic survey process for a minimum of one year and incorporating scavenging and searcher efficiency bias corrections. These studies yielded avian mortality rates ranging from 0.63 to 7.7 birds per wind turbine per year. At the two sites located in the eastern U.S., the bird mortality rates ranged from 4.04 to 7.7 birds per wind turbine. Arnett et al. (2007) also summarized the bat mortality rates from 22 wind facilities in North America where recent standardized mortality monitoring was conducted using a systematic survey process for a minimum of one year and incorporating scavenging and searcher efficiency corrections. The bat mortality rates ranged from 0.1 to 69.6 bats per wind turbine. Of the seven sites located in the eastern U.S., the bat mortality rates ranged from 20.8 to 69.6 bats per wind turbine. Few studies are available that identify the indirect or disturbance effects of wind power on birds. Leddy et al. (1999) determined that densities of breeding grassland birds were lower within 80 m of turbines in southwestern Minnesota, and other studies indicate that bird use is lower within 100 m of turbines than in areas further away (Arnett et al., 2007). A recent review of published and unpublished studies concluded that the extent and significance of disturbance effects to grassland birds are unknown but could range from zero to several hundred metres from a wind turbine generator (Arnett et al., 2007). 3.2 DIRECT EFFECTS MORTALITY Scavenger Correction The results of the spring and fall scavenger trials are summarized in Table 3.1 (Appendix B). By pooling the spring s 6-day search period results, the overall predicted proportion of carcasses not removed by scavengers over the search period was calculated to be 37%. By pooling the fall 3-day and 4-day search period results, the overall predicted proportion of carcasses not removed by scavengers over the search period to be 61%. 3.7

15 Results June Searcher Efficiency Correction Individual searcher efficiency ranged from 50% to 87.5% (Table 2.2, Appendix B). The overall searcher efficiency was subsequently calculated for the Melancthon I Wind Plant by weighting the individual searcher efficiencies according to the proportion of wind turbine generators surveyed by each individual over the course of the study. The overall searcher efficiencies for the spring and fall were calculated as 67.0% and 71.6%, respectively Percentage of Area Searched During the carcass searches, an area with a 35 m radius was searched around each wind turbine generator. As noted above, EC has indicated that most birds and bats will fall within 50 m of the wind turbine generator base (Environment Canada, 2008b). This distance was used to determine the percent of area searched (Ps). Accordingly, 49% of the area where the majority of carcasses are expected to land was searched. This represents a conservative correction factor as the relationship between density of carcasses and the distance from the wind turbine generator base appears to be logarithmic (Jain et al, 2007). This relationship suggests that a higher proportion of carcasses will be found in closer proximity to the wind turbine generator base Direct Effects Bird Mortality Direct mortality to birds was found to be low in both spring and fall. Only two bird carcasses were observed during the spring migration period. Correcting for searcher efficiency, scavengers, and percent of area searched, these results would represent approximately 16 bird fatalities for the entire wind plant or 0.4 bird fatalities per wind turbine during the course of the spring survey. During the fall migration period, a total of 10 bird carcasses were observed. When corrected for searcher efficiency, scavengers, and percent of area searched, these results represent approximately 47 bird fatalities for the entire wind plant or 1.0 bird fatality per turbine during the course of the fall survey. One additional carcass was found outside the survey period. A Red-tailed Hawk carcass was observed by maintenance crews on March 30, This record is included in the discussion of mortality in Section , but could not be included in the calculation of overall mortality as it was not observed as part of the controlled surveys. Raw data collected on the bird carcasses is provided in Appendix C. Six bird species were represented: Red-tailed Hawk, Red-eyed Vireo, Tree Swallow, Northern Rough-winged Swallow, Grey-cheeked Thrush, Magnolia Warbler (Table 3.2 and Table 3.3, Appendix B). All species have an Ontario General Status of Secure and are ranked as S5 (i.e., Secure Common, widespread, and abundant in the province), except for the single fatality of the Greycheeked Thrush which is ranked S3S4 (i.e., Vulnerable in Ontario due to a restricted range, relatively few populations [often 80 or fewer], recent or widespread declines or other factors making it vulnerable to extirpation). 3.8 cs w:\active\ \reports\2007 final report\2007 m1 post-construction monitoring report_final_ doc

16 Results June 2008 During the fall migration period, fatalities of Tree Swallows were more common than any other species, comprising 30% of the carcasses found. Correcting for searcher efficiency, scavengers, and percent of area searched, the number of Tree Swallow carcasses found represents approximately 14 fatalities for the entire wind plant, or 0.3 birds per wind turbine, during the fall migration period. All of the Tree Swallow fatalities occurred between August 2 and 16, 2007 and were concentrated in the southwest corner of the wind plant area. Four Red-tailed Hawk carcasses were observed during 2006 and 2007; one during the 2006 breeding season, one in March 2007 (observed by maintenance crews), one during spring 2007 and one in fall Based upon the carcass search results for 2006 and 2007, this equates to approximately 0.02 Red-tailed Hawks per wind turbine in 2006 and 0.07 Red-tailed Hawks per wind turbine in Three of the carcasses were adults and one was indeterminate, as the full carcass was not present. The four Red-tailed Hawk fatalities are considered to represent the actual mortality rate without correction factors, for the following reasons: searcher efficiency rates are higher than average for larger birds larger and heavier birds are more likely to land closer to the wind turbine generators scavenger rates are lower for larger birds as they are harder for predators to carry off. Two of the Red-tailed Hawks carcasses appeared to be several months old, one of which had been visited by scavengers with large parts of the bird left behind. The Red-tailed Hawk fatalities occurred at four different wind turbine generators (T2, T10, T37 and T45), distributed throughout the wind plant area. All four wind turbines were located in agricultural fields, at various distances from woodlands (approximately 50, 210, 240, and 265 m). Due to the small number of fatalities it was not possible to test for statistical significance. However, no clear relationship between mortality rate and distance to woodlands was observed Woodlands Of the 12 bird carcasses observed during the 2007 study period, no wind turbine generator was responsible for more than one fatality. The 12 wind turbines that had bird fatalities were distributed throughout the wind plant area with no evident geographical pattern. An analysis was conducted to look for a relationship between mortality rate and distance to a woodland. Wind turbine generators were separated into three distance regimes from woodlands to facilitate the analysis: less than 50 m 50 m to 100 m greater than 100 m 3.9

17 Results June 2008 For turbines less than 50 m, between 50 and 100 m, and greater than 100 m from a woodland, 2007 mortality rates were calculated to be 0.38, 0.25, and 0.24 bird fatalities per wind turbine generator, respectively. Due to the small number of fatalities it was not possible to test for statistical significance. However, no clear relationship between mortality rate and distance to woodlands was observed Aviation Safety Lights Fifteen of the wind turbine generators at the Methancthon I Wind Plant have flashing red aviation safety lights installed as per the requirements of Transport Canada Aviation. When comparing the bird mortality rates at wind turbines with and without aviation lights, no clear relationship was observed. Specifically, five of the bird fatalities were observed at lit wind turbine generators, or 0.33 bird fatalities per lit wind turbine, compared to 0.27 bird fatalities per unlit wind turbine. No species or group of birds appeared to be more attracted to the lit turbines versus the unlit wind turbines Direct Effects Bat Mortality During the spring migration period, direct mortality to bats was also very low. A single bat carcass was observed during the four week monitoring program in the spring. Correcting for searcher efficiency, scavengers, and percent of area searched, approximately eight bat fatalities for the facility or 0.2 bat fatalities per wind turbine generator was estimated for the spring period. In the fall bat mortality was found to be notably higher in comparison to the spring results. A total of 40 bat carcasses were observed during the eight week monitoring program. When corrected for searcher efficiency, scavengers, and percent search area approximately 187 bats or 4.2 bat fatalities per wind turbine generator occurred during the peak fall migration period. Bat fatalities were more common earlier in the fall migration period in comparison to the latter stages of migration. The majority of the bat carcasses (approximately 80%) were found in August. Only three bat carcasses were observed after the September 6, 2007 survey. Raw data collected on the bat carcasses is provided in Appendix C. Four bat species were represented: hoary, little brown, eastern red, and silver-haired bats (Table 3.2 and Table 3.3, Appendix B). Hoary bats were the most common species found representing 48% of fatalities. The next most common species was little brown bat representing 28% of fatalities. Eastern red bat and silver-haired bat were found in smaller numbers (5% and 13% respectively). All species have an Ontario General Status of Secure and are ranked either S5 (i.e., Secure Common, widespread, and abundant in the province) or S4 (i.e., Apparently Secure Uncommon but not rare) cs w:\active\ \reports\2007 final report\2007 m1 post-construction monitoring report_final_ doc

18 Results June Mortality Patterns During the fall migration period, no bat fatalities were observed at 20 of the 45 wind turbine generators. At the remaining wind turbines, one to two bat carcasses were generally found over the course of the eight week study. Wind turbines T33 and T46 had three bat carcasses and turbines T10 and T48 had four bat carcasses. Wind turbine locations are shown on Figure 2.1 (Appendix A). Overall, bat fatalities were disturbed throughout the wind plant area with no wind turbine generators responsible for significantly higher mortality than the plant average Woodlands An analysis was conducted to determine if a relationship existed between mortality rate and distance to a woodland. As with the bird analysis, wind turbine generators were separated into three distance regimes from woodlands. For turbines less than 50 m, between 50 and 100 m, and greater than 100 m from a woodland, mortality rates were calculated to be 1.00, 0.75, and 0.91 bat fatalities per wind turbine, respectively. Due to the small number of fatalities it was not possible to test for statistical significance. However, there was no strong relationship between mortality rate and distance to woodlands Aviation Safety Lights When comparing bat mortality rates at wind turbine generators with and without aviation safety lights no clear relationship was observed. Nine of the bat fatalities were observed at lit wind turbines, or 0.60 bat fatalities per lit wind turbine, compared to 32 bat fatalities, or 1.03 bat fatalities per unlit wind turbine. All of the bat fatalities at lit wind turbines were either little brown bats or hoary bats. However, given the small number of fatalities it is not possible to draw any definitive conclusions. 3.3 INDIRECT EFFECTS BIRD USAGE Species Density Fifty-two bird species were observed on point counts in 2007, 54 in 2006, and 75 in All species have an Ontario General Status of either S5 (i.e., common, widespread, and abundant in Ontario), S4 (i.e., uncommon but not rare, and apparently secure) or SE (i.e., exotic and not considered a native part of Ontario s fauna). A complete list of all birds observed and their abundance for all point counts completed in 2007 is provided in Appendix D. Across all three years (i.e., 2004, 2006, and 2007) most species were observed at densities of less than 2 pairs/10 ha. During 2007 post construction monitoring Bobolink and Savannah Sparrow were most common followed by Red-winged Blackbird with densities of 8.2, 7.2 and 3.2 pairs/10 ha, respectively, across all habitat types. Densities were calculated by habitat and results compared for pre- (2004 and 2006) and post- (2007) construction monitoring. Due to a small sample size, comparisons were not possible for point counts conducted in wetland habitat. 3.11

19 Results June Field and Crop The results of the field investigation during both pre- and post-construction monitoring estimate that in comparing the different habitats, overall bird density was highest in field habitat and that the wind plant area supports a diverse community of grassland birds. The most abundant species recorded in field and crop habitat are listed in Table 3.4. During all three years of monitoring the most commonly recorded species included Bobolink, Savannah Sparrow, Red-winged Blackbird, and Song Sparrow. Densities of some species were significantly higher in 2007 with the Bobolink, Savannah Sparrow, and Red-winged Blackbird recorded at densities of 16.3, 13.4, and 3.4 pairs/10 ha, respectively. In 2006, densities of the same three species were 3.8, 4.2, and 3.8 pairs/10 ha, respectively. A one-way ANOVA analysis of the results indicated that the difference among 2004, 2006, and 2007 densities for Bobolink and Savannah Sparrow were statistically significant at a 90% confidence interval with p values [p = probability] of 0.09 and 0.02, respectively. The average Bobolink density in 2007 was somewhat influenced by a single high count of 21 at point count A2, which likely reflects coverage of a colony of this species. The 2007 average Bobolink density, excluding this point, was 14.2 pairs/10 ha compared to 3.8 pairs/10 ha in The results for Red-winged Blackbird, American Goldfinch, Horned Lark, and Song Sparrow were not statistically significant Forest The most abundant species recorded in forest habitat are listed in Table 3.5. Red-eyed Vireo was the most common species in both 2006 and 2007 with densities of 8.5 and 5.8 pairs/10 ha, respectively. A t-test analysis of the results indicated that the difference between 2006 and 2007 densities of each species listed in Table 3.5 are not statistically significant at a 90% confidence interval. (A t-test is mathematically identical to a one-way ANOVA but can only be performed when there are two categories for comparison.) Woodlands Relationship An analysis was conducted to examine avian use of the study area in relation to distance from wind turbine generator locations. Densities in forest habitat were calculated for postconstruction point counts that were located 150 m from a wind turbine and compared to those located >150m from a wind turbine. The results for the most abundance species are listed in Table 3.6. Red-eyed Vireo was the most common species in both distance categories with a density of 6.4 pairs/ 10 ha on forest point counts 150 m from a turbine and 5.6 pairs/ 10 ha on forest point counts >150m from a wind turbine. For all species, densities were within similar ranges for both distance regimes. T-test analyses of the results indicated that the difference in density for each 3.12 cs w:\active\ \reports\2007 final report\2007 m1 post-construction monitoring report_final_ doc

20 Results June 2008 species between point counts 150 and those >150m is not statistically significant at a 90% confidence interval Height Analysis In 2007, 39 of the 44 species (i.e., 89%) recorded at the 24 point count stations were only observed below blade sweep height. Similarly, during the 2006 surveys 43 of 54 species (i.e., 80%) were only observed below blade sweep height. Table 3.7 provides a summary of the height observations of all birds observed during pre- and post- construction point counts that were observed at blade sweep height. Table 3.7 also includes data for occurrences at blade sweep, above blade sweep, and well above blade sweep heights. Full height analysis results for 2007 are provided in Appendix E. Based upon the data collected in 2006, 93% of all birds observed were on the ground or below the height of blade sweep of the wind turbine generators. This number increased to 98% of all bird observations in In 2006, six percent were observed at the height of the blade sweep and one percent was observed above the blade sweep. In 2007, only two percent were at the height of blade sweep and no birds were observed above the blade sweep. No statistically significant differences were noted between 2006 and 2007 height observations. During both years the vast majority of birds were observed below the height of wind turbine blade sweep. 3.13

21 Results June 2008 This page left blank intentionally cs w:\active\ \reports\2007 final report\2007 m1 post-construction monitoring report_final_ doc

22 4.0 Discussion 4.1 DIRECT EFFECTS BIRDS The operation of the Melancthon I Wind Plant in 2007 resulted in little direct mortality to birds. Mortality during the fall migration period was found to be slightly higher than that during the spring migration. The 2007 avian mortality rates at the Melancthon I Wind Plant were 0.4 birds per wind turbine during spring migration and 1.0 bird per wind turbine during fall migration. A review of recent avian fatality rates from 14 facilities across North America with modern turbines was conducted by Arnett et al. (2007). Results from these facilities were based upon standardized mortality monitoring using a systematic survey process for a minimum of one year and incorporating scavenging and searcher efficiency bias corrections. These studies yielded fatality rates ranging from 0.63 to 7.7 birds per wind turbine per year. The 2007 avian mortality rate at the Melancthon I Wind Plant (1.4 birds per turbine over a twelve-week period) is at the low end of this range. While still yielding a low overall mortality rate, during the fall migration period the mortality rate of Tree Swallows was estimated to be higher than other species. It is theorized that Tree Swallows may be at higher risk of collision with wind turbine generators during flocking behaviour, which coincides with their early fall migration. Overall, the estimated number of Tree Swallow mortalities is very low (i.e., 0.3 birds per turbine during the fall migration) and would result in negligible effects at the local population scale. A total of four Red-tailed Hawk fatalities were observed during 2006 and Based upon the estimated date of collision, one Red-tailed Hawk (fresh specimen found on August 9, 2007) was possibly a migrant passing through the wind plant area. The other three individuals, one in 2006 and two in 2007, were likely residents. Based upon the carcass search results for 2006 and 2007, this equates to approximately 0.02 Red-tailed Hawks per wind turbine in 2006 and 0.07 Red-tailed Hawks per wind turbine in However, due to low replacement rate, raptor populations can be slow to recover from loss of adults. Overall, the low number of bird fatalities observed over four weeks of spring migration and eight weeks of fall migration demonstrate that there were no significant unanticipated bird mortality events during the migration periods (such as multiple fatalities on a single night), locations (high numbers of fatalities at a single turbine), at lit wind turbines (aviation safety lighting), or in areas proximal to woodlands. Direct effects are consistent with those predicted in the Environmental Screening Report (Stantec, 2005d). 4.1

23 Discussion June DIRECT EFFECTS BATS The 2007 post-construction monitoring program for bats at the Melancthon I Wind Plant focused on direct effects to bats during their fall migration period. The fall monitoring occurred in late summer and early fall when bats are thought to be most at risk of turbine collisions. Johnson (2004, as cited by Ontario Ministry of Natural Resources, 2006) indicated that over 90% of bat fatalities at wind plants occur between mid-july and the end of September. The 2007 results confirm that most of the bat mortality observed at the Melancthon I Wind Plant occurred during the fall migration period in comparison to the spring migration (i.e., approximately 25 times higher than spring migration). However, the overall mortality rate of bats at the Melancthon I Wind Plant was low in 2007 at 4.4 bats per wind turbine generator over the twelve-week survey period. Tree bats (i.e., hoary, silver-haired, and eastern red bats) that migrate long distances and do not hibernate comprised approximately 65% the 2007 fatalities. This proportion is less than that observed at other wind facilities in North America, where more than 80% of fatalities can be comprised of species from this group (Johnson and Strickland, 2004; Ontario Ministry of Natural Resources, 2006). As with birds, bat mortality varies considerably by location in North America (US Government Accountability Office, 2005). Wind turbines in forested landscapes, particularly those on forested ridges such as high-profile sites in the Appalachian Mountains of West Virginia, tend to have significantly higher bat mortality rates than wind turbine generators placed in open country. Arnett et al. (2007) summarized the results of bat fatalities from 22 wind facilities in North American where recent standardized mortality monitoring was conducted using a systematic survey process for a minimum of one year and incorporating scavenging and searcher efficiency corrections. The fatality rates ranged from 0.1 to 69.6 bats per wind turbine. Of the seven sites located in the eastern U.S., the fatality rates ranged from 20.8 to 69.6 bats per wind turbine. The results of the 2007 data for Melancthon I Wind Plant yield a bat mortality rate (4.4 bats per wind turbine across the spring and fall migrations) that is well below that of higher concern wind facilities in North America. There was no correlation between bat mortality rates and wind turbine lighting or the distance to the nearest woodlot. 4.3 INDIRECT EFFECTS BIRD USAGE Point counts conducted in 2007 identified a suite of common birds using natural and agricultural habitats in the wind plant area. No unusual species or densities were detected. There is no evidence from the study results to indicate that forest or grassland birds are nesting in lower densities in the wind plant area. In fact, densities of the two most common grassland bird species observed in the wind plant area (i.e., Bobolink and Savannah Sparrow) significantly increased in 2007 compared to 2004 and This result is likely influenced by two factors. 4.2 cs w:\active\ \reports\2007 final report\2007 m1 post-construction monitoring report_final_ doc

24 Discussion June 2008 The 2007 average density of Bobolinks was influenced by a high count at a single point (Section ). As well, habitat changed at some of the point count stations between 2004 and Five of the 11 points surveyed in both 2004 and 2007 were identified as crop in 2004 and as hay in One of the 21 points surveyed in both 2006 and 2007 was identified as crop in 2006 and as hay in These changes could result in more suitable habitat for grassland bird species such as Bobolink and Savannah Sparrow in the Field and Crop habitat category. No direct evidence of avoidance is evident based on the survey results. There was no significant difference in the density of forest birds in woodlots within 50 m, between m, or more than 100 m from a wind turbine. There was no significant difference in the density of grassland birds in fields and crops within 150 m or more than 150 m of a wind turbine. Results of the height analysis indicate that very small numbers of birds are flying at turbine height. 4.3

25 Discussion June 2008 This page left blank intentionally. 4.4 cs w:\active\ \reports\2007 final report\2007 m1 post-construction monitoring report_final_ doc

26 5.0 Conclusions The 2006 and 2007 post-construction bird and bat monitoring program has demonstrated that direct mortality of bird and bats species is low at the Melancthon I Wind Plant. Additionally, there is no direct evidence of avoidance, no relationship of mortality to turbine lighting or distance from woodlands, and demonstration of limited bird use in the blade sweep area. Population densities have remained generally consistent within the wind plant between the preconstruction and post-construction observations. Further, two grassland species thought to be particularly sensitive to wind turbines (i.e., Bobolink and Savannah Sparrow) have increased in density from the pre-construction data to the post-construction data. These increases may be related to habitat changes, but nonetheless suggest that breeding bird density has not been adversely affected by the construction and operation of the wind plant. Finally, it is noted that the post-construction monitoring data support the level of potential effects predicted in the Environmental Screening Report (Stantec, 2005). Additionally, based upon two years of post-construction monitoring, no significant unanticipated effects of the wind plant have been observed on birds or bats. This report concludes the post-construction monitoring activities for the Melancthon I Wind Plant. However, following commercial operation of the Melancthon II Wind Plant, supplementary post-construction monitoring activities will be implemented to assess the potential effects of the combined wind plant. STANTEC CONSULTING LTD Valerie Wyatt, M.Sc., Senior Project Manager 5.1

27 Conclusions June 2008 This page left blank intentionally. 5.2 cs w:\active\ \reports\2007 final report\2007 m1 post-construction monitoring report_final_ doc

28 6.0 References Arnett, E. B., D. B. Inkley, D. H. Johnson, R. P. Larkin, S. Manes, A. M. Manville, J. R. Mason, M. L. Morrison, M. D. Strickland, and R. Thresher Impacts of wind energy facilities on wildlife and wildlife habitat. Wildlife Society Technical Review The Wildlife Society, Bethesda, Maryland, USA. Environment Canada. 2007a. Wind Turbines and Birds: A Guidance Document for Environmental Assessment. Prepared by the Canadian Wildlife Service. Final Report, February Environment Canada. 2007b. Recommended Protocols for Monitoring Impacts of Wind Turbines on Birds. Prepared by the Canadian Wildlife Service. Final Report, February 19, Environment Canada. 2008a. Personal Communications ( ) from Denise Fell, Environmental Assessment Officer; Environment Canada to Mark Kozak, Stantec Consulting Ltd.. January 8, Environment Canada. 2008b. Personal Communications ( ) from Charles Francis, Environment Canada to Valerie Wyatt, Stantec Consulting Ltd.. January 29, Jain, A., P. Kerlinger, R. Curry, and L. Slobodnik Annual Report for the Maple Ridge Wind Power Project. Post-construction Bird and Bat Fatality Study June 25, Johnson, G. D A review of bat impacts at wind farms in the United States. Presented at: Proceedings of the wind energy and birds/bats workshop: understanding and resolving bird and bat impacts. Washington, DC. May 18-19, (Schwartz, ed.). Prepared by RESOLVE, Inc., Washington, DC. Johnson, G. D. and M. D. Strickland An assessment of potential collision mortality of migrating Indiana bats and Virginia big-eared bats traversing between caves. NedPower Mount Storm Wind Project, Grant County, West Virginia. Western Ecosystems Technology Inc., April 14, Leddy, K. L., K. F. Higgins and D. E. Naugle Effects of wind turbines on upland nesting birds in Conservation Reserve Program grasslands. Wilson Bulletin 111(1): Ontario Ministry of Natural Resources Wind turbines and bats: bat ecology background information and literature review of impacts. December Fish and Wildlife Branch, Wildlife Section. Lands and Waters Branch, Renewable Energy Section. Peterborough, ON. 61 p. 6.1

29 References June 2008 Stantec Consulting Ltd. 2005a. Spring Migration and Breeding Bird Report, Melancthon Grey Wind Project. February, Prepared for: Canadian Hydro Developers, Inc. Stantec Consulting Ltd. 2005b. Fall Migration and Bats Report, Melancthon Grey Wind Project. February, Prepared for: Canadian Hydro Developers, Inc. Stantec Consulting Ltd. 2005c. Winter Raptor Survey, Melancthon Grey Wind Project. February, Prepared for: Canadian Hydro Developers, Inc. Stantec Consulting Ltd. 2005d. Melancthon Grey Wind Project Environmental Screening Report. February, Prepared for: Canadian Hydro Developers, Inc. Stantec Consulting Ltd Melancthon I Wind Plant Post-construction Bird and Bat Monitoring Report: March United States Government Accountability Office Wind Power: Impacts on Wildlife and Government Responsibilities for Regulating Development and Protecting Wildlife. GAO (Available at cs w:\active\ \reports\2007 final report\2007 m1 post-construction monitoring report_final_ doc

30 Appendix A Figures

31 km W:\active\ \graphics\Corel\ _Figures\ _03.cdr 1:220,000 SHELBURNE Study Area Boundary Base Map Source: Ministry of Transportation and Communications, Ontario, 1979, Dufferin and Grey, Original Scale 1:100,000. INDEX MAP OF SOUTHERN ONTARIO REVISION NO. REVISION DATE - - DESCRIPTION REVISED BY: - - Sudbury NIPISSING PROJECT NAME: PARRY SOUND Algonquin Park MELANCTHON I WIND PROJECT CLIENT NAME: PRESCOTT & RUSSELL Ottawa RENFREW MUSKOKA OTTAWACARLETON HALIBURTON DURHAM DUFFERIN YORK WELLINGTON HURON DURHAM HALTON WATERLOO WA TERLOO OXFORD Sarnia London ELGIN Detroit Lake St. Clair ESSEX BRANT NIAGARA NORFOLK HALDIMAND KENT LOCATION OF DETAIL CANADIAN HYDRO DEVELOPERS, INC. DATE INITIATED: FILENAME: FEBRUARY, _04.cdr LE VIL N FIGURE NO. 1.1 Kingston NORTHUMBERLAND Toronto Hamilton Niagara Falls Buffalo HAMILTON WENTWORTH MIDDLESEX LAMBTON S ED LE R G &E PEEL METRO TORO TORONTO PERTH FRONTENAC ON GT D IN AD VICTORIA Lake Simcoe GREY BRUCE X& NO LEN HASTINGS PETERBOROUGH SIMCOE LANARK S DA N DU RY T, AR O NG RN E M GL O ST & 100 Kilometres PROJECT LOCATION AND STUDY AREA SCALE: PROJECT NO.: 1:220,000 REV. NO. SHEET NO. 0 1 OF CHECKED BY: APPROVED: NK VW DRAWN BY: CEW

32 B7 B8 B21 B20 A4 B9 A5 B19 A6 A3 B3 B4 A7 A1 A2 A B18 A9 A11 B B5 B11 A10 B10 B12 B13 B1 B14 15 B6 W:\active\ \graphics\Corel\2008_Figures\ _03.cdr 1 B2 Turbine Locations Maintenance Shop/ Control Building 2006 Point Count Locations 2004 Point Count Locations B15 B16 Melancthon I Wind Plant TURBINE LOCATIONS AND POINT COUNT LOCATIONS _03.cdr