Benton Lake Refuge Shelterbelt Habitat Suitability and Bird Use Study

Service Unit: Benton Lake NWR Reporting Office: Benton Lake Complex Species or group: Upland habitat – Shelterbelts 2007

Introduction

In the recent past, planting shelterbelts was advocated in the Great Plains as a method of increasing species diversity (Schroeder 1986 and others), particularly bird diversity. Shelterbelts were planted on Benton Lake NWR initially in the 1960s, but have been maintained and expanded as recently as the 1990s. Since that time, many of the shelterbelt trees and shrubs have died, which may be the result of recent drought conditions. No attempt has been made to irrigate or replant the shelterbelts. However, during prescribed burning, the shelterbelts have been protected with firebreaks.

Since the planting of shelterbelts on the refuge, additional research and understanding of the effects of woody vegetation in the Great Plains indicates that shelterbelts may have undesirable, negative effects on grassland bird species, in particular (see Bakker 2003 for review). Grassland species of concern at Benton Lake refuge include marbled godwits, chestnut-collared longspurs, Baird’s sparrow, grasshopper sparrow, lark bunting and Sprague’s pipit. In addition, most of the species of shrubs and trees planted on the refuge are not native, and maintenance of these would likely be in direct contradiction to the Service’s biological integrity policy (601 FW 3.10).

In 2008, Benton Lake NWR will start its planning process to develop a 15 year management plan. Making management decisions about the shelterbelts on the refuge will be a key topic in this planning process. The key question will be whether or not to remove shelterbelts or maintain them. Having sufficient background data on shelterbelt condition, quality and bird use will inform future management of these shelterbelts.

The initial inventory of shelterbelts described here will address four management questions (objectives):

1) What is the current condition of shelterbelts on the refuge in terms of their contribution to bird species diversity (assuming that was the original management goal)?

2) What bird species are currently using shelterbelts on the refuge and their relative relative abundance?

3) How important are shelterbelts at BNL to species of management concern?

4) Are shelterbelts having a negative impact on grassland birds on BLNWR?

Methods

In order to assess the current condition of shelterbelts on the refuge in terms of their contribution to bird species diversity we used a habitat suitability model described by Schroeder (1986), and later revised (Schroeder et al 1992). This model identified several characteristics related to the quality of shelterbelts for wildlife, particularly species richness of birds. Although the habitat suitability model focused on species richness, the authors speculated that the same variables defined in the model are positively correlated with wildlife productivity.

Key variables in the revised Habitat Suitability Model that affect species richness in shelterbelts include shelterbelt size, number of snags/ha, average height of tallest row and foliage height diversity (MacArthur and MacArthur 1961).

The area of each shelterbelt was measured by mapping the perimeter on foot using Trimble GPS unit (sub-meter accuracy). The perimeter was defined as the maximum exterior extent of the canopy of woody vegetation. Any breaks in shelterbelts greater than 50m were mapped as separate units. The 50m cut-off is somewhat arbitrary, but based in part on studies showing that this is an appropriate scale for the effect of trees on grassland songbirds (Bakker 2003).

Snags (dbh>10cm) in each shelterbelt were simply counted. The 10cm cut-off is based on data on natural nest cavity selection by tree swallows, the cavity nesters most likely to be found on the refuge. We measured average tree height and foliar height diversity in the tallest row at 10 locations. We used a clinometer to measure the tree or shrub height at each point. Foliar height diversity was assigned at each point by indicating the number of strata (0-1m, 1-10m and >10m) in which vegetative layers reach their maximum height.

The data for each variable corresponds to a suitability index variable (SI) (Schroeder 1992). The SI values are then entered in the following equation to get an overall Habitat Suitability Index value for the shelterbelt:

HSIshelterbelt = SIfoliarheightdiversity x SIsnag x SIheight x SIsize 3

The HSI values for each shelterbelt are between 0 to 1, with 1 signifying an optimal shelterbelt for maximizing bird species diversity. Although not part of the HSI model, we also estimated the percentage of the shelterbelt that was dead and took several photographs.

In order to identify what bird species are currently using shelterbelts on the refuge, and their relative abundance, each shelterbelt was surveyed four times between May 15 and June 16. One random, 100m x 40m belt transect per ha (approximately) was placed in each shelterbelt parallel to the nearest tree row (Quamen and Naugle, In press). We counted all birds seen or heard within 20m on either side of the transect, and noted bird movements to avoid double counting. We walked transects slowly (approximately 1 km/hr). All surveys occurred between sunrise and 11:00am and starting locations were rotated among all transects. Counts were not conducted on days with precipitation or excessive wind (>20km.hr; Ralph 1993).

We used presence/absence and densities of birds within shelterbelts versus grassland areas as an indicator of negative impacts from shelterbelts. If grassland bird species occur in the grassland plots, but not the shelterbelt plots, or at higher densities in grassland plots, we will infer that these bird species are avoiding shelterbelts and the available grassland habitat for these passerine species on the refuge is reduced by the presence of shelterbelts. To do this, four grassland plots, with one 100mx40m transect each, were surveyed in the same manner and on the same days as the shelterbelt plots. Plots were selected randomly within native grasslands >800m from any shelterbelts (Quamen and Naugle, In press). The results of the four surveys of each shelterbelt and grassland transect were averaged to avoid pseudoreplication. Density differences between shelterbelts and grasslands were compared using t- tests.

Each shelterbelt was completely searched for active nests twice during the breeding season (May 23-Jun 8 and Jun 18- Jul 6). For most nests, only the bird and tree species were recorded. However, for loggerhead shrikes, the primary species of concern that has historically nested in refuge shelterbelts, nests were checked every 3-5 days until the fate was determined. Nest and egg success were calculated based on exposure days (Mayfield 1975, Johnson 1979).

Results

The Habitat Suitability Index score for refuge shelterbelts ranged from a high score of 0.43 to a low of 0.09 on a scale from 0 to 1 (Figure 2). For most of the shelterbelts, less than 25% of the existing vegetation was dead (Figure 3). For details on each shelterbelts’ condition, size and photos see Appendix 1.

Thirty-five bird species were detected in shelterbelt and grassland plots combined (Table 1). Twenty-four species were detected only in shelterbelts and two species, horned lark and clay-colored sparrow, were detected only in grasslands. Western meadowlark was the most commonly detected species (25%, n=104) in the shelterbelts and across all plots (23%, n=111). Chestnut- collared longspur was the most commonly detected bird in the grassland plots (34%, n=20). The habitat suitability model did a good job of relating shelterbelt condition to bird species richness for Benton Lake NWR (R2 =0.49, p=0.001). However, the shelterbelt next to headquarters, which scored the highest, had a strong influence on this relationship (Figure 4).

We found a total of 44 active nests in the refuge shelterbelts. The most common species’ nest was Brewer’s blackbird, followed by eastern kingbirds and loggerhead shrikes (Table 2). Shelterbelt 12 had the highest number of nests (7). We found no active nests in six of the shelterbelts. The tree species most commonly used for nests was Russian olive (Table 3). Dead trees were used for 11% of the nests we found.

We found 9 loggerhead shrike nests in 8 different shelterbelts. All nests successfully fledged young with an average of 5.78 ± 0.83 fledglings/nest. Shrikes nested in Russian olive (Elaeagnus angustifolia), both alive and dead, and caragana shrubs (Caragana arborescens). Swainson’s hawks are also a species of concern, but their nests were not monitored. They also nested in Russian olive and caragana.

There were five species that were recorded on both shelterbelt and grassland transects and had at least 10 total detections. Of these, densities of chestnut- collared longspurs were significantly higher on grassland transects (Figure 4)(p=0.001). Densities of other species were not significantly different between the two habitat types (BHCO p=0.69, SAVS p=0.69, VESP p=0.23, WEME p=0.44).

Discussion

Based on the habitat suitability ranks for refuge shelterbelts, a few shelterbelts are in moderate condition and most shelterbelts are in poor condition relative to their potential to increase bird species diversity on the refuge. Shelterbelts ranked uniformly low for each of the key factors of size, structural diversity, snags and height. Most of these factors will not improve within shelterbelts unless there is active management to plant new trees/shrubs or irrigate to increase height, size and structural diversity. Such actions would be in direct violation of the Service’s biological integrity policy (601 FW 3.10). Some of the shelterbelts with caragana shrubs have expanded naturally, increasing size and structural diversity, which would likely continue slowly if left unmanaged. Russian olive does not appear to be expanding rapidly on the refuge as it can in certain areas (Tu 2003). Since the majority of shelterbelts on the refuge are less than 25% dead, this suggests that the natural attrition, due to drought and unsuitable conditions, has already occurred and the remaining trees and shrubs are the hardy survivors. The HSI rank may increase over time as trees die, however they probably won’t become snags until the natural lifespan of these species has expired.

Although this study was only conducted for one year, the results suggest that there are several native, migratory bird species that readily use shelterbelts during the breeding season. We documented 24 species that were found only in shelterbelts (Table 1). Since we only had four grassland transects in this study, we cross-referenced this data with a four-year dataset on grassland bird point counts done on the refuge from 1994-97(Martin et al unpublished data) . Of the 24 species found in shelterbelts, but not detected in grassland transects in our study, 18 of these were never detected in grassland counts during the 1990s either. This suggests that shelterbelts are ‘important’ to these 18 species on the refuge. Of these 18 species, we found active nests in the shelterbelts for 7 of them. A review of the refuge’s historical bird records indicate that only the European starling, American crow and house sparrow would be expected to occur in shelterbelts, based on previous records and the appropriate habitat, but were not detected in this study.

Three of the species that actively nested in shelterbelts, and were not found in grasslands, are species of concern. These are the loggerhead shrike, Swainson’s hawk and mourning dove. The mourning dove is a USFWS Focal Species because of the management responsibilities the Service has for harvested populations (Dolton et al 2007, USFWS 2004). The central population is considered to be a ‘game bird above desired condition’ (USFWS 2004). However, both the loggerhead shrike and Swainson’s hawk are priority species for the USFWS due to habitat loss and population declines (USFWS 2004, Partners in Flight 2005).

Loggerhead shrikes have been monitored intermittently on the refuge in the recent past. In our survey, we found 9 nests that all successfully fledged and average of 5.78 ± 0.83 young. In past years, shrike nests were not searched for systematically, which limits the conclusions that can be drawn from the data. In 10 years, 4 of 30 nests failed. At a minimum, this suggests that shrikes are not always successful at Benton Lake. A study in Minnesota estimated an average of 5.5 fledglings/breeding pair would be needed to maintain a stable population (Brooks and Temple 1990), which is similar to our average of 5.78. However, much more detailed studies would be needed to determine if the refuge is a population source or sink. For example, loggerhead shrikes may be especially vulnerable during the period from fledging to independence, which we did not track (Yousef 1996). The data does not suggest that they prefer a certain tree species for nesting as we found nests in Russian olive, alive and dead, as well as caragana shrubs.

Swainson’s hawks have not been as closely monitored on the refuge. Nests have been monitored opportunistically which makes any conclusions difficult. Swainson’s hawk nests were not followed in this study, but 5 were found during nest searches. At least one of these nests failed. Swainson’s hawks have adapted to agricultural development and will readily nest in planted trees and hunt in nearby fields (England et al 1997). However, urban development and farm consolidation, which lead to loss of the these habitats, are thought to be a primary reason for Swainson’s hawk population declines (England et al 1997). Almost but one of the Swainson’s hawk nests on the refuge were found in Russian olive trees, the failed nest was found in a caragana shrub.

Although this study focused data collection on shelterbelts themselves, a strong body of literature suggests that there are likely negative impacts to grassland- dependent birds on the refuge as well (Bakker 2003). Negative impacts may be displacement of grassland species that avoid woody and shrubby habitat or decreased nest success due to increased presence of predators and parasitic birds that use shelterbelts.

Because chestnut-collared longspurs were found in significantly lower densities in shelterbelts than grasslands, this species’ habitat on the refuge may be reduced by the presence of shelterbelts. At least two studies have found that chestnut-collared longspurs are negatively associated with woody vegetation within a grassland patch and the percent of the grassland patch encompassed by woody vegetation (Bakker 2003). It may be possible to re-analyze the data from the grassland inventories from 1994-97 to test for effects of distance from shelterbelts on chestnut-collared longspur abundance.

Even if grassland bird species are still found near shelterbelts, their nest success may be reduced. Brown-headed cowbirds were the eighth most common bird species out of 33 detected in our shelterbelt transects. Several studies have found that shrubby and woody edge habitat can significantly increase brood parasitism of grassland species of concern by brown-headed cowbirds (Davis and Sealy 2000, Gates and Gysel 1978 and Winter et al 2000).

Nest predation has been shown to increase with the presence of edge habitats created by shelterbelts (Bakker 2003). Prior to this study, avian predators were a concern for refuge managers. However, we detected relatively few black-billed magpies on our transects (6%), and only one active nest, and we did not record any American crows or their nests. We did not investigate mammalian predators in this study. However, the refuge has an active predator removal program and trap data could be analyzed to see if more predators have been caught in shelterbelts versus other habitat types.

Distance to wooded edge has been shown in many studies to be related to both nest predation and exclusion of certain species (Bakker 2003). The distance varies by study area and species. In order to further assess the potential negative impacts of shelterbelts on the refuge, we used the most conservative distance (50m) and the most liberal (800m) to determine what percentage of refuge grassland habitat is either unavailable or less desirable for grassland species (Bakker 2003). If we buffer the shelterbelts by 50m, the original 5,790 acres of grassland habitat on the refuge is reduced by 5%. If we buffer by 800m, the effective grassland habitat is reduced by 63%.

Management Recommendations

From the results of this study, there is not a clear or obvious direction for shelterbelt management. Based on the habitat suitability model, almost all of the refuge shelterbelts are in poor condition. If we do nothing, most of the shelterbelts on the refuge may eventually disappear due to natural attrition. However, the lifespan of caragana shrubs and Russian olive is 40+ years, which gives our trees at least another 30 years (MT DNRC 2007, SD Ag Service, no date) A few of the shelterbelts have shown new growth of caragana shrubs and may persist indefinitely.

With the presence of these shelterbelts in their current condition, the effective grassland habitat on the refuge has been reduced between 5-63%. This study found that the chestnut-collared longspur may be displaced from grasslands by the presence of refuge shelterbelts. A review of the literature suggests other grassland species of concern, including grasshopper sparrows, grassland nesting ducks, marbled godwits and long-billed curlews, that nest within close proximity to the shelterbelts, may experience increased rates of nest predation and/or nest parasitism.

If we remove the shelterbelts, up to eighteen species of birds that are dependent on woody habitat may cease to be found on the refuge. Two of these are species of concern because of declining habitat, the loggerhead shrike and Swainson’s hawk. Additional analysis, consultation and discussion with outside experts would be necessary to determine if this is likely to have a significant population-level effect and if this is an acceptable consequence.

Shelterbelt removal could occur in stages, starting with the shelterbelts with the least value to woody habitat dependent species and the most benefit to grassland species. Shelterbelts 3 and 8 fit this profile. They had few species detected during surveys and are situated within the center of large grassland blocks on the refuge. The remaining shelterbelts along the edges of the refuge could be removed systematically, leaving those with documented Swainson’s hawk and loggerhead shrike nests for last. The shelterbelt in the headquarters area could stay since it is in a highly disturbed area and provides other benefits for refuge equipment and staff.

Shelterbelt removal will be a major undertaking. Recent prescribed fires on the refuge have demonstrated that these fires are not effective for removing trees. Mechanical removal, possibly followed by herbicide, will likely be needed. The understory of most of these shelterbelts is also non-native. Ideally, removal of trees would be followed with restoration of the understory with either dense nesting cover or native grasses. Removal should occur outside of the breeding season, preferrably in fall. Prescribed fire could be used after tree extraction to remove slash piles. A full analysis of the procedure and costs for tree removal should be developed before any trees are removed. In addition, monitoring of bird use should follow any shelterbelt removal. Figure 1. Extent and location of shelterbelts at Benton Lake National Wildlife Refuge. Shelterbelts were mapped along the edge of the canopy on foot using hand-held GeoX-T Trimble units. Any breaks in canopy cover greater than 50m were designated separate subunits of a shelterbelt (e.g. 11a and 11b). Smaller shelterbelts have been buffered to make them more visible on the map. Grassland transect locations (at least 800m from any refuge shelterbelts) are denoted by (▲). Habitat Suitability of BNL Shelterbelts

1.00

0.90

0.80

0.70

0.60

0.50 HSI 1a 0.40 2 15 0.30 14 12 18 6 13 17b 0.20 8 1b 3 5 16 7 9 11a 17a 0.10 4 10 11b

0.00 Shelterbelt

Figure 2. Habitat Suitability Index score for each shelterbelt on the refuge.

Summary of Dead Vegetation in Shelterbelts

14

12

10

8

6

# Shelterbelts # 4

2

0 0 - 25% 26 - 50% 51-75% 76-100%

Percentage of Dead Shrubs/Trees

Figure 3. Number of shelterbelts that had 0-25%, 26-50%, 51-75% and 76-100% dead vegetation. There are more than 18 shelterbelts because we included subunits that were separated by more than 50m. To identify which shelterbelts had subunits, see Appendix 1. HSI vs Bird species richness

16

14

12

10

8

6 # Bird species Bird # 4

2

0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 HSI

Figure 4. Plot of bird species richness in 18 shelterbelts at Benton Lake NWR versus habitat suitability index (R2 = 0.49, p=0.001) (Schroeder et al 1992).

Bird Densities

0.0005000 0.0004500 0.0004000 0.0003500 0.0003000 shelterbelt 0.0002500 grassland 0.0002000 0.0001500

Density (birds/0.4ha) Density 0.0001000 0.0000500 * 0.0000000 BHCO CCLO SAVS VESP WEME Bird Species

Figure 5. Densities of common bird species on shelterbelt transects and grassland transects. Chestnut-collared longspurs had significantly higher densities on grassland transects than shelterbelt transects (p=0.001). There were no significant differences for any of the other species.

Table 1. The number of detections and frequency (%) of birds in shelterbelts and grasslands (>800m from trees) on Benton Lake NWR. Species in bold italics were not detected during grassland transects in this study or refuge grassland point counts (1994-97). Species in italics were not detected on grassland transects in our study, but were detected in previous refuge grassland point counts.

Shelterbelts Grasslands Total Priority Species (n=417) n=59 n=476 Species n % n % n % western meadowlark 104 24.9 7 12 111 23 Brewer's blackbird 60 14.4 0 0 60 13 eastern kingbird 45 10.8 0 0 45 9 savannah sparrow 34 8.2 6 10 40 8 vesper sparrow 25 6.0 1 2 26 5 mourning dove 22 5.3 0 0 22 5 FWS Focalb loggerhead shrike 19 4.6 0 0 19 4 FWS Focal, BCCa brown-headed cowbird 14 3.4 1 2 15 3 yellow-headed blackbird 10 2.4 0 0 10 2 western kingbird 8 1.9 0 0 8 2 chestnut collared longspur 7 1.7 20 34 27 6 FWS Focal, BCC tree swallow 7 1.7 0 0 7 1 American robin 6 1.4 0 0 6 1 black-billed magpie 6 1.4 0 0 6 1 willet 6 1.4 3 5 9 2 FWS Focal, USShorec ruby-crowned kinglet 5 1.2 0 0 5 1 common yellowthroat 4 1.0 0 0 4 1 sharp-tailed grouse 4 1.0 0 0 4 1 PIFd gray partridge 3 <1 0 0 3 1 FWS Focal, BCC, long-billed curlew 3 <1 5 8 8 2 USShorec FWS Focal, BCC, marbled godwit 3 <1 5 8 8 2 USShore Swainson's hawk 3 <1 0 0 3 1 BCC, PIF lark sparrow 2 <1 0 0 2 <1 northern pintail 2 <1 0 0 2 <1 FWS Focal, BCC FWS Focal, BCC, PIF, Sprague's pipit 2 <1 3 5 5 1 S2e American crow 1 <1 0 0 1 <1 great-horned owl 1 <1 0 0 1 <1 gray catbird 1 <1 0 0 1 <1 American kestrel 1 <1 0 0 1 <1 ring-necked pheasant 2 <1 0 0 2 <1 short-eared owl 1 <1 0 0 1 <1 FWS Focal, BCC, PIF FWS Focal, BCC, upland sandpiper 1 <1 0 0 1 <1 USShore white-crowned sparrow 1 <1 0 0 1 <1 horned lark 0 0.0 6 10 6 1 clay colored sparrow 0 0.0 2 3 2 <1 (a) Birds of Conservation Concern (USFWS 2002) and/or North American Waterfowl Management Plan (USFWS 1998) (b) USFWS Migratory Bird Program Focal Species (USFWS 2005) (c) US Shorebird Plan priority species (Brown et al 2001) (d) Partners in Flight species of concern (Rich et al 2004) (e) State species of concern (MTNHP 2007 ) Table 2. Active nests by bird species for each shelterbelt at Benton Lake NWR Shelterbelt # Bird Species 1a 1b 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Total Brewer's blackbird 1 6 1 2 10 eastern kingbird 1 1 1 2 2 1 8 loggerhead shrike 1 2 1 1 1 1 1 1 9 mourning dove 4 1 1 6 unidentified spp 1 1 2 4 Swainson's hawk 1 1 2 1 5 western kingbird 1 1 black-billed magpie 1 1 great-horned owl 1 1 Total nests 7 0 5 2 0 0 1 0 0 0 1 6 7 2 6 3 1 3 1 45

Table 3. Species and condition of trees that active nests were found in during searches in 2007 at Benton Lake NWR.

Tree species Alive Dead caragana 12 Russian olive 19 5 honeysuckle 1 black spruce 2 sand cherry 2 unknw 3 39 5 REFERENCES

Bakker, K. 2003. A synthesis of the effect of woody vegetation on grassland nesting birds. Proceedings of the South Dakota Academy of Science 82:119-141.

Brooks, BL and SA Temple. 1990. Dynamics of a loggerhead shrike population in Minnesota. Wilson Bulletin. 102:441-450.

Brown, S., C. Hickey, B. Harrington, and R. Gill, eds. 2001. The U.S. Shorebird Conservation Plan, 2nd ed. Manomet Center for Conservation Sciences, Manomet, MA.

Dieni, JS and SL Jones. 2002. A field test of the area search method for measuring breeding bird populations. Journal of Field Ornithology 73(3): 253-57.

Dolton, D.D., R.D. Rau, and K. Parker. 2007. Mourning dove population status, 2007. U.S. Fish and Wildlife Service, Laurel, Maryland, USA.

England, A. Sidney, Marc J. Bechard, and C. Stuart Houston. 1997. Swainson's Hawk (Buteo swainsoni), The Birds of North America Online (A. Poole, Ed.). Ithaca: Cornell Lab of Ornithology; Retrieved from the Birds of North America Online: http://bna.birds.cornell.edu/bna/species/265doi:bna.265

Johnson, D. 1979. Estimating nest success: the Mayfield method and an alternative. Auk 96: 651-661.

MacArthur, RH and JW MacArthur. 1961. On bird species diversity. Ecology 42: 594-98.

Martin, S., Piercy, M., Farrina, T. and R. Jordan. Unpublished data. Benton Lake National Wildlife Refuge Non-game Monitoring Points. Informal Report. US Fish and Wildlife Service, Great Falls, MT.

Mayfield, H. 1975. Suggestions for calculating nest success. Wilson Bulletin 87: 456-466.

Montana Department of Natural Resources. 2007. Spring 2007 Seedling Availability and Seed Sources. Helena, MT. http://dnrc.mt.gov/Forestry/ Nursery/Shrubs.asp

MTNHP [Montana Natural Heritage Program]. 2007. Montana animal species of concern. Montana Fish, Wildlife and Parks. Helena, MT. 17pp. http://nhp.nris.state.mt.us/reports/2006_MASOC.pdf

Partners in Flight. 2005. Species assessment database. Partners in Flight Technical Series no. 3. Brighton, CO. http://www.rmbo.org/pif/pifdb.html

Quamen, F and D Naugle. In press. Assessing the impacts of tree plantings on grassland birds in North and South Dakota. Completion report for state wildlife grant #T19. 25pp.

Ralph, CJ, GR Geupel, P Pyle, TE Martin and DF DeSante. 1993. Handbook of field methods for monitoring landbirds. USDA Forest Service General Technical Report PSW-144. Albany, CA

Rich, T. D., C. J. Beardmore, H. Berlanga, P. J. Blancher, M. S. W. Bradstreet, G. S. Butcher, D. W. Demarest, E. H. Dunn, W. C. Hunter, E. E. Iñigo-Elias, J. A. Kennedy, A. M. Martell, A. O. Panjabi, D. N. Pashley, K. V. Rosenberg, C. M. Rustay, J. S. Wendt, T. C. Will. 2004. Partners in Flight North American Landbird Conservation Plan. Cornell Lab of Ornithology. Ithaca, NY. Partners in Flight website. http://www.partnersinflight.org/cont_plan/ (VERSION: March 2005).

Schroeder, RL. 1986. Habitat suitability index models: wildlife species richness in shelterbelts. USFWS Bio Report 82(10.128). 17pp

Schroeder, RL, TT Cable and SL Haire. 1992. Wildlife Species Richness in Shelterbelts: a test of a habitat model. Wild Soc Bull: 20:265-73.

South Dakota Department of Agriculture. No date. Russian-olive fact sheet. Division of Forestry, Pierre, SD. http://www.northern.edu/natsource/ TREESA1/Russia1.htm

Tu, Mandy. 2003. Element stewardship abstract for Elaeagnus angustifolia L. , Russian olive. The Nature Conservancy. Arlington, VA. 9pp.

USFWS [US Fish and Wildlife Service]. 1998. North American Waterfowl Management Plan, 1998 Update. Washington DC. 43pp.

USFWS [US Fish and Wildlife Service]. 2002. Birds of conservation concern 2002. Division of Migratory Bird Management, Arlington, VA. 99pp. http://migratorybirds.fws.gov/ reports/bcc2002.pdf

USFWS [US Fish and Wildlife Service]. 2004. A blueprint for the future of migratory birds: Migratory Bird Program strategic plan 2004-2014. 15pp + appendices. http://www.fws.gov/migratorybirds/mbstratplan/MBStratPlanTOC.htm

USFWS [US Fish and Wildlife Service]. 2005. Migratory bird focal species. Division of Migratory Bird Management, Arlington, VA. http://migratorybirds. fws.gov/mbstratplan/GPRAMBSpecies.pdf

Yosef, Reuven. 1996. Loggerhead Shrike (Lanius ludovicianus), The Birds of North America Online (A. Poole, Ed.). Ithaca: Cornell Lab of Ornithology; Online: http://bna.birds.cornell.edu/bna/species/231 Appendix 1

Shelterbelt Habitat Condition and Bird Inventory Shelterbelt 1a

HSI = 0.73 % Dead= 26-50% Size = 8.6 acres

Tree species: caragana, Russian olive, blue spruce, honeysuckle

Bird Inventory American crow American kestrel American robin Black-billed magpie Brown-headed cowbird Eastern kingbird* Great horned owl Loggerhead shrike Mourning dove* Ruby-crowned kinglet Tree swallow Vesper sparrow Western kingbird* Western meadowlark Yellow-headed blackbird Shelterbelt 1b

HSI = 0.18 % Dead= 76-100% Size = 0.10 acres

Tree species: Elm, lilac

Bird Inventory Gray partridge Western meadowlark

Shelterbelt 2

HSI = 0.39 % Dead= 51 - 75% Size = 15.8 acres

Tree species: caragana, Russian olive, sand cherry, honeysuckle

Bird Inventory Loggerhead shrike* Eastern kingbird Mourning dove* Tree swallow Vesper sparrow Western meadowlark

Shelterbelt 3

HSI = 0.18 % Dead= 0 - 25% Size = 2.1 acres

Tree species: caragana

Bird Inventory Brown-headed cowbird Brewer’s blackbird* Common yellowthroat Eastern kingbird Savannah sparrow Vesper sparrow Western meadowlark Yellow-headed blackbird

Shelterbelt 4

HSI = 0.11 % Dead= 0 - 25% Size = 0.7 acres

Tree species: caragana, Russian olive

Bird Inventory Chestnut-collared longspur Eastern kingbird Marbled godwit Mourning dove Savannah sparrow Western meadowlark Willet

Shelterbelt 5

HSI = 0.18 % Dead= 26 - 50% Size = 2.1 acres

Tree species: caragana

Bird Inventory Eastern kingbird Lark sparrow Long-billed curlew Northern pintail Ring-necked pheasant Vesper sparrow Western meadowlark

Shelterbelt 6

HSI = 0.23 % Dead= 0 - 25% Size = 5.1 acres

Tree species: caragana, Russian olive

Bird Inventory Black-billed magpie* Brewer’s blackbird Common yellowthroat Loggerhead shrike* Ruby-crowned kinglet Savannah sparrow Vesper sparrow Western meadowlark Willet

Shelterbelt 7

HSI = 0.16 % Dead= 0 - 25% Size = 1.4 acres

Tree species: caragana, Russian olive

Bird Inventory Chestnut-collared longspur Eastern kingbird Savannah sparrow Vesper sparrow Western meadowlark

Shelterbelt 8

HSI = 0.22 % Dead= 51 – 75 % Size = 1.4 acres

Tree species: caragana

Bird Inventory Chestnut-collared longspur Eastern kingbird Long-billed curlew Marbled godwit Ruby-crowned kinglet Savannah sparrow Tree swallow Vesper sparrow Western meadowlark Willet

Shelterbelt 9

HSI = 0.14 % Dead= 26 - 50 % Size = 1.1 acres

Tree species: caragana

Bird Inventory Brown-headed cowbird Brewer’s blackbird Eastern kingbird Sharp-tailed grouse Vesper sparrow Western meadowlark Yellow-headed blackbird

Shelterbelt 10

HSI = 0.1 % Dead= 0 - 25 % Size = 0.4 acres

Tree species: Russian olive

Bird Inventory Chestnut collared longspur Eastern kingbird Savannah sparrow Swainson’s hawk* Vesper sparrow Western meadowlark

Shelterbelt 11a

HSI = 0.14 % Dead= 26 - 50% Size = 1.4 acres

Tree species: Russian olive, sand cherry, honeysuckle

Bird Inventory Brown-headed cowbird Brewer’s blackbird Eastern kingbird* Loggerhead shrike* Mourning dove* Sprague’s pipit Swainson’s hawk* Western meadowlark

Shelterbelt 11b

HSI = 0.11 % Dead= 0 – 25 % Size = 0.6 acres

Tree species: Russian olive

Bird Inventory Brown-headed cowbird Brewer’s blackbird Eastern kingbird* Loggerhead shrike* Mourning dove* Sprague’s pipit Swainson’s hawk* Western meadowlark

Shelterbelt 12

HSI = 0.27 % Dead= 26 - 50 % Size = 5.4 acres

Tree species: Russian olive, sand cherry

Bird Inventory Brown-headed cowbird Brewer’s blackbird* Eastern kingbird Loggerhead shrike* Western meadowlark

Shelterbelt 13

HSI = 0.23 % Dead= 0 - 25% Size = 5.1 acres

Tree species: Russian olive, caragana

Bird Inventory Brewer’s blackbird Eastern kingbird* Mourning dove Savannah sparrow Western meadowlark

Shelterbelt 14

HSI = 0.3 % Dead= 0 - 25% Size = 8.6 acres

Tree species: Russian olive, caragana

Bird Inventory Brewer’s blackbird* Chestnut-collared longspur Eastern kingbird* Loggerhead shrike* Savannah sparrow Sprague’s pipit Western meadowlark

Shelterbelt 15

HSI = 0.34 % Dead= 26 - 50% Size = 13.8 acres

Tree species: Russian olive, caragana, sand cherry

Bird Inventory Brown-headed cowbird Brewer’s blackbird* Eastern kingbird Lark sparrow Loggerhead shrike* Sharp-tailed grouse Vesper sparrow Western meadowlark

Shelterbelt 16

HSI = 0.18 % Dead= 0 - 25% Size = 2.4 acres

Tree species: Russian olive, sand cherry

Bird Inventory Chestnut-collared longspur Eastern kingbird* Swainson’s hawk Vesper sparrow Western meadowlark

Shelterbelt 17a

HSI = 0.14 % Dead= 0 - 25% Size = 1.6 acres

Tree species: caragana

Bird Inventory Common yellowthroat Eastern kingbird Mourning dove Savannah sparrow Short eared owl Vesper sparrow Western meadowlark

Shelterbelt 17b

HSI = 0.23 % Dead= 0 - 25% Size = 5.7 acres

Tree species: caragana, Russian olive

Bird Inventory Great horned owl* Swainson’s hawk* Loggerhead shrike?*

Shelterbelt 18

HSI = 0.29 % Dead= 0 - 25% Size = 6.8 acres

Tree species: caragana

Bird Inventory Eastern kingbird Gray partridge Long-billed curlew Loggerhead shrike* Ruby-crowned kinglet Ring-necked pheasant Upland sandpiper Vesper sparrow Western meadowlark

Shelterbelt species list

Elm Ulmus spp. blue spruce Picea pungens sand cherry Prunus besseyi caragana Caragana arborescens lilac Syringia vulgaris Russian olive Elaeagnus angustifolia honeysuckle Lonicera maackii