Progress Report
Calendar Year 2003
Principle Investigator:
Dr. Rolf Koford
Iowa Coop. Fish and Wildl. Res. Unit
Science Hall II
Iowa State University
Ames, IA 50011
Graduate Student
Aaftab Jain
Dept. of Natural Resource Ecology and Management.
Science Hall II
Iowa State University
Ames, IA 50011
Co-Investigators
Guy Zenner, Natural Resource Biologist
Alan Hancock, Natural Resource Technician
Iowa Dept. of Natural Resources
1203 N. Shore Drive
Clear Lake, IA 50428
Abstract
We examined bird and bat mortality at a new 89-turbine windfarm constructed in an environmentally sensitive area in north-central Iowa. The windfarm became operational in November 2001. It is located in cropland between three Wildlife Management Areas (WMA’s) with historically high bird use. In the past, migrant and resident waterfowl, shorebirds, raptors, and songbirds moved between the WMA’s through the area now occupied by the windfarm. Studies of bird collision mortality in California and elsewhere raised concerns about the possibility of mortalities in this area. From April 15, 2003 and December 15, 2003 we searched for dead animals under 26 randomly selected turbines. Six 76.2 m by 3.0 m transects were maintained as bare ground under each of these turbines. Access roads and construction pads under turbines were also searched. We found two birds (a yellow-throated vireo and a tree swallow) and 31 bats (hoary, red, little brown, big brown and silver-haired bats). Spring and summer search efficiency and scavenge rates were evaluated. During observer efficiency trials, observers found 77% of bird carcasses. In spring and summer and fall scavenging trials, scavengers removed 7%, 12% and 7% of carcasses respectively, within the duration of observer search cycles (i.e., 2 days). Point counts were conducted to compare bird activity in fields with and without turbines. We monitored waterfowl activity and behavior in the fall. 1.2 million total goose-use days and 194,029 total duck-use days were recorded in the WMA’s, from September 15 to December 25, 2003. Canada goose foraging behavior was monitored for a total of 270 flocks, from September 27 to December 1, 2003. Bat detectors were used to compare bat activity at turbine versus adjacent non-turbine sites. No significant differences were found between relative bird and bat activity at turbine, turbine with transect and non-turbine sites. Results presented are preliminary. The study will continue until December 2004. While we have yet to evaluate the significance of the mortality data, proper siting of these facilities remains a priority.
Introduction
Electrical generation in the Midwest has historically been accomplished through the burning of fossil fuels. Burning fossil fuels, however, can have negative environmental impacts such as increasing atmospheric CO2 emissions, which are known to contribute to global warming, or degrading air quality through release of particulate matter. Additionally, fossil fuels are a nonrenewable limited source of energy. In northern Iowa, wind power is a financially competitive alternative source of renewable energy, as it is in other parts of the U.S., that is generally viewed as having few negative effects on the environment (Erickson et al. 2001). There are, however, concerns that bird and bat mortality resulting from collisions with wind turbines could be substantial in some areas (Orloff and Flannery 1992).
Study Area:
In December 2001, construction was completed on the 89-turbine Top of Iowa Wind Farm near Joice in Worth County, Iowa. Turbines are located on private land in Sections 14, 15, 16, 21, 22, 23, 24, 25, 26, 35 and 36 of Bristol Twp. (T- 99N, R-22W) and Sections 1, 2, and 11 of Fertile Twp. (T-98N, R-22W), an area comprising about 2,137 ha (5,280 acres) (Fig. 1). Turbines are mounted on 71.6 m (235-foot) high tubular towers and turned by three 25.9 m (85-foot) blades. Blade speed at the tips is approximately 337 kmph (130 mph).
Unlike other wind farms in Iowa, the Top of Iowa Wind Farm is located in an area that historically has had very high bird use, particularly water bird use. The wind farm is situated between three large state-owned WMA’s that are complexes of wetland, grassland and forest habitat: the Rice Lake (~2,500 acres), Elk Creek (~2,500 acres) and Hanlontown Slough (~1,000 acres) WMAs (Fig. 1). The close proximity of these three large WMAs provides attractive habitat on a large scale, for north-central Iowa, and offers islands of strategic habitat. In addition to hosting numerous migrating birds during spring and fall, the Rice Lake – Elk Creek – Hanlontown Slough complex (RL-EC-HS) provides habitat for many breeding birds during the spring and summer, particularly wetland and grassland birds. In the past, migrant and resident shorebirds, rails, raptors, sparrows and icterids moved freely between the WMAs, their flight paths routinely taking them through the area that is now occupied by the wind farm. Additionally, part of the wind farm is contained within an area that has been closed to Canada goose hunting for 30 years to increase Canada goose use in the area. Finally, 2 of the adjacent WMA’s contain inviolate waterfowl refuges that attract up to 40,000 Canada geese and 20,000 ducks to the area each year, resulting in 2.5 million waterfowl-use days in the vicinity of the wind farm. There is no other existing wind farm site in Iowa with higher potential bird use.
Objectives:
1. Determine bird and bat mortality resulting from impacts with wind-generator towers and turbine blades, with emphasis on mortality during the spring and fall migration periods.
2. Determine bird and bat species composition, relative abundance, habitat use, flight patterns and the relative mortality risk at turbine sites versus non-turbine sites.
3. Determine impacts of the wind turbines on waterfowl use of croplands by comparing waterfowl use of quarter sections containing wind turbines to similar quarter sections without turbines during the fall, with emphasis on Canada goose use of the area closed to Canada goose hunting around the Rice Lake WMA.
Methods:
Researchers looked for evidence of collision-induced mortality under 26 of 89 wind towers. Six 3m-wide transects were maintained vegetation free, using herbicides and manual weeding techniques, on each 76m x 76 m search plot under each of the 26 randomly selected towers. Transects were parallel to existing corn/soybean rows. Total transect area searched under each tower was 1371.6 m2 (6 x 3m x 76.2m). This area comprised 24% of the search plot. Access roads and construction pads under turbines were also searched. Starting March 15, the turbine sites were being set up, and standardized searches of all mortality transects began April 15, 2003. Searching under each tower started immediately after transects were setup at its base. The search frequency was once every three days. From June 13 onwards, the search pattern was standardized to once every two days, in order to increase accuracy of the searches. Scavenger removal rates of collision evidence were measured during spring, summer, and fall by placing birds of 3 sizes (house sparrow, mallard/pigeon and Canada goose) on mortality transects under each of the 26 monitored wind towers. Carcasses were monitored daily for two weeks for evidence of scavenging. The status of each carcass was reported as intact, partially scavenged, or completely removed. Search efficiency trials were conducted for each observer by having an independent DNR wildlife technician place small birds, such as house sparrows, on transects without the field assistants’ knowledge. The field assistants recorded all evidence of bird or bat collisions that they discovered, including evidence planted by the independent wildlife technician. Planted evidence of collisions was later removed from the database and a search efficiency rate calculated for each field assistant.
We estimated relative avian abundance and activity using fixed radius (100m) point counts from May 1, 2003 to December 1, 2003 (Ralph et al. 1995). Starting March 15, observers standardized their point counting techniques to reduce variance between observers. Additional point count sites were added as landowner permission was obtained until by May 1, all point count sites were in place in accordance with the experimental design. Observers recorded species, activity, and location of all birds within a 100m radius of the point count site. The duration of each point count was 10 minutes. Point counts were conduced in the morning from one-half an hour after sunrise until 4.5 hours after sunrise, during mid-day from 11 AM to 2:00 PM, in the evening from 4:00 PM to 7:30 PM, and at night from half an hour before sunset to an hour and a half after sunset. These times were periodically adjusted to account for changes in daylight hours. During morning and evening point count periods, observers usually completed 10 point counts. Due to the shorter time frame for conducting mid-day and night point counts, observers usually only completed 8 point counts for those periods. We conducted point counts under wind towers without mortality transects, as well as wind towers with mortality transects, and in adjacent fields without wind towers. We also conducted point counts in crop fields in an area approximately 4 miles southwest of the windfarm to contrast bird activity within the windfarm to similar sites outside the windfarm. Each site was visited approximately once every 6 days for each of the four time periods (morning, mid-day, evening and night).
Due to difficulties in acquiring functional bat monitoring equipment (shipping delays, malfunctioning equipment) we did not begin remotely monitoring bat activity around wind turbines until September. Prior to that time, bat activity was monitored in conjunction with evening and night point counts. Anabat ultrasonic bat detectors were used to monitor bat activity from September 4, 2003 to October 9, 2003 (O’Farrell 1999). Monitors were placed at wind tower sites as well as in adjacent fields without wind towers. The bat detectors were waterproofed for passive monitoring (O’Farrell 1998) and left overnight at each site. The collected data was downloaded (in a digital format) the following morning onto a computer. Using software programs and statistical analysis, the digital information was used to estimate relative bat activity and species on the windfarm (Jolly 1997).
Waterfowl activity was monitored in the fall, from September 15 to December 25, 2003. Waterfowl use of crop fields within the area closed to Canada goose hunting around Rice Lake was estimated twice weekly. Waterfowl behavior was also observed during morning and evening foraging periods in the same area. Behavior noted was time spent foraging (stationary and mobile) versus time spent vigilant (stationary and mobile) using scan sampling techniques (Altmann 1974). A flock was scanned for 2 minutes, and each bird observed was assigned a combination of behaviors (foraging, vigilant, mobile, stationary or other). For both activity and behavior, observations were made from county roads using a vehicle and spotting scope to keep disturbance to a minimum. Cropping and/or tillage practices and turbine activity were recorded for each field during each count along with any observations of other human activity that may have influenced waterfowl use of quarter sections. Fields with and without wind turbines in the area closed to Canada goose hunting were sampled for waste grain to estimate the relative amounts of grain available for foraging waterfowl in the fields. This was done to determine if the fields with wind turbines had similar amounts of waste grain in them, i.e., were as attractive to feeding geese, as fields without wind turbines, an important consideration in goose use of these fields. Farming practices were also documented for sampled fields. Farmers whose fields were not sampled were contacted to determine the farming practices they used in 2003 to predict the attractiveness of unsampled fields to foraging geese.
In addition, numbers of waterfowl using Rice Lake, Elk Creek and Hanlontown Slough were estimated twice weekly between 10:00 AM and 2:00 PM to get a relative picture of waterfowl use of the WMA’s adjacent to the wind farm. These counts were conducted in a manner similar to past waterfowl counts, thereby providing indices to waterfowl populations that were comparable to previous years. To examine the feasibility of classifying waterfowl flight behavior relative to the wind turbines and to provide ideas for future observation methods, observations of waterfowl morning flight patterns into the wind farm area were made from October 1 through November 1. Observation data have been entered into a computer database (Excel) and/or ArcGIS and analysis is ongoing.
Results:
Between April 15, 2003, and December 15, 2003, we found two birds (a yellow-throated vireo and a tree swallow) and 31 bats (11 hoary, 9 little brown, 6 eastern red, 3 big brown and 2 silverhaired bats) on our mortality search transects. All appeared to have died as a result of collisions with the wind turbines.
The following adjustments have been made to extrapolate the mortality findings to the entire wind turbine. The mortality findings were adjusted for:
a) Proportion of the total plot area searched (76 m by 76 m plots for 89 towers)
b) Percent of test carcasses removed by scavengers within the search period
c) Percent of carcasses missed by observers in the search efficiency trials
Approximately 26% of the 76 m by 76 m plot under 26 towers was searched. Spring, summer and fall scavenging trials indicated scavengers removed 7%, 12% and 7% of carcasses, respectively, within the duration of the search frequency (2 days) in this study. Observer efficiency trials for mortality transect searches indicated that, as a group, observers found 77% of bird carcasses, although efficiency ratings varied from 40% to 95% for individual observers. Adjusted estimates of bird and bat mortality for the Top of Iowa Windfarm during the 2003 field season indicated that 10.8 birds and 167.2 bats died as a result of collisions with the wind turbines between April 15 and December 15, 2003 (Table 1).
Table 1. Adjustments for estimating total mortality at the Top of Iowa Windfarm.
| adjustment level | bats | birds |
| raw mortality findings | 31 | 2 |
| area adjusted | 119.23 | 7.69 |
| scavenge study adjusted | 128.77 | 8.31 |
| search efficiency adjusted | 167.23 | 10.79 |
Preliminary analysis of point count data using ANOVA (Analysis Of Variance, SAS Institute Inc, 2001.) indicated there was no significant difference between bird activity for any combination of the three treatments (wind turbine site, wind turbine site with transects, and crop field without wind turbine) when analyzed for all bird species combined or for the 5 most common bird species observed at the windfarm (Table 2).
Table 2. Results of bird activity comparisons between 3 sites
| category | p value |
| all species | 0.46 |
| common grackle | 0.67 |
| brown-headed cowbird | 0.79 |
| horned lark | 0.53 |
| vesper sparrow | 0.67 |
| red-winged blackbird | 0.15 |
The pilot study using Anabat ultrasonic bat detectors also indicated there was no significant difference between bat activity at wind turbine sites and crop fields without turbines (Paired Ttest, p = 0.63, SAS Institute Inc, 2001).
Fall waterfowl use observations are currently being analyzed for the 270 flocks of geese that were observed foraging in fields with and without wind turbines. Approximately 1.2 million goose-use days and 194,000 duck-use days were recorded from September 15 to December 25, 2003 for the adjacent RL-EC-HS complex (Figure 2). Waterfowl use of the adjacent WMA’s was below average for both ducks and geese in 2003 compared to historical counts due to a late summer drought in 2003. To compare goose use of fields with wind turbines to those without turbines, GIS analysis will be used to overlay goose counts in fields within the area that is closed to Canada goose hunting around Rice Lake WMA.
RECOMMENDATIONS FOR 2004 FIELD SEASON:
Mortality search methodology, including measurements of observer efficiency and scavenging rates, appeared to be satisfactory and will be continued as in 2004 as conducted in 2003. A substantial amount of staff time was needed for weeding and mowing during the 2003 field season to maintain transects vegetation-free. Transect maintenance will be reduced in 2004 by removing all crops immediately after germination and making transects more accessible for mowing (i.e., paying the landowners for some additional crop loss to maintain a 1 meter wide path to connect the transects on one end).
The pilot bat study indicated various problems with identifying bats by their calls. Several bat biologists have been contacted to provide assistance and advice for this aspect of the study.
The point counts serve to answer two questions. First, to estimate bird-use in the windfarm using on-windfarm point counts in the morning, mid-day, evening and night. Second, to contrast bird abundance on the windfarm with abundance at similar sites off the windfarm. For the later comparison, the accepted methodology is to conduct only morning point counts, since bird activity is highest at that time (Ralph et al. 1995). In 2004, only morning counts will be conducted to make comparisons of bird activity on and off the windfarm. To ensure a better comparison of bird activity on and off the windfarm, more morning counts will be conducted in 2004.
Literature Cited:
Altmann, J. 1974. Observational study of behavior: sampling methods. Behaviour. 49: 227-267
Anderson, R., Morrison, M., Sinclair, K. and Strickland, D. 1999. Studying wind energy/bird interactions: A guidance document. National Wind Coordinating Committee.
(http://www.nationalwind.org)
Bishop, R. A. 1981. Iowa’s wetlands. Proc. Iowa Acad. Sci. 88:11-16.
Byrne, S. 1983. Bird movements and collision mortality at a large horizontal axis wind turbine.
Cal-Neva Wildlife Transactions, 1983: 76-83.
California Energy Commission. 1989. Avian mortality at large wind energy facilities in California: Identification of a problem. Staff Report, Sacramento, CA.
Erickson, W.P., Johnson, G.D., Strickland, M.D., Young, Jr. D.P., Sernka, K.J. and Good, R.E. 2001. Avian collisions with wind turbines: A summary of existing studies and comparisons to other sources of avian collision mortality in the United States. National Wind Coordinating Committee. (http://www.nationalwind.org)
Jolly, S. 1997 Analysis of bat echolocation calls recorded by Anabat bat detectors. M.S. Thesis.
Department of Computer and Mathematical Sciences, Victoria University of Technology.
Melbourne, Australia.
Larsen J.K. and Madsen J. 2000. Effects of wind turbines and other physical elements on field utilization by pink-footed geese (Anser brachyrhynchus): A landscape perspective.
Landscape Ecology 15: 755-764.
Leddy K.L., Higgins K.F. and Naugle D.E. 1999. Effects of wind turbines on upland nesting birds in conservation reserve program grasslands. Wilson Bulletin, 111(1) 100-104.
O’Farrell, M. J. 1998 A passive monitoring system for Anabat II using a laptop computer. Bat Research News 39: 147-150.
O’Farrell, M.J. 1999. Qualitative identification of free-flying bats using the Anabat detector.
Journal of Mammalogy, 80(1): 11-23.
Orloff S. and Flannery. 1992. Wind turbine effects on avian activity, habitat use, and mortality in Altamont Pass and Solano County Wind Resource Areas, 1989-1991. Final Report to the California Energy Commission, Sacramento Calif. 150 pp.
Orloff, S. 1992. Tehachapi Wind Resource Area avian collision baseline study. California Energy Commission, Sacramento, CA.
Osborn R.G., Dieter C.D., Higgins K.F. and Usgaard R.E 1996. Bird flight characteristics near wind turbines in Minnesota. The American Midland Naturalist, 139: 29-38.
Osborn R.G., Higgins K.F., Usgaard R.E. and Dieter C.D. 2000. Bird mortality associated with wind turbines at the Buffalo Ridge Wind Resource Area, Minnesota. The American
Midland Naturalist, 143: 41-52.
Ralph, C. J., J. R. Sauer, and S. Droege. 1995. Monitoring bird populations by point counts. U.S. Forest Service General Technical Report PSW-GTR-149, Albany, California.
SAS System for Windows, Version 8.2. Copyright © 2001 SAS Institute Inc. SAS and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of
SAS Institute Inc., Cary, NC, USA.
Thelander, C.G. and Rugge, L. 2000. Avian risk behavior and fatalities at the Altamont Wind Resource Area. National Renewable Energy Laboratory, Golden, Co.
Winkleman, J.E. 1985. Bird impact by middle-sized wind turbines on flight behavior, victims, and disturbance. Limosa, 58: 117-121.
Fig 1. Location of the Top of Iowa Wind Farm relative to the Rice Lake, Elk Creek and Hanlontown Slough Wildlife Management Areas in northern Iowa.
Fig 2. Bi-Weekly counts of Canada geese and ducks using Rice Lake, Hanlontown Slough and Elk Creek WMAs.
