GRASSLAND BIRD RESPONSE TO DISKING/INTERSEEDING OF LEGUMES IN CONSERVATION RESERVE PROGRAM LANDS IN NORTHEAST NEBRASKA LUCAS PAUL NEGUS

Transcription

1 GRASSLAND BIRD RESPONSE TO DISKING/INTERSEEDING OF LEGUMES IN CONSERVATION RESERVE PROGRAM LANDS IN NORTHEAST NEBRASKA By LUCAS PAUL NEGUS Bachelor of Science University of Nebraska at Kearney Kearney, Nebraska 2002 Submitted to the Faculty of the Graduate College of the Oklahoma State University in partial fulfillment of the requirements for the Degree of MASTER OF SCIENCE May, 2006

2 GRASSLAND BIRD RESPONSE TO DISKING/INTERSEEDING OF LEGUMES IN CONSERVATION RESERVE PROGRAM LANDS IN NORTHEAST NEBRASKA Thesis Approved: Craig Davis Timothy O Connell Tim McCoy A. Gordon Emslie Dean of the Graduate College ii

3 ACKNOWLEDGEMENTS I would like to thank the Nebraska Game and Parks Commission for funding this project through a State Wildlife Grant. I would specifically like to thank Scott Wessel for preparing the grant, as well as for providing guidance, advice, and all levels of support for the project. I also thank Pheasants Forever for providing field equipment and funding for the project. The Lower Elkhorn Natural Resources District also provided funding for field technicians, for which I am very thankful. I am also thankful to the Oklahoma State University Zoology Department for administering the funds for the project. This project could not have been completed without the assistance and guidance of many individuals. I sincerely thank my major advisor, Dr. Craig A. Davis, for guidance, advice, and support through all aspects of this project. I truly appreciate the assistance and advice my graduate committee members, Dr. Timothy O Connell and Dr. Tim McCoy, provided during the study. This project could not have been completed in entirety without the help of my field technicians, Jamie Bachmann, Cassidy Goc, Jordan Johnson, and Adam Schole. I am very thankful for their enthusiasm and tireless work ethic, enduring early, wet mornings and hot afternoons with no complaints. I would also like to thank the graduate students of the Zoology department for providing help and support when needed, and especially for the friendship and many good memories. I would particularly like to thank Sabrina Rust, who lovingly stuck by my side and supported me in everything I did the last several years. I would like to thank my family for all the support and visits, as well as my dog, Tyson, for putting up with city life for iii

4 the last couple of years. Finally, I would like to thank my parents for not only letting me stay at home once again during the summers, but also for instilling the love of the outdoors in me at an early age and supporting me through everything I have done. iv

5 TABLE OF CONTENTS Chapter Page ACKNOWLEDGEMENTS... iii LIST OF TABLES... vii LIST OF FIGURES... ix I. GRASSLAND BIRD RESPONSE TO DISKING/INTERSEEDING LEGUMES IN CONSERVATION RESERVE PROGRAM LANDS IN NORTHEAST NEBRASKA...1 Introduction...1 Grassland Avifauna...1 Habitat Loss and Degredation...1 Habitat Fragmentation...3 Conservation Reserve Program...5 Justification...8 Objectives...12 Methods...12 Study Area...12 Disking and Interseeding...13 Bird Abundance Surveys...15 Nest Searches and Monitoring...15 Vegetation Sampling...16 Statistical Analysis...17 Results...21 Grassland Bird Community...21 Nesting Success...23 Vegetation Characteristics...26 Vegetation Influences...27 Discussion...28 Conservation Implications and Management...41 Literature Cited...46 APPENDICES...85 v

6 Appendix A...85 Appendix B...87 Appendix C...89 vi

7 LIST OF TABLES Table Page 1.1. Overall relative abundance, species richness, and species diversity of breeding grassland birds in treatment and reference fields in Stanton County, Nebraska, Relative abundance (birds/transect) of breeding grassland birds that were observed in >1% of surveys in treatment and reference fields in Stanton County, Nebraska, Relative abundance (birds/transect) of breeding grassland birds that were observed in >1% of surveys in 3 portions of Conservation Reserve Program fields managed by disking/interseeding in Stanton County, Nebraska, Number of nesting species, number of nests, and nest density (nests/ha) of grassland birds in treatment and reference fields in Stanton County, Nebraska, Nest success probabilities for incubation, nestling, and overall nesting period for all bird species in treatment and reference fields in Stanton County, Nebraska, Nest success probabilities for all bird species, dickcissels, and red-winged blackbirds in Conservation Reserve Program fields in Stanton County, Nebraska, 2004 and Vegetation characteristics of successful and unsuccessful nests of all grassland bird species in Conservation Reserve Program fields in Stanton County, Nebraska, Vegetation characteristics of successful and unsuccessful dickcissel nests in Conservation Reserve Program fields in Stanton County, Nebraska, Vegetation characteristics of successful and unsuccessful red-winged blackbird nests in Conservation Reserve Program fields in Stanton County, Nebraska, Vegetation characteristics of successful and unsuccessful grasshopper sparrow nests in Conservation Reserve Program fields in Stanton County, Nebraska, vii

8 1.11. Vegetation characteristics of successful and unsuccessful bobolink nests in Conservation Reserve Program fields in Stanton County, Nebraska, Vegetation characteristics of treatment and reference fields in Conservation Reserve Program fields in Stanton County, Nebraska, Vegetation characteristics of 3 portions of Conservation Reserve Program fields managed by disking/interseeding in Stanton County, Nebraska, Logistic regression models for vegetation variables that best predicted grassland bird presence in treatment sites in Stanton County, Nebraska, Logistic regression models for vegetation variables that best predicted grassland bird presence in reference sites in Stanton County, Nebraska, viii

9 LIST OF FIGURES Figure Page 1.1. Location of study area (represented by black box) in Stanton County, Nebraska Distance frequency histograms for all species, dickcissels, grasshopper sparrows, and bobolinks in treatment and reference fields in Stanton County, Nebraska, 2004 and Mean (+ S.E.) nest density (nests/ha) for all species combined for 0- (n = 2), 1- (n = 4), and 2-year (n = 2) post treatment and no treatment (n = 11) nest search plots in Conservation Reserve Program fields in Stanton County, Nebraska, 2004 and. Means with different letters are significantly different (P < 0.05)...83 ix

10 CHAPTER I: GRASSLAND BIRD RESPONSE TO DISKING/INTERSEEDING OF LEGUMES IN CONSERVATION RESERVE PROGRAM LANDS IN NORTHEAST NEBRASKA INTRODUCTION Grassland Avifauna Throughout the Midwest, grassland bird populations are declining faster than any other group of birds (Samson and Knopf 1994). From 1969 to 1991, grassland bird populations in Illinois, Minnesota, Wyoming, Nebraska and Missouri declined from 24 to 91% (Samson and Knopf 1994). These declines have been attributed mainly to the loss of prairie habitat. Samson and Knopf (1994) estimated that as little as 1% of native prairie habitat remains in the Midwest. Unfortunately, the prairie that remains often exists in small fragments and receives insufficient management. Additionally, modern agricultural practices that favor reduced crop diversity and increased field sizes may contribute to declines in grassland bird numbers (Best et al. 1998). Habitat loss and degradation Habitat loss is the primary cause for declines in grassland bird numbers. There were approximately 162 million ha of prairie in the Great Plains prior to European settlement, with the Tallgrass Prairie Ecoregion constituting approximately 60 million of these hectares (Samson and Knopf 1994, Steinauer and Collins 1996). Today, most of the tallgrass prairie has been plowed and converted to agricultural lands, with losses as high as 99.9% in some states (Steinauer and Collins 1996). Habitat loss is not the only factor contributing to the decline in grassland birds. 1

11 Improper or inadequate management of remaining grasslands may also play a role in grassland bird declines. Fire historically played a major role in the maintenance of prairies. Periodic fires (every 2-5 years) set by lightning or Native Americans restricted woody vegetation from encroaching into the prairies and increased plant species diversity (Steinauer and Collins 1996). Since European settlement, fires have been suppressed allowing woody vegetation to encroach on and often dominate prairies. The importance of fire is evident by the response of grassland birds to this disturbance factor. Dechant et al. (2003) recommended burning every 2-4 years to improve habitat for grasshopper sparrows (Ammodramus savannarum). Bobolinks (Dolichonyx oryzivorus) also respond positively to properly timed burns (2-4 yrs) (Herkert 1994). Johnson and Temple (1990) found lower rates of nest depredation on grasshopper sparrow nests in recently burned areas in Minnesota. Many grasslands are included in farming and ranching operations, often to the detriment of the ecological attributes of those grasslands. Specifically, many agricultural grasslands and forage crops are mowed or hayed annually, with entire fields being cut in mid-summer during the peak of nesting season for many grassland birds. Annual mowing during the breeding season results in high rates of nest failure for dickcissels (Spiza americana; Frawley and Best 1991). In a study in New York, mowing during the breeding season accounted for 51% of nest failures for bobolinks (Bollinger et al. 1990). Annual mowing of entire fields provides habitat for only a select few grassland bird species, limits changes in vegetation structure, and promotes grass succession (Horn and Koford 2000, McCoy et al. 2001a). Improper grazing strategies also adversely affect grassland birds (Zimmerman 1997). The primary impact of grazing is reduction or 2

12 elimination of above-ground vegetation. Season-long grazing or intensive grazing during the nesting period reduces vegetation height and density at a critical period, resulting in reduced avian abundance and productivity (Zimmerman 1997). For example, Swanson (2003) reported that nesting success of savannah sparrows (Passerculus sandwichensis) was significantly higher on ungrazed grasslands than on grasslands grazed continuously. Temple et al. (1999) reported dickcissels were more abundant in ungrazed grasslands than in continuously or rotationally grazed pastures in southwestern Wisconsin. Habitat fragmentation Habitat fragmentation may also be a contributing factor to declining grassland bird numbers (Herkert 1994, Winter and Faaborg 1999). Johnson and Igl (2001) defined habitat fragmentation in prairies as the division of large, contiguous areas of prairie habitat into smaller patches isolated from one another. Three effects of fragmentation are reduced patch-size, increased edge, and increased isolation (Johnson and Igl 2001). Several studies have found decreasing patch size to have negative effects on the presence and nest success of grassland birds (Herkert 1994, Helzer and Jelinski 1999, Winter and Faaborg 1999, Johnson and Igl 2001). Johnson and Igl (2001) found 6 of 15 grassland bird species consistently favored larger patches of habitat over smaller pathces. Herkert et al. (2003) studied nest predation in relation to patch size for 4 grassland bird species (dickcissel, grasshopper sparrow, Henslow s sparrow [Ammodramus henslowii], and eastern meadowlark [Sturnella magna]) and reported predation rates of these species consistently declined with increasing patch size. Moreover, predation rates were consistently lowest in prairies larger than 1,000 ha. Increased edge caused by habitat fragmentation also increases brood parasitism by brown-headed cowbirds (Molothrus ater). Brown-headed cowbirds, one of the most 3

13 common edge species in the Great Plains, seem to favor small grassland patches (Johnson and Igl 2001). Nest productivity is reduced by cowbird parasitism because fewer host young fledge from parasitized nests than non-parasitized nests (Johnson and Temple 1986). Johnson and Temple (1990) found brood parasitism increased with decreasing distance to edge. It is speculated that brown-headed cowbirds are more effective near edges because they use high perches in trees to locate the host s nests (Johnson and Temple 1990). Davis and Sealy (2000) observed that cowbird parasitism was highest in small, irregularly shaped plots with high amounts of edge. Two proposed management strategies that could reduce cowbird parasitism in grasslands are providing and creating large tracts of grassland habitat and removing woody edges and woody vegetation from grasslands (Johnson and Temple 1990, Davis and Sealy 2000, Koford et al. 2000). The effects of habitat isolation are difficult to quantify. In the field of landscape ecology, biologists attempt to label isolated habitats as source or sink habitats for animal populations. A source population is a population in which fecundity is greater than that required to maintain a stable population. Conversely, a sink population occurs if fecundity is below the level that is required to maintain a stable population (Pulliam 1988). Although some grasslands have been identified as source or sink habitats for grassland birds, critical size or specific characteristics of the grassland habitat that influence source/sink status are not easily identified. Management recommendations generally encourage large, diverse grasslands to benefit grassland birds, but exact sizes and habitat conditions required by many species still remain unclear. Johnson and Igl (2001) suggest that creating grassland habitat near existing grasslands, or establishing 1 large field rather than several small ones would benefit more grassland bird species than 4

14 creating small, isolated fields. Johnson and Igl (2001) also noted regional differences may play a role in the development of management strategies (i.e., habitat requirements in 1 region may not be applicable in another region). Identifying the regional differences in habitat requirements and developing management strategies is an important component in the conservation of grassland birds. Conservation Reserve Program With the tremendous losses of native prairie throughout the Midwest, surrogate grasslands such as Conservation Reserve Program (CRP) fields have become increasingly important to grassland wildlife. The CRP was established as a provision of the Food Security Act of 1985 (1985 Farm Bill) and has been retained in both the 1996 and 2002 Farm Bills (U.S. Department of Agriculture 2003a). The original goals of CRP were to reduce erosion and improve water quality of highly erodible cropland, with a co-equal objective of creating and enhancing wildlife and fish habitat added in 1996 (Johnson and Schwartz 1993a). CRP pays farmers annual rental payments to retire highly erodible cropland from production and plant it to grasses, trees, or other perennial cover for years (U.S. Department of Agriculture 2003a). CRP differs from previous farm set-aside programs (e.g., the Payment in Kind Program) because it is a long-term program that retires cropland for greater than 10 years by planting the cropland to permanent vegetation providing substantial wildlife benefits (Hays et al. 1989). In the Midwest, CRP lands have been found to be important to a variety of grassland wildlife species. Game species such as ring-necked pheasants (Phasianus colchicus; King and Savidge 1995), northern bobwhite (Colinus virginianus; King and Savidge 1995), white-tailed deer (Odocoileus virginianus; Luttschwager and Higgins 5

15 1992) and ducks (Reynolds et al. 1994) all use CRP fields. A study by King and Savidge (1995) revealed pheasant abundance was higher in areas with a high percentage of CRP than areas with a low percentage of CRP, probably due to the increased nesting and brood-rearing habitat. Reynolds et al. (1994) found duck nest success in CRP fields to be as high or higher than that of cover planted specifically for duck nesting. Nest success for ducks in CRP fields was 2-9% greater than nest success rates needed to maintain stable populations. Non-game grassland songbirds also benefit from CRP (Johnson and Schwartz 1993b, Patterson and Best 1996). Several studies have attributed increases or at least stable trends in specific grassland bird species to CRP (Igl and Johnson 1995, Herkert 1998, Ryan et al. 1998). Herkert (1998) reported grasshopper sparrow population trends that were negative in 13 mid-continental states prior to CRP were positive following CRP enrollment in those states. Igl and Johnson (1995) attributed a rapid increase in Le Conte s sparrow (Ammodramus leconteii) in North Dakota to the increased nesting habitat offered by CRP when favorable wet weather conditions occurred. In an analysis of North American Breeding Bird Survey results before and after CRP, Reynolds et al. (1994) concluded that 4 of 8 declining grassland bird species showed positive population trends, with only 1 species continuing to decline. Not surprisingly, Ryan et al. (1998) found relative abundance of birds in CRP fields was as much as 10 times higher than in crop fields. In a similar study, Best et al. (1997) also found the abundance of birds to be as much as 10.5 times higher in CRP than crop fields. Moreover, CRP fields supported 3 times more nesting species and 13.5 times the total number of nests than row crop fields. 6

16 Although CRP fields provide habitat for grassland birds, the size and connectivity of fields influence their suitability for grassland birds. McCoy et al. (1999) determined whether CRP lands in Missouri were source or sink habitats for grassland bird species. They found that source-sink status differed among grassland bird species using CRP fields. Among their results, they found that CRP fields were source habitats for grasshopper sparrows, field sparrows (Spizella pusilla), eastern meadowlarks, and American goldfinches (Carduelis tristis) and sink habitats for dickcissels and red-winged blackbirds (Agelaius phoeniceus). McCoy et al. (1999) concluded that CRP created source habitats for many species, especially where alternative-breeding habitat was in poor condition or absent from the area. Other studies, however, have documented declines in grassland bird populations associated with CRP, indicating these lands may act as a sink habitat (Reynolds et al. 1994). Source/sink dynamics of CRP may be important to the conservation of grassland birds and need to be investigated more thoroughly. Although CRP has benefited numerous wildlife species, the types of plantings used in CRP fields can have a major effect on grassland bird species. Conservation Reserve Program fields are planted to a variety of cover types, referred to as conservation practices (CPs). Two of the most commonly implemented CPs are CP1, a cool-season grass mix, and CP2, a warm-season grass mix. CP1 s are frequently dominated by smooth brome (Bromus inermis), orchard grass (Dactylis glomerata), or timothy (Phleum pratense), while CP2s are frequently dominated by switchgrass (Panicum virgatum). The vegetation in CP1s is typically short and dense compared to the tall CP2 vegetation. Recently, natural resource agencies have promoted planting CP2s instead of CP1s to 7

17 benefit wildlife. However, several studies in the Midwest have found no difference in total abundance or species richness of grassland birds between CP1 and CP2 plantings (Johnson and Schwartz 1993b, King and Savidge 1995, Delisle and Savidge 1997, McCoy et al. 2001b). Although avian richness and abundance may not differ between CP1s and CP2s, some grassland bird species prefer certain habitat types that may not be provided separately by each CP. For example, dickcissels prefer CRP fields with tall grasses and high forb content (Patterson and Best 1996, Hughes et al. 1999), while grasshopper sparrows prefer habitats of moderate height and density (McCoy et al. 2001a). Ideally, CRP fields should have separate portions planted to CP1 and CP2 to benefit the full array of grassland bird species. JUSTIFICATION Similar to other states, CRP in Nebraska has become increasingly important to grassland wildlife. The 472,000 ha of CRP in Nebraska is nearly four-times the 123,000 ha of remaining native tallgrass prairie in the state (Steinauer and Collins 1996, U.S. Department of Agriculture 2003b). This CRP acreage provides additional grassland habitat and complements the remaining native grasslands for grassland-dependent species. However, nearly 90% of this CRP acreage in Nebraska is >5 years old (Scott Wessel, Nebraska Game and Parks Commission, personal communication). As a result, these older CRP fields have become senescent, providing limited habitat to wildlife. In eastern Nebraska, typical CP1 plantings in the initial enrollment consisted of a single cool season grass species, smooth brome, combined with several legume species (Trifolium spp.) (Clayton Stalling, Nebraska Game and Parks Commission, personal communication). In the early years of the program, these plantings provided a diverse 8

18 habitat, consisting of a mixture of grasses, and various forbs and legumes that provided excellent habitat for many wildlife species. These diverse and heterogeneous grasslands are desired because they provide multiple niches for a myriad of grassland species. However, due to succession and the inhibiting nature of smooth brome, these once productive grassland fields have become a monoculture of smooth brome that has a limited value to wildlife (Millenbah et al. 1996, McCoy et al. 2001b). Specifically, these monocultures of smooth brome are characterized by short and dense vegetation with high amounts of litter accumulation that provide habitat for a limited number of grassland species. There have been many studies focused on the effects of mowing, grazing, fire, or a combination of these management practices on grassland birds in CRP fields (Zimmerman 1992, Knopf and Samson 1997, Swanson et al. 1999, Horn and Koford 2000, Madden et al. 2000, McCoy et al. 2001a). These management practices can be effective in enhancing CRP, but are frequently ineffective due to logistical and/or social constraints. Mowing and haying are often avoided because of the steep terrain and presence of pocket gophers (Geomys bursarius) in the CRP fields. Pocket gophers make large mounds of soil, which often damage haying equipment to a point that it may be too costly to hay or mow (Dale Clark, CRP landowner, personal communication). Grazing is often avoided because fences must be constructed, water sources provided, and cattle must be monitored daily. Prescribed burning is rarely used for several reasons. First, most landowners are not properly trained to conduct prescribed burns, and local fire marshals are skeptical about providing burn permits to unqualified people. There are also few burn crews that can provide professional burning services, and those that do are 9

19 expensive to hire. Finally, fire is still viewed by many landowners as being destructive and dangerous, even on grasslands. Because these management practices are rarely used, alternative management practices have been explored. Plant diversity has been shown to increase through mechanical means such as disking and interseeding legumes. Shallow disking can be an effective tool for enhancing CRP fields, acting to set back succession in the fields. Disking and interseeding is favored by wildlife managers because of the diverse habitat created and is favored by landowners/farmers because of the availability of equipment and convenience of the practice. However, disking and interseeding is an ephemeral practice and provides diverse habitat for only 4-5 years before grass becomes the dominant vegetation again (Scott Wessel, Nebraska Game and Parks Commission, personal communication). In May 2002, the Nebraska Game and Parks Commission initiated a program to curb declining ring-necked pheasant populations in the state. The program, entitled Focus on Pheasants, placed an emphasis on creating nesting and brood-rearing habitat in aging CRP fields (Taylor 2002). The best habitats for nesting and brood rearing include a weedy, diverse grassland. This can most easily be accomplished in CRP fields by lightly disking the soil to break up the existing mature grass stand which promotes growth of annual broad-leafed forbs (weeds), increases bare ground, and sets back succession (Manley et al. 1994). As an added benefit, alfalfa and sweetclover (legumes) are also interseeded after disking to increase forb abundance and encourage an influx of insects (Whitmore et al. 1986). Manley et al. (1994) found arthropod biomass to be significantly higher in disked portions of a field compared to undisked portions. A study 10

20 conducted by Leathers (2003) evaluated invertebrate abundance on CRP fields in eastern Nebraska that had been disked and interseeded. Although his results were highly variable, Leathers (2003) found invertebrate and forb abundance were often higher in disked and interseeded portions of the fields than untreated CRP fields. The influx in insects is important because they provide critical food resources for both game bird chicks and grassland birds (Whitmore et al. 1986, Jackson et al. 1987, Kobal et al. 1998, McIntyre and Thompson 2003). Mid-term management, a new rule for CRP sign-up 26 and subsequent sign-ups, requires management on CRP fields to enhance habitat diversity (U. S. Department of Agriculture 2003a). Mowing, grazing, burning, and disking/interseeding are all approved practices for this rule. According to the rule, management must be conducted at least once on the entire CRP field during the life of the contract, and management can only be conducted on a maximum of one-third of the field per year. A management regime that includes disking one-third of a field in years 4, 5, and 6 of a contract could create desirable field conditions for wildlife by having part of the field in different successional stages. Millenbah (1993) found avian diversity and density was highest in 1-3 year old CRP fields in Michigan, which was attributed to the diverse vegetation in the younger fields. Mid-term management is also cost-shared under U. S. Department of Agriculture (USDA) rules. Up to 50% of the incurred cost for the performed management practice will be reimbursed to the landowner. Cost-share payments combined with the convenience of disking makes this management attractive to farmers. Resource professionals in the area expect that mandated management on CRP along with the 11

21 efforts of conservation agencies and organizations will significantly increase the practice of disking and interseeding not only in Nebraska, but also throughout the Midwest. Currently, little knowledge exists on the effects of disking and interseeding on grassland bird abundance and nest productivity. With CRP fields increasingly providing key grassland habitat for wildlife throughout the Midwest, disking and interseeding has the potential to improve the quality of existing CRP fields. While I studied the effects of disking/interseeding legumes on a fine geographic scale, the response of the grassland bird community may be extrapolated to a much broader scale to predict the effects of this management practice throughout the Midwest. This study presents an opportunity to evaluate grassland bird population response to different levels of succession in grassland habitats, as well as evaluate the effects of a new management practice on grassland birds. Moreover, information from this study will be important in guiding future decisions concerning management of CRP fields and ultimately may influence future USDA policies for the Conservation Reserve Program. OBJECTIVES 1) Compare avian richness and abundance in disked/interseeded CRP fields to unmanaged CRP fields. 2) Compare avian nest productivity in disked/interseeded CRP fields to unmanaged CRP fields. 3) Evaluate differences in vegetation structure, composition, and cover between disked/interseeded and unmanaged CRP fields. METHODS Study Area 12

22 This study was conducted in Stanton County, Nebraska during the summers of 2004 and (Figure 1). Stanton County lies in the Loess Uplands and Till Plains Major Land Resource Area in northeast Nebraska. Precipitation in the area averages 50 to 65 cm per year, with most occurring during the growing season (Natural Resource Conservation Service 2003). Soils in the area are of the Crofton-Nora and Nora-Crofton complexes that include silty, loamy, and sandy textural classes (Hammond 1982). Stanton County is mainly an agricultural county with corn and soybeans being the predominant crops. Other common crops include oats and alfalfa (Hammond 1982). Nearly 11,200 ha of CRP occur in Stanton County (U.S. Department of Agriculture 2003b). This study was conducted on privately-owned CRP within a 51.5 km 2 study area. The study area location was selected because of the large percentage of older CRP fields in the area, the potential to improve habitat quality from management practices, and landowner cooperation (Taylor 2002). Sixteen CRP fields ranging in size from 16 to 64 ha were selected for the study. Eight of the fields were manipulated (treatment fields), and 8 fields were unmanipulated and used as reference fields. A portion (25-33%) of each manipulated field was disked in 2003 and 2004, resulting in 50-66% of each field occurring in different successional stages. Conservation Reserve Program fields in the study area were originally planted to smooth brome, alfalfa, and yellow sweetclover, but are now dominated by smooth brome. No planned management (disking, haying, or grazing) has occurred on the reference fields. Selection of study fields was based on previous management history, field size, proximity to other fields, and landowner cooperation. Disking and Interseeding 13

23 Disking and interseeding of the treatment fields began in September Disking was performed by a contractor using a tandem disc designed specifically for sod breakup. Portions (up to one-third) of each field were selected and flagged prior to disking. Selection of portions to be disked was based on topography and landowner preference. Wherever possible, portions that had been hayed in August 2002 were selected because the reduction of residual cover made disking more effective. Disking was maintained at a depth of 7.6 to 10.2 cm. Additionally, each portion was disked twice to effectively break-up the sod. After disking, a no-till drill was used to interseed legumes. The legume mix, provided by Pheasants Forever, consisted of alfalfa, red clover, and yellow sweetclover. The legumes were seeded at a rate of 6.75 kg/ha (1.69 kg of red clover, 1.69 kg of yellow sweetclover, and 3.38 kg of alfalfa). According to USDA guidelines, interseeding was terminated on 15 September 2002 to avoid winter kill of late germinating plants (Natural Resource Conservation Service 2002). Seeding was continued after 1 November, when consistent freezes occurred at night. Consistent freezes ensured a low soil temperature was maintained even during times with warm daytime temperatures, preventing seed germination and winter-kill. Disking resumed in April 2003 and continued until 1 May No disking or interseeding was conducted between 1 May and 1 August Disking on another portion of each field resumed in fall 2003 and was completed in the spring of All disking depths and seeding rates were performed in accordance with USDA guidelines (Natural Resource Conservation Service 2002). Treatment study fields consisted of CRP fields that had portions disked and interseeded over the span of 2 years. A typical treatment study field had 25-33% disked 14

24 in 2003 and 25-33% disked in 2004, with the remainder of the field undisked. The reference study fields did not have any disking performed on any portion of the field. Bird Abundance Surveys I surveyed birds using the belt transect method similar to Best et al. (1997) and McCoy et al. (2001b). Transects were 200-m long with a fixed 100-m width (Helzer and Jelinski 1999). Transects were not located <50 m from any field edge or boundary between disked portions of the field. By maintaining transects >50 m from the edge of any field (including roads, tree lines, and agricultural fields), I minimized edge effects on survey results (Helzer 1996). In the treatment fields, I established 3 transects with 1 transect located in each portion of the field (i.e., 2 transects in disked/interseeded portions and 1 transect in the undisked portion). In the reference fields, I randomly established 2 transects in each field in In, an additional transect was established in each reference field to be more consistent with surveys in treatment fields. I surveyed each transect for avian abundance once during 3 periods (25 May-14 June, 15 June-30 June, and 1 July-15 July) (Best et al. 1997, McCoy et al. 2001b). Avian abundance was determined by counting birds seen and/or heard while walking along each transect. All birds encountered within the belt transect were counted. Birds flying over the transect were noted, but were not included in abundance data for the fields. In, I recorded the perpendicular distance from the transect to each bird to determine detection probability for each species encountered. Surveys were conducted between sunrise and 3 hours after sunrise on days when the wind was <16 km/hr with no fog or rain. Nest Searches and Monitoring 15

25 In 2 treatment fields, I established a 4-ha nest search plot centered around each avian survey transect, resulting in 3 nest search plots per field. In 2004, I also established 2, 4-ha nest search plots in reference fields. In, the number of search plots in reference fields was increased to 6 to be more comparable to treatment fields. Nest searches were conducted using 4 methods: 1) observing birds engaged in breeding behavior indicating nest building, incubation, or feeding of nestlings (Martin and Geupel 1993, Giuliano and Daves 2002), 2) chain or rope dragging, 3) systematic walking with or without a sweeping stick (Winter et al. 2003), and 4) random searches through field plots (Giuliano and Daves 2002). Plots were searched periodically (>2 times per sampling period) throughout the summer on a rotational basis to minimize disturbance. Additional nests were located while conducting avian abundance surveys or sampling vegetation. Once a nest was located, it was marked with flagging 5 m to the north and 5 m to the south of the nest (Giuliano and Daves 2002), and a description of the nest location was recorded on a nest data sheet. Additionally, Universal Transverse Mercator (UTM) coordinates for each nest were obtained using a Global Positioning System (GPS) unit. Nests were checked every 2-5 days until a final outcome was determined (Martin and Geupel 1993). Outcome was recorded as 1) successful (at least 1 young fledged), 2) unsuccessful (no young fledged due to depredation, weather, nest parasitism, or unknown causes), or 3) undetermined. Vegetation Sampling Vegetation sampling occurred along the avian abundance survey transects and at each nest location. I located 4 sampling points (spaced evenly at 40 m, 80 m, 120 m, and 16

26 160 m) along each transect to record vegetation data. Vegetation data was collected along survey transects immediately following avian abundance surveys during each period and at nests within 10 days of nest completion. At each sampling point on the transect and at each nest, I recorded the following vegetative characteristics: percent cover, maximum vegetation height, litter depth, and horizontal visual obstruction. Percent vegetation cover was estimated using a 20 x 50 cm Daubenmire (1959) frame. Each frame was centered around each sampling point and percent grass, forb, dead material, and bare ground were estimated (Hughes et al. 1999). Planted legumes were included with all other forbs in the percent forb cover estimate. Horizontal visual obstruction was measured using a Robel pole centered on each sampling point and at each nest (Robel et al. 1970). Horizontal visual obstruction readings were taken from a height of 1 m and a distance of 4 m from the 4 cardinal directions. Litter depth and maximum vegetative height within the Daubenmire frame were measured at each corner of the frame using a meter stick (Hughes et al. 1999). Additionally, the plant species each nest occurred in and the height of each nest was recorded. Statistical Analysis Prior to conducting statistical analysis, I determined total bird and individual species relative abundances for treatment and reference fields by averaging count data from each survey period for each year. Bobolink relative abundance was calculated from the first and second survey periods because bobolinks were in large groups preparing for migration during the third survey period. I only included bird species that occurred in >1% of all surveys for the calculation of individual species abundances. Shannon- 17

27 Weiner species diversity index was used to calculate overall bird diversity for treatment and reference fields for each year (Krebs 1999). To examine similarities in grassland bird communities between treatment and reference fields and among portions of treatment fields, I used Morisita s index of similarity (Brower and Zar 1977): I M = 2x i y i / ( ) N 1 N 2, where x i is the number of individuals in species i for community 1, y i is the abundance of species i in community 2, 1 is Simpson s dominance index for community 1, 2 is Simpson s dominance index for community 2, N 1 is the total number of individuals in community 1, and N 2 is the total number of individuals in community 2. The range of I M is from 0 (no similarity) to 1 (complete similarity). Avian species richness was also calculated for each treatment and reference field for both years and included all species observed at any time in study fields during each breeding season. I used Levene s test to test for homogeneity of variance for the avian abundance variables (Zar 1999). Most data sets did not meet assumptions of homogeneity, so I used a square-root transformation to correct for heteroscedacity (Zar 1999). Because only total bird abundance met assumptions after transformation, I rank-transformed individual species data sets that did not meet the assumptions of homogeneity (Conover and Iman 1981). I used a 2-way analysis of variance (ANOVA) to examine differences in total bird abundance, individual species abundance, diversity, and richness between treatment and reference fields and years (2004 and ) (SAS Institute 2003). To further evaluate the effects of disking/interseeding on avian species, I also used a 2-way ANOVA to examine differences in the total bird abundance and individual species abundance of 5 species (dickcissel, red-winged blackbird, bobolink, grasshopper sparrow, and common yellowthroat) among treatments within each manipulated field (disked-2003, disked- 18

28 2004, and undisked) and year. I used Duncan s Multiple Range Test to examine differences between the 3 portions of treatment fields (SAS Institute 2003). To assess differences in detectability of individual species between treatment and reference fields, I constructed frequency histograms with 10-m increments using the perpendicular distances for each bird observation that was recorded in (Rotella et al. 1999, McCoy et al. 2001b). Frequency histograms were constructed for all species combined, as well as for dickcissels, bobolinks, and grasshopper sparrows individually because they had >20 observations for both treatment and reference fields. Frequencies were computed by calculating the proportion of observations in each distance increment for treatment and reference fields. In general, comparisons of these histograms indicated detectability for all species combined and individual species were similar between treatment and reference fields (Figure 2). Nearly all the vegetation data failed to meet assumptions of homogeneity. Percent dead material was the only vegetation variable that was homogenous. All other variables (maximum vegetation height; horizontal visual obstruction; litter depth; and percent forb, grass, and bare ground cover) were square-root transformed (Zar 1999), but still did not meet assumptions of homogeneity. Therefore, I rank-transformed those variables (Conover and Iman 1981). I used a 2-way ANOVA to examine differences in vegetation variables between treatment and reference fields and years. To further evaluate effects of disking/interseeding on vegetation characteristics, I used a 2-way ANOVA to examine differences in vegetation variables among different treatments within each manipulated field and year. I used Duncan s Multiple Range Test to examine differences among the 3 portions of treatment fields (SAS Institute 2003). 19

29 To determine the effect of vegetation characteristics on nest success, I used a 2- way ANOVA to examine differences in vegetation variables for years and nest fate (successful and unsuccessful) of all bird species combined, as well as individual dickcissel, red-winged blackbird, bobolink, and grasshopper sparrow nests. Nest success for all bird species combined, dickcissels, and red-winged blackbirds was determined using the Mayfield method (Mayfield 1975, Johnson 1979). Nest success probability between reference and treatment fields was examined using only all species combined for 2004 and because <20 nests were found in reference fields in Dickcissel and red-winged blackbird nest success probability could not be compared between treatment and reference fields because <20 nests for each species were found in reference fields during both years. I used a Chi-square contingency table to examine differences in nest success, number of nests, and number of species nesting between treatment and reference fields (Dow 1978, SAS Institute 2003). I used a 2-way ANOVA to test for differences in nest density between treatment and reference fields and years (SAS Institute 2003). Additionally, I used a 1-way ANOVA to test for differences in nest density among plots that were in different post-treatment stages (1- and 2- years post-treatment, current-year treatment, and no treatment). For all analyses performed, I inferred a significance level at P < I used logistic regression to develop explanatory models for the occurrence of individual grassland bird species occurring in >1% of surveys in treatment and reference fields based on vegetation characteristics (i.e., forb, grass, dead material, and bare ground cover; horizontal visual obstruction; litter depth; and maximum vegetation height). I selected logistic regression over linear regression because the individual species 20

30 abundance data were heavily weighted with zeros and violated assumptions of linear regression (Zar 1999, Madden et al. 2000). I used the logistic model: P(presence) = 1/ (1 + exp{ - [b o + b 1 (x)]}) where P(presence) was the probability that a bird species was present, b o and b 1 were intercept and slope coefficients, and x was the predictor variable (vegetation variable). I used a backward-elimination routine to create the best model for each bird species using all vegetation variables. A variable was eliminated from the model if its observed significance level for the regression coefficient (based on Wald chisquare) was P > I used the Hosmer and Lemeshow (1989) test to assess the goodness-of-fit of each model. RESULTS Grassland Bird Community Over the 2 years of the study, I observed 28 bird species in treatment fields and 25 species in reference fields (Appendix A). There was high overlap in species assemblages (I M = 0.72) between treatment and reference fields, yet several species were unique to treatment and reference fields. Orchard orioles (Icterus spurious), American robins (Turdus migratorius), cedar waxwings (Bombycilla cedrorum), house sparrows (Passer domesticus), and killdeer (Charadrius vociferous) were only observed in treatment fields, while eastern meadowlarks and Baltimore orioles (Icterus galbula) were only observed in reference fields. Mean abundance of all grassland birds and species diversity were higher in treatment than reference fields (Table 1). There was a treatment x year interaction for species richness (F 1,28 = 4.47, P = 0.044). In 2004, species richness was higher in treatment fields, but in, there was no significant difference for species richness between treatment and reference fields. 21

31 During the study, 2 species (common yellowthroats [Geothlypis trichas] and redwinged blackbirds) were significantly more abundant in treatment fields, 2 species (bobolinks and Henslow s sparrows) were significantly more abundant in reference fields, and 1 species (sedge wren [Cistothorus platensis]) had similar relative abundances in treatment and reference fields (Table 2). Three species (dickcissels [F 1,260 = 5.13, P = 0.024], grasshopper sparrows [F 1,260 = 8.55, P = 0.004], and western meadowlarks [Sturnella neglecta] [F 1, 260 = 5.29, P = 0.022]), showed significant treatment x year interactions. Dickcissels were more abundant in treatment fields for each year. Grasshopper sparrow abundance did not significantly differ between treatment and reference fields in 2004, but they were more abundant in reference than treatment fields in. Western meadowlark abundance was significantly higher in treatment fields in 2004, but was not significantly different between treatment and reference fields in. Dickcissels, grasshopper sparrows, and bobolinks were the most abundant species for both years in treatment and reference fields, accounting for 78% of the total bird abundance. High overlap occurred in grassland bird communities among portions of treatment fields. Species assemblages between disked-2003 and disked-2004 portions (I M = 0.90) and between disked-2004 and undisked portions (I M = 0.84) exhibited high overlap. The least overlap occurred between disked-2003 and undisked portions (I M = 0.66) of treatment fields. Though Morista s index of similarity indicated grassland bird communities in different portions of treatment fields were similar, there were several differences in bird abundances within treatment fields (Table 3). Bobolinks were significantly more abundant in the undisked portion than the disked-2003 and disked- 22

32 2004 portions (P < 0.001). Treatment x year interaction effects were observed for all other species (all species combined [F 2, 138 = 16.76, P < 0.001], dickcissel [F 2,138 = 47.98, P < 0.001], red-winged blackbird [F 2,138 = 10.41, P < 0.001], grasshopper sparrow [F 2,138 = 5.77, P = 0.004], and common yellowthroat [F 2,138 = 10.12, P < 0.001]) (Table 3). Therefore, subsequent analyses for those species were performed within years. In 2004, all species combined, dickcissels, red-winged blackbirds, and common yellowthroats were significantly more abundant in the disked-2003 portion (P < 0.001) than the disked or undisked portions of the fields. Grasshopper sparrows were significantly more abundant in the disked-2004 portion (P = 0.002) of the fields than the disked-2003 and undisked portions of the fields in In, all species combined, dickcissels, and common yellowthroats were significantly more abundant in the disked-2004 portion (P < 0.001) than the disked-2003 or undisked portions. Red-winged blackbirds did not significantly differ among treatment types (P = 0.522) in. In, grasshopper sparrows were significantly more abundant in the undisked portion than the disked-2003 portion (P = 0.008), but abundance in the disked-2004 portion did not differ significantly from the other 2 portions (Table 3). Dickcissels dominated the observations in the disked-2003 and disked-2004 portions of treatment fields, accounting for 56% and 43% of bird observations, respectively. In the undisked portions of treatment fields, grasshopper sparrows were observed most often (34%), followed closely by dickcissels (27%) and bobolinks (26%). Nesting Success I located a total of 247 nests (112 nests in 2004, 135 nests in ) during the 2 years of the study. Of those 247 nests, 206 nests of 8 species occurred in treatment fields, 23

33 and 41 nests of 10 species occurred in reference fields (Table 4). The majority of the nests located were dickcissel (89) and red-winged blackbird (107) nests. Nest density did not differ between treatment and reference fields (P = 0.3) or years (P = 0.850) (Table 4). When further evaluated by years post-treatment instead of by treatment or reference field, nest density was significantly higher in the 1-year post-treatment plots than the currentyear post-treatment and no treatment plots (P = 0.008). Nest density in the 2-year posttreatment plots did not differ from any other plots (Figure 3). Overall nest success probability for all species was similar between treatment and reference fields (Table 5). Mayfield nest success probability for all bird species was higher in 2004 than (Table 6). Dickcissels had nearly twice the nest success probability in 2004 than, and red-winged blackbirds were nearly 3 times more successful in 2004 than (Table 6). Apparent nest success for all birds was 31% in 2004 and 19% in. For dickcissels and red-winged blackbirds, apparent nest success was 33% and 29% in 2004, and 16% and 9% in, respectively. During both years, nest failures for all birds were attributed to several causes: depredation (64% , 75% - ), abandonment (8% , 13% - ), weather (7% , 9% - ), and unknown causes (21% , 3% - ). Dickcissel nest failures were attributed to depredation (52% , 73% - ), abandonment (7% , 19% - ), weather (15% , 3% - ), and unknown causes (26% , 5% - ). Red-winged blackbird nest failures were attributed to depredation (71% , 75% - ), abandonment (9% , 12% - ), weather (3% , 2% - ), and unknown causes (17% , 12% - ). 24

34 I observed many differences in vegetation characteristics between successful and unsuccessful nests for all species combined (Table 7). Litter depth and percent grass cover were significantly higher at successful nests. Unsuccessful nests had a significantly higher percentage of forb cover than successful nests. Litter depth, percent grass cover, and percent dead material cover were significantly higher at all nests in than The percentage of forb cover and bare ground was higher at all nests in 2004 than (Table 7). A fate x year interaction was observed in nest height (F 1,234 = 5.55, P = 0.019), horizontal visual obstruction (F 1,236 = 5.48, P = 0.020), and maximum vegetation height (F 1,236 = 8.19, P = 0.005). Subsequent analyses for these characteristics were performed within years. In 2004, there was no difference in nest height, horizontal visual obstruction, or maximum vegetation height between successful and unsuccessful nests. In, nest height, horizontal visual obstruction, and maximum vegetation height were significantly higher at unsuccessful nests than successful nests. There were few differences in vegetation characteristics between successful and unsuccessful dickcissel nests (Table 8). Fate x year interactions were observed for maximum vegetation height (F 1,78 = 6.26, P = 0.014) and percent dead material (F 1,78 = 4.09, P = 0.047). Further analyses were performed within years for these characteristics. In 2004, neither maximum vegetation height nor percent dead material differed between successful and unsuccessful dickcissel nests. In, maximum vegetation height was higher at unsuccessful than successful dickcissel nests. Nest height, horizontal visual obstruction, percent forb cover, and percent bare ground were significantly higher at dickcissel nests in 2004 than. Litter depth and percent grass cover were significantly higher at dickcissel nests in than Nest height did not differ 25

35 between successful and unsuccessful dickcissel nests, but dickcissel nests were higher in 2004 than (Table 8). Red-winged blackbird nests showed several differences between successful and unsuccessful nests (Table 9). No fate x year interactions were observed. Litter depth and percent grass cover were higher at successful than unsuccessful red-winged blackbird nests (Table 9). Percent forb cover was higher at unsuccessful than successful red-winged blackbird nests. Horizontal visual obstruction, maximum vegetation height, percent forb cover, and percent bare ground were higher at all nests in 2004 than. Litter depth and percent grass cover were higher at redwinged blackbird nests in than Grasshopper sparrow nests exhibited 1 difference between successful and unsuccessful nests; percent bare ground was higher at successful than unsuccessful grasshopper sparrow nests (Table 10). There were no significant differences between successful and unsuccessful bobolink nests (Table 11). Vegetation Characteristics Many vegetation characteristics differed between treatment and reference fields (Table 12). Forb cover was significantly higher in treatment fields, while grass cover was significantly higher in reference fields. Grass cover was higher in all fields (treatment and reference) in than Treatment x year interactions occurred in horizontal visual obstruction (F 1,1040 = 8.65, P < 0.001), litter depth (F 1,1040 = 30.17, P < 0.001), maximum vegetation height (F 1,1040 = 61.70, P < 0.001), percent dead material (F 1,1040 = 25.20, P < 0.001), and percent bare ground (F 1,1040 = 47.13, P < 0.001). Further analyses of these vegetation characteristics were performed within years. Maximum vegetation height, horizontal visual obstruction, and percent bare ground were all significantly higher in treatment fields than reference fields in 2004 and. Litter depth and 26

36 percent dead material were significantly higher in reference than treatment fields in 2004 and (Table 12). There were also many differences in vegetation characteristics within treatment fields (Table 13). Treatment x year interactions occurred in all analyses (horizontal visual obstruction [F 2,570 = , P < 0.001]; litter depth [F 2,570 = 81.33, P < 0.001]; maximum vegetation height [F 2,570 = , P < 0.001]; percent forb [F 2,570 = , P < 0.001], grass [F 2,570 = 46.53, P < 0.001], dead material [F 2,570 = 18.59, P < 0.001], and bare ground [F 2,570 = 56.38, P < 0.001] cover). Thus, further analyses were performed within years. Horizontal visual obstruction was highest in the disked-2003 portion of the treatment fields for both years (P < 0.001) (Table 13). Litter depth and percent dead material were highest in the undisked portion of the treatment fields for 2004 and (P < 0.001). In 2004, litter depth was similar between disked-2003 and disked-2004 portions, while in disked-2003 portions had more litter than disked-2004 portions. In 2004, maximum vegetation height was highest in the disked-2003 portion, but in, maximum vegetation height was highest in disked-2003 and disked-2004 portions (P < 0.001). Forb cover was highest in the disked-2003 portion of treatment fields in 2004 (P < 0.001), but shifted to being highest in the disked-2004 portion in (P < 0.001). In 2004, grass cover was higher in the undisked portion of the fields than the other 2 portions (P < 0.001). But, in, grass cover was highest in undisked and disked-2003 portions (P < 0.001). Percent bare ground was highest in the disked-2004 portion of the fields in 2004 and (P < 0.001) (Table 13). Vegetation Influences 27

37 Vegetation had considerable influence on the presence of individual grassland bird species in CRP fields. In treatment fields, 7 of the 8 most common grassland bird species had significant explanatory models (Table 14). Presence of all the species, except red-winged blackbirds, was best explained by more than 1 variable. Red-winged blackbird occurrence was explained by increasing maximum vegetation height. Dickcissel occurrence was best explained by 2 variables, increasing forb cover and visual obstruction. Common yellowthroat occurrence was explained by increasing forb cover and vegetation height, while sedge wrens were associated with increasing forb cover and litter depth. Bobolinks were associated with increasing dead material and decreasing bare ground. Western meadowlarks were associated with decreasing forb and grass cover. Grasshopper sparrows were associated with increasing forb cover and litter depth, and decreasing vegetation height (Table 14). Henslow s sparrows did not have a significant model in treatment fields. In reference fields, 3 of the 8 most common grassland bird species had significant explanatory models (Table 15). Common yellowthroats were associated with increasing vegetation height, and red-winged blackbirds were associated with increasing litter depth. Bobolink occurrence was explained by 2 variables, increasing grass and dead material cover. Dickcissels, grasshopper sparrows, Henslow s sparrows, sedge wrens, and western meadowlarks did not have significant explanatory models in reference fields. DISCUSSION Overall, CRP grasslands in northeastern Nebraska provided habitat for a wide variety of grassland birds. Of the 30 bird species observed in my study fields, 8 were grassland bird species that commonly occurred in the fields and used the fields for 28

38 nesting, foraging, and raising young. In an analysis of North American Breeding Bird Surveys (BBS) from , Herkert (1995) reported that over 75% of all grassland bird species, including the 8 most commonly observed species in my study, have shown significant declines throughout the Midwest. Recent BBS data ( ) indicate that grasshopper sparrows, western meadowlarks, dickcissels, and common yellowthroats continue to decline (Sauer et al. ). However, these more recent data also show an increasing trend for bobolinks and sedge wrens and a stable or a slightly decreasing trend for red-winged blackbirds. Henslow s sparrows, which were only observed in my reference fields, may be one of the fastest declining grassland birds (Herkert 1997), especially in the east and northeast portions of its range (Pruitt 1996). Recently, however, local Henslow s sparrow populations in the Midwest have increased (Sauer et al. ), possibly due to habitat created by CRP (Herkert et al. 2002). These recent BBS results, along with my findings, suggest that CRP grasslands will continue to be important for these grassland bird species into the future and may be critical in maintaining or increasing current populations. Although CRP fields appear to provide critical habitat for grassland birds, the lack of appropriate management often reduces their value to individual species. Disking and interseeding creates a diverse mosaic of vegetation that positively benefits many grassland bird species. Avian species diversity, richness (in 2004), and mean abundance of all species combined were higher in the treatment than reference fields, indicating the vegetation structure resulting from treatment of fields accommodated more species than the reference fields. Millenbah et al. (1996) found similar results in newly planted CRP fields, presumably because of the vegetation changes from the disturbance of planting 29

39 them. In my study, disked/interseeded portions of treatment fields had a diverse vegetation composition and structure, consisting of various heights of forbs, weeds, and grasses. Additionally, treatment fields consisted of 3 portions that were in different vegetation successional stages. This diverse vegetation composition and structure within disked/interseeded portions and variety of successional stages throughout treatment fields is likely responsible for the high avian diversity, richness, and abundance. Dickcissels were the most common grassland birds in my study area and were most abundant in treatment fields. Dickcissels prefer grasslands that have dense, tall cover that provide many elevated song perches (Hughes et al. 1999, Temple 2002, Dechant et al. 2003). Additionally, areas with high proportions of forbs, especially legumes, provide nesting cover, nest support (Appendix B), and an increased abundance of invertebrate foods (Frawley and Best 1991, Patterson and Best 1996, Temple 2002). Treatment fields provided the dense, tall legumes important to dickcissels for nesting and foraging, along with live and dead forbs (sunflowers [Asteraceae], ragweed [Asteraceae], hemp [Cannabaceae], and thistles [Asteraceae]) that were frequently used for perches (Lucas Negus, personal observation). More specifically, the disked/interseeded portions of the treatment fields, particularly the 1-year post-treatment portions, provided vegetation structure that was nearly ideal for dickcissels. This preference for the vegetation structure in the 1-year post-treatment portions within treatment fields is shown by the consistent high abundance of dickcissels in the 1-year post-treatment portions of treatment fields over the 2 years of the study. Dickcissels were significantly more abundant in the 1-year post-treatment portion of treatment fields during both years of the study. Red-winged blackbirds and common yellowthroats were also more abundant in 30

40 treatment fields. Like dickcissels, both species showed preferences for the 1-year posttreatment portion of the treatment fields during both years of the study. Red-winged blackbirds, which generally breed in a wide range of wetland and upland habitats (Yasukawa and Searcy 1995), were observed almost exclusively in treatment fields. They commonly placed their nests high (> 50 cm) in the legumes, and used live and dead forbs for perching. Common yellowthroats also used the tall vegetation for perching, but were more commonly observed hidden in the lower, dense legumes. This preference for the low, dense vegetation by common yellowthroats is consistent with that reported by Stewart (1953) for common yellowthroats in Michigan. Grasshopper sparrows showed no preference between treatment and reference fields in 2004, but preferred reference fields in. Grasshopper sparrows generally prefer grasslands of moderate height and density with patchy bare ground (Vickery 1996, McCoy et al. 2001a). With the exception of bare ground, reference fields and undisked portions of the treatment fields offered this suitable grassland habitat of moderate height and density. However, within treatment fields in 2004, grasshopper sparrows displayed a preference for areas with bare ground and were significantly more abundant in the portions of treatment field that had been disked in the spring 2004 than any other field portions. In the disked portions, grasshopper sparrows were frequently observed perched on dirt clumps, where vegetation was sparse and short or intermediate in height throughout the season and bare ground was abundant (Lucas Negus, personal observation). In, disked/interseeded portions of treatment fields had undergone at least 1 full growing season, resulting in the majority (50 66%) of each field having tall, 31

41 dense vegetation that was less suitable for grasshopper sparrows compared to reference fields. Western meadowlarks were more abundant in treatment fields in 2004, but showed no difference between treatment and reference fields in. The preference of treatment fields in 2004 was somewhat unexpected and contradictory to the findings of Wiens and Rotenberry (1981), who reported western meadowlarks preferred areas with high amounts of grass and litter cover. Reference fields were characterized more by high amounts of grass and litter than treatment fields. Perhaps the treatment fields provided adequate areas of grass and litter cover for nesting and were more attractive for foraging than reference fields. In, western meadowlarks showed no preference between treatment and reference fields, possibly indicating that the vegetation changes in treatment fields between 2004 and resulted in less desirable vegetation in. Bobolinks and Henslow s sparrows were more abundant in reference fields than treatment fields during both years of the study. Both species prefer habitats with high percentages of grass and dead material, and high amounts of litter (Bollinger and Gavin 1992, Herkert and Glass 1999). This habitat was primarily found in reference fields and the portion of treatment fields that was undisked. Henslow s sparrows were most often observed perched and nesting on hilltops where vegetation was not as high or dense as it was in valleys. This is contrary to the findings of Zimmerman (1988) who found Henslow s sparrow males in Kansas set up territories in areas with taller vegetation than surrounding areas that were unused. This may indicate Henslow s sparrows use a wider range of habitats than originally thought. Bollinger and Gavin (1992) found bobolinks in New York preferred old hayfields (>8 years old), and fields >30 ha supported twice the 32

42 bobolink density of fields <10 ha. The reference fields in my study were all >10 ha and had not been disturbed for >10 years; it is not surprising bobolink abundance was higher in these fields than the treatment fields that had smaller undisked portions. Although results from my study primarily support Bollinger and Gavins (1992) findings, bobolinks were also observed nesting in small (<10 ha) undisked portions of treatment fields, as well as disked/interseeded portions of the fields. This indicates that although treatment fields were not used as much as reference fields were used by bobolinks, they still provided adequate nesting habitat for a reduced number of bobolinks. Not surprisingly, vegetation characteristics significantly affected presence of many of the grassland bird species in CRP fields. Nearly every species (7 of 8 species) had significant explanatory models in treatment fields, while only 3 species (red-winged blackbirds, bobolinks, and common yellowthroats) had significant models in reference fields. Reference fields likely had few explanatory models because vegetation was very uniform throughout the entire fields, and species presence could not be attributed to distinct vegetation characteristics. Dickcissels and common yellowthroats were associated with similar vegetation variables in treatment fields. Dickcissels were positively associated with vegetation height and forb cover, while common yellowthroats were positively associated with visual obstruction and forb cover in treatment fields. Both species were most commonly observed in disked/interseeded portions of the treatment fields that had undergone at least 1 full growing season. Although these 2 species were associated with similar vegetation variables and were commonly observed in the same portions of the treatment fields, they were likely able to coexist without competition in these portions because the vertical vegetation structure present provided a 33

43 separate foraging space for each species. Cody (1968) concluded that vegetation height was important because avian species foraging strategies are stratified by the vertical vegetation structure. Dickcissels used the upper levels of the vertical structure in CRP fields, while common yellowthroats used the lower levels. Maximum vegetation height has been reported as being an important factor for dickcissels (Zimmerman 1966) and was likely a key variable in my study because dickcissels were commonly observed perching and nesting high in the vegetation. In addition to the height of vegetation, the amount of forb cover present was also important for dickcissels. Delisle and Savidge (1997) concluded dickcissel abundance was positively correlated to forb occurrence. Forbs were used by dickcissels for both perching and nesting during my study. Visual obstruction was likely a key variable for common yellowthroats because they prefer the low, dense vegetation (Stewart 1953). A high visual obstruction measurement would indicate the vegetation is both tall and dense, providing larger amounts of the preferred dense vegetation. Common yellowthroats were most likely associated with increasing forbs because they provided this tall, dense vegetation. Red-winged blackbirds were positively associated with maximum vegetation height in treatment fields and litter depth in reference fields. It was not unexpected that red-winged blackbirds would be associated with increasing vegetation height in treatment fields since they frequently selected the tallest vegetation for nest placement (Lucas Negus, personal observation). The tall vegetation also provided prominent perch sites for males to display and to alert others when potential predators are near (Yasukawa and Searcy 1995). In reference fields, red-winged blackbirds were quite often observed in transects parallel to or crossing valleys or water-ways in fields. These valleys and water- 34

44 ways generally consisted of dense, lush stands of grass compared to other areas in the fields, presumably due to the higher amounts of moisture in these areas. Because of the density of the vegetation in these areas, litter accumulation was considerably higher in these areas than other areas of the field, explaining the positive association with litter depth in reference fields. Grasshopper sparrow presence in treatment fields was associated with increasing forb cover and litter depth, and decreasing vegetation height. This model fits well with observations reported by others. While habitat preferences of grasshopper sparrows vary by region, they generally prefer somewhat disturbed areas with short or intermediate vegetation height (Vickery 1996, Dechant et al. 2003). Whitmore (1981) noted grasshopper sparrows preferred sparse, patchy habitat and nested in vegetation clumps with high amounts of litter, and Schneider (1998) found increasing litter to be one of the strongest predictors of grasshopper sparrow presence in grasslands. Bobolinks were associated with increasing dead material cover and decreasing bare ground in treatment fields and increasing grass and dead material cover in reference fields. These predictive models seem to fit very well with the abundance surveys and observations from this study, as well as other studies. In New York, bobolink density was higher in areas with low legume cover and high amounts of litter and grass cover (Bollinger and Gavin 1992). Similar to Henslow s sparrows, bobolinks used undisked CRP more often than disked/interseeded CRP. With no disturbance, the undisked CRP had high amounts of litter, dead material, and grass cover. Sedge wrens were associated with 2 vegetation variables in treatment fields, increasing forb cover and litter depth. Sedge wrens prefer areas with tall, dense 35

45 vegetation (Johnson and Schwartz 1993b, Herkert et al. 2001). Delisle and Savidge (1997) also found sedge wrens were positively associated with litter depth in Nebraska. In treatment fields, the disked/interseeded portions had high amounts of forbs that likely provided the tall, dense vegetation that was preferred by sedge wrens rather than the undisked portions. Western meadowlarks in treatment fields were negatively associated with forb and grass cover, which is somewhat puzzling. Western meadowlarks occur in a diversity of grassland habitats, but are generally thought to prefer areas of short, less dense grasses (Lanyon 1994, McCoy 1996). One explanation may be that the forbs were too tall and dense in disked portions of treatment fields. In a study on Montana grasslands, Dieni and Jones (2003) reported that western meadowlarks avoided nesting in patches with high percentages of forb cover. Another explanation may be that landscape characteristics of treatment fields such as patch size, core area, and edge area, which were not included in the model, were responsible for the presence of western meadowlarks rather than the measured vegetation characteristics that were used to predict presence. Perhaps the best explanation for the model comes from Wiens and Rotenberry (1981), who reported that western meadowlarks did not correlate well with any habitat characteristics. Vegetation characteristics also likely influenced nest densities in my study fields. The nest density in treatment plots was nearly 3 times the nest density in reference plots. Additionally, the nest plots in the undisked portion of the treatments fields had very low densities (0.77 nests/ha), while the plots in disked/interseeded portions had moderate to very high nest densities ( nests/ha). A further evaluation of nest densities by years post-treatment may better represent the differences in nest density. The density of 36

46 nests peaked in the 1-year post-treatment plots and decreased in all other stages of treatment (no treatment, current-year, and 2-year post-treatment) (Figure 3). This trend is due to the high number of red-winged blackbird and dickcissel nests that were observed nearly exclusively in the disked/interseeded portions of treatment fields. As mentioned above, the vegetation characteristics of disked/interseeded portions of treatment fields were nearly ideal for dickcissels, and red-winged blackbirds were attracted to the same characteristics. Dickcissels and red-winged blackbirds are both polygynous and as many as 15 red-winged blackbird females have been reported nesting in 1 male territory, resulting in high nest densities (Zimmerman 1966, Yasukawa and Searcy 1995). Additionally, it appeared both species re-nested in the same area soon after a nest failure rather than leaving the disked/interseeded portions of fields (Lucas Negus, personal observation). This is somewhat contrary to that reported by Zimmerman (1982), who reported only 17% of dickcissel females re-nested in the same territory they originally nested in. This use of the same field portions for re-nesting may have also contributed to the high nest density in the disked/interseeded portions of treatment fields. Although differences in nest density between treatment and reference fields were detected, there were no differences in nest success probability between treatment and reference fields. The vegetation characteristics of treatment fields positively influenced both abundance and nest density of many grassland bird species, but appear to not enhance the nest success. The lack of difference in nest success suggests that vegetation characteristics that were attractive for nesting did not offer more protection from predators than reference field vegetation. 37

47 As mentioned, there was no difference in nest success probability between treatment and reference fields, but nest success probability of all species combined and red-winged blackbirds in was only half of that in Possible factors that may have caused the decline in nest success include climate/weather differences, increased search effort during the field season, increased nest densities, increased predator numbers, and improved predator search images. Climate/weather differences were not likely responsible; the weather conditions in 2004 and were similar and nest failure due to weather was similar between the 2 years. In, I hired 2 additional field technicians that increased our nest searching efforts. Although we had more people in the fields in, the same methods for finding and checking nests were used as were used the previous summer, so there should not have been an increase in disturbance affecting nest success probability in. Although previous research is inconclusive, nest density may be correlated to depredation rates. Robertson (1972) reported predation rates decreased with increasing nest density, while Fretwell (1977) reported increased predation rates with increased density. Zimmerman (1984) found no correlation between dickcissel nest density and predation rates. Regardless, nest densities did not differ between 2004 and, thus nest densities likely had little influence on the nest success probability difference between the 2 years. Because the percentage of nests that were depredated increased from 64% in 2004 to 75% in, it is possible that predators may have increased in abundance or developed better search images in the field season. Although nest success probability was similar between treatment and reference fields, more nests were 38

48 depredated in treatment fields simply because there were many more nests (primarily redwinged blackbird and dickcissel nests) in these fields. Because treatment fields were first disked/interseeded in 2004, it was the first time a large number of red-winged blackbirds and dickcissels nested in these CRP fields. Therefore, it is possible that predator populations may have been low or had not developed the search image to find these particular nests. After the influx of nests in 2004, predator populations may have increased or predators may have developed a search image for nests by the field season. Others have suggested yearly predation differences are due to fluctuations in predator populations (Davis 2003). In a study on Saskatchewan grasslands, Davis (2003) partially attributed abnormally low nest success in 1997 to an influx of prairie voles (Microtus ochrogaster) during that year. While the prairie voles were not always the nest predator, they likely also attracted other predators that foraged opportunistically on grassland bird nests. Additionally, predators select prey from a large prey community that includes grassland birds. Annual changes in grassland bird nest predation may be reflective of changes in the overall prey community. Schmidt (1999) predicted that increasing abundance of alternative foods (prey) would decrease grassland bird nest depredation. A study on artificial nest success in CRP fields by Vander Lee et al. (1999) supported this prediction, reporting increased nest success in plots in which supplemental prey was supplied. While I did not evaluate changes in the prey community during this study, it is likely that the prey community, like the bird and vegetation communities, changed over the years of the study due to the disking/interseeding. 39

49 Overall, 71% of the nests were depredated over the 2 years of this study. This is similar to findings of other studies in the Midwest (Patterson and Best 1996, Winter 1999). In my study, few of the nests were disturbed or destroyed, suggesting the predators involved were most likely snakes or small mammals (Thompson et al. 1999, Pietz and Granfors 2000). A depredated field sparrow nest was observed with 1 dead, decapitated young, suggesting a mouse likely depredated the nest (Pietz and Granfors 2000). Thirteen-lined ground squirrels (Spermophilus tridecemlineatus) and Franklin s ground squirrels (Spermophilus franklinii), known nest predators, were also frequently observed in the fields. Other nests were unsuccessful due to abandonment, weather, or unknown causes. Surprisingly, no nests were unsuccessful due to cowbird parasitism. Only a few nests were parasitized over the 2 years of the study, but all of these nests were depredated before the young fledged. This is surprising, however, because power lines, shelter belts, and farm buildings, all reported as perch fields for cowbirds (Johnson and Temple 1990, Schaffer et al. 2003), are abundant in the study area. Most likely, very few nests were parasitized simply because there were very few cowbirds in the area. My survey results indicated there was a very low abundance of cowbirds throughout the study area. Vegetation characteristics also may have affected nest success of individual species. For dickcissels, the only vegetation difference between successful and unsuccessful nests occurred in. Contrary to what was expected, maximum vegetation height around unsuccessful nests was higher than that around successful nests. It is expected that higher vegetation around nests would provide better nest concealment, increasing the probability of nest success. However, this pattern may indicate predators 40

50 that use olfactory senses or audio cues rather than visual cues, such as snakes, were responsible for depredation. Zimmerman (1984) concluded in his study on dickcissels in Kansas that snakes were most likely a major predator on nests and nest density or concealment of nests were not important factors influencing the rate of predation. Instead, snake density was likely the most important factor influencing depredation. In my study, most depredated nests were not disturbed or destroyed indicating snakes could have been the primary nest predator (Thompson et al. 1999). Nest success of the 2 most abundant ground-nesting species (grasshopper sparrows and bobolinks) did not appear to be significantly affected by vegetation. Successful grasshopper sparrow nests had more bare ground surrounding them than unsuccessful nests. Whitmore (1981) reported grasshopper sparrow territories had higher amounts of bare ground than areas not used for territories. Grasshopper sparrows generally nested in clumps of vegetation, usually surrounded by bare ground. This also indicates grasshopper sparrows may be more successful in areas that have a diversity of grasses, forbs, and bare ground, rather than an evenly distributed monoculture of vegetation. Bobolink nests exhibited no vegetation differences between successful and unsuccessful nests. Bobolinks primarily nested in undisked CRP, where there were few forbs and vegetation was uniform throughout the field. Contrary to other studies, bobolinks in my study concealed their nests in litter surrounded by evenly-distributed grasses and not in clumps of forbs or grasses (Martin and Gavin 1995). This is likely the reason there were no vegetation differences between successful and unsuccessful bobolink nests. CONSERVATION IMPLICATIONS AND RECOMMENDATIONS 41

51 As indicated by my results and those of many others, CRP fields provide critical habitat for grassland breeding birds (Johnson and Schwartz 1993b, Igl and Johnson 1995, Best et al. 1997, Delisle and Savidge 1997). However, as CRP fields age without management, their benefits to grassland birds may decrease (Millenbah et al. 1996). Aging CRP fields can be enhanced using several management methods including haying, burning, grazing, and disking/interseeding legumes (McCoy et al. 2001b). In my study, disking/interseeding legumes was promoted primarily because of the diverse vegetation created in CRP fields dominated by smooth brome compared to other practices. Although the costs of management vary by region and practice, disking/interseeding is generally more expensive than the other practices, with fuel costs contributing greatly. However, if entire CRP fields are hayed or grazed, a 25% reduction in CRP payments is imposed (U. S. Department of Agriculture 2003a). Depending on land rental rates, this reduction in CRP payments may make haying or grazing more costly than disking/interseeding. Burning is the most cost effective of the practices, but is rarely used in the region because of lack of burn crews, landowner experience, and social factors (fear of burning). Additionally, burning smooth brome does not create the vegetation diversity that can be created using disking/interseeding (Scott Wessel, Nebraska Game and Parks Commission, personal communication). According to my results, disking/interseeding legumes benefited many grassland bird species in several ways. Disked/interseeded fields accommodated higher abundances, more species, and higher diversities of grassland birds than the reference fields. The manner in which the fields were disked/interseeded seemed to create a range of vegetation characteristics, or successional stages, that were desirable to a wide array of 42

52 grassland bird species. Disking/interseeding produced the weedy, sparse vegetation that is similar to that of newly seeded CRP (Millenbah et al. 1996). In the first growing season following disking, typical vegetation characteristics included little to no litter, high amounts of bare ground, sparse and short stature legumes (alfalfa and red clover) early in the season, moderately dense and medium height planted legumes later in the season, with patches of unplanted forbs (sunflowers, ragweed, hemp, and thistles) occurring throughout the field (Lucas Negus, personal observation). Grasses present in the first growing season were typically annuals such as foxtails (Setaria spp.). Vegetation in the 1-year post-treatment portions was predominantly tall, dense planted legumes. Yellow sweetclover provided tall vegetation stature, whereas the red clover and alfalfa filled in the lower levels of the vertical structure. There were few other forbs and grasses present after 1 full growing season. However, many unplanted forb stalks remained from the previous season and were frequently used for perching. In the 2-years post treatment portions, the vegetation consisted mainly of tall smooth brome, with sparse alfalfa and red clover interspersed in the lower levels of the vegetation. There was little bare ground in the third year after disking. Reference fields were typically dominated by short to medium height smooth brome, few or no forbs, moderate amounts of standing dead material, high amounts of litter, and no bare ground. Vegetation characteristics in reference fields did not change during the 2 years of the study. It is likely factors other than vegetation, such as patch characteristics and invertebrate communities, had impacts on the grassland bird community in the CRP fields. Core area, field or patch size, edge area, and distance to edge have been reported to affect grassland bird communities throughout the Midwest (Herkert 1994, Helzer and 43

53 Jelinski 1999, Winter and Faaborg 1999, Sporrong 2004). While I did not examine patch characteristics such as field or patch size and edge area, it is possible that these factors influenced the grassland bird communities in my study. Moreover, it is also possible that disked/interseeded portions were too small to accommodate some area sensitive species such as grasshopper sparrows, bobolinks, and western meadowlarks (Helzer and Jelinski 1999, Winter and Faaborg 1999). Additionally, the increased amount of edge created in treatment fields by disking/interseeding portions of the fields may have reduced nest success due to increased predation. Winter et al. (2000) suggested that dickcissels and other habitat generalists are affected by decreased nest success near edges due to increased depredation (primarily by mesocarnivores) near these edges. However, nearly all edge studies focused on hard edges such as tree rows or roads, whereas the edges in my project were soft edges, representing changes in herbaceous vegetation structure. Further research on the effects of soft edges on grassland birds is needed. I attempted as best as I could to control for field size by selecting similar sized reference and treatment fields, as well as selecting treatment fields in which portions were disked/interseeded in similar sizes and shapes, but ultimately could not select fields that were identical in size and shape. Although not likely, these small differences in size and shape may have influenced bird communities in my treatment and reference fields. Invertebrates are important food resources for breeding grassland birds, as well as many other bird species (Hull et al. 1996). Several studies have reported that grassland birds opportunistically forage on invertebrates from the orders Coleoptera, Lepidoptera, Orthoptera, and Aranae (Kaspari and Joern 1993, Kobal et al. 1998, McIntyre and Thompson 2003). An examination of the invertebrate community in my study fields in 44

54 2004 indicated treatment fields provided higher abundances and biomasses of these orders than reference fields (Appendix C). Grassland birds are likely not only responding to the vegetation differences in the treatment fields, but also to the invertebrate community that is influenced by the vegetation changes. To maximize vegetation and grassland bird diversity in CRP fields, I recommend using disking/interseeding legumes as a management tool. Moreover, I recommend establishing an annual rotation of this management practice to maximize benefits. An annual rotation would create a mosaic of vegetation characteristics, which would provide habitat for several species simultaneously (Herkert 1996, Madden et al. 2000, Winter et al. ). Ideally, fields could be divided into fourths, with one-fourth of the field being disked/interseeded annually. To accommodate the suite of grassland birds, I also recommend maintaining one-fourth of the field out of the annual rotation to provide the mature grassland habitat that is preferred by several species (e.g., Henslow s sparrows and bobolinks) (Herkert 1996). Additionally, fields that are difficult to disk/interseed because of rough terrain, poor accessibility, or other reasons should not be disked/interseeded, thus providing important habitat to those species that may be displaced by disking/interseeding. This study was relatively short in duration, and the benefits of disking/interseeding beyond 2 years after the initial disking/interseeding are unknown. Petersen and Best (1999) concluded that 2-3 years of post-disturbance data for perturbation experiments may be insufficient for grassland birds. Vegetation characteristics indicate that in the third year after disking/interseeding, vegetation becomes more similar to that in reference fields. It can be predicted that the grassland 45

55 bird community will likely reflect these vegetation changes as the field matures. However, exactly how long it takes for grassland bird communities in treatment fields to become similar to those of reference fields is not known. Researching the disked/interseeded fields for several years (> 5) following management would yield more confident conclusions. Future grassland bird studies in CRP should not only focus on vegetation characteristics, but should also focus on the landscape characteristics and invertebrate communities as mentioned previously. Additionally, studies should be long-term (> 3 years) (Petersen and Best 1999) and cover a wide geographic range (Winter et al. ). Grassland habitats vary highly between years and regions, as do grassland bird populations. Thus it is important that studies extend over wide geographic regions and years to accurately measure grassland bird communities (Igl and Johnson 1999). These types of studies are important to better understand grassland bird populations and trends in the future, especially as required management of CRP becomes more common and widespread. Future studies on management practices, specifically disking/interseeding, should also examine source-sink population implications. Although disked/interseeded portions had higher abundances and nest densities than undisked portions, they may be acting as sink habitats rather than source habitats. Finally, future studies of CRP vegetation management should focus on techniques that will increase the longevity of the vegetation created by disking/interseeding. Disking/interseeding legumes may be more effective when used in conjunction with other management techniques such as herbicide application, burning, haying, or grazing. LITERATURE CITED 46

56 Best, L. B., H. Campa, III, K. E. Kemp, R. J. Robel, M. R. Ryan, J. A. Savidge, H. P. Weeks, Jr., and S. R. Winterstein Bird abundance and nesting in CRP fields and cropland in the Midwest: a regional approach. Wildife Society Bulletin 25: Avian abundance in CRP and crop fields during winter in the Midwest. American Midland Naturalist 139: Bollinger, E. K., P. B. Bollinger, and T. A. Gavin Effects of cropping on eastern populations of the bobolink. Wildlife Society Bulletin 18: , and T. A. Gavin Eastern bobolink populations: ecology and conservation in an agricultural landscape. Pages in J. M. Hagan III and D. W. Johnston, editors, Ecology and conservation of neotropical migrant landbirds, Smithsonian Institute Press, Washington, D. C., USA. Brower, J. E., and J. H. Zar Field and laboratory methods for general ecology. William C. Brown Publishers, Dubuque, Iowa, USA. Cody, M. L On the methods of resource division in grassland bird communities. American Midland Naturalist 102: Conover, W. J., and R. L. Iman Rank transformations as a bridge between parametric and nonparametric statistics. American Statistician 35: Daubenmire, R. F A canopy coverage method of vegetational analysis. Northwest Science 33: Davis, S. K., and S. G. Sealy Cowbird parasitism and nest predation in fragmented grasslands of southwestern Manitoba. Pages in J. N. Smith, 47

57 T. L. Cook, S. I. Rothstein, S. K. Robinson, and S. G. Sealy, editors, Ecology and management of cowbirds and their hosts. University of Texas Press, Austin, USA Nesting ecology of mixed-grass prairie songbirds in southern Saskatchewan. Wilson Bulletin 115: Dechant, J. A., M. L. Sondreal, D. H. Johnson, L. D. Igl, C. M. Goldade, M. P. Nenneman, and B. R. Euliss Effects of management practices on grassland birds: Grasshopper Sparrow. Northern Prairie Wildlife Research Center, Jamestown, ND. Jamestown, ND: Northern Prairie Wildlife Research Center Home Page. (Version 12AUG2004). Accessed 5 December,. Delisle, J. M., and J. A. Savidge Avian use and vegetation characteristics of Conservation Reserve Program fields. Journal of Wildlife Management 61: Dieni, J. S., and S.L. Jones Grassland songbird nest site selection patterns in northcentral Montana. Wilson Bulletin 115: Dow, D. D A test of significance for Mayfield s method of calculating nest success. Wilson Bulletin 90: Frawley, B. J., and L. B. Best Effects of mowing on breeding bird abundance and species composition in alfalfa fields. Wildlife Society Bulletin 19: Fretwell, S. D Is the dickcissel a threatened species? American Birds 31:

58 Guiliano, W. M., and S. E. Daves Avian responses to warm-season grass use in pasture and hayfield management. Biological Conservation 106:1-9. Hammond, C. L Soil Survey of Stanton County Nebraska. United States Department of Agriculture- Soil Conservation Service in cooperation with University of Nebraska, Conservation and Survey Division. Hays, R. L., R. P. Webb, and A.H. Farmer Effects of the Conservation Reserve Program on wildlife habitat: results of 1988 monitoring. Transactions of the North American Wildlife and Natural Resources Conference 54: Helzer, C. J The effects of wet meadow fragmentation on grassland birds. Masters Thesis, University of Nebraska, Lincoln., and D. E. Jelinski The relative importance of patch area and perimeter-area ratio to grassland breeding birds. Ecological Applications 9: Herkert, J. R The effects of habitat fragmentation on Midwestern grassland bird communities. Ecological Applications 4: An analysis of Midwestern breeding bird population trends: American Midland Naturalist 134: Population trends of the Henslow s sparrow in relation to the Conservation Reserve Program in Illinois, Journal of Field Ornithology 68: The influence of the CRP on grasshopper sparrow population trends in the mid-continental United States. Wildlife Society Bulletin 26:

59 , and W. D. Glass Henslow s sparrow response to prescribed fire in an Illinois prairie remnant. Studies in Avian Biology 19: , D. E. Kroodsma, and J. P. Gibbs Sedge wren (Cistothorus platensis). No. 582 in A. Poole and F. Gill, editors, The birds of North America, The Birds of North America, Inc., Philadelphia, Pennsylvania, USA., P. D. Vickery, and D. E. Kroodsma Henslow s sparrow (Ammomdramus henslowii). No. 672 in A. Poole and F. Gill, editors, The birds of North America, The Birds of North America, Inc., Philadelphia, Pennsylvania, USA., D. L. Reinking, D. A. Wiedenfeld, M. Winter, J. L. Zimmerman, W. E. Jensen, E. J. Finck, R. R. Koford, D. H. Wolfe, S. K. Sherrod, M. A. Jenkins, J. Faaborg, and S. K. Robison Effects of prairie fragmentation on the nest success of breeding birds in the Midcontinental United States. Conservation Biology 17: Horn, D. J., and R. R. Koford Relation of grassland bird abundance to mowing of Conservation Reserve Program fields in North Dakota. Wildife Society Bulletin 28: Hosmer, D. W., Jr., and S. Lemeshow Applied logistic regression. John Willey and Sons, New York, New York, USA. Hughes, J. P., R. J. Robel, K. E. Kemp, and J. L. Zimmerman Effects of habitat on dickcissel abundance and nest success in Conservation Reserve Program fields in Kansas. Journal of Wildlife Management 63:

60 Hull, S. D., R. J. Robel, and K. E. Kemp Summer avian abundance, invertebrate biomass, and forbs in Kansas CRP. Prairie Naturalist 28:1-12. Igl, L. D., and D. H. Johnson Dramatic increase of Le Conte s Sparrow in Conservation Reserve Program fields in northern Great Plains. Prairie Naturalist 27:89-94., and Le Conte s sparrows breeding in Conservation Reserve Program fields: precipitation and patterns of population change. Studies in Avian Biology 19: Jackson, J. R., G. A. Hurst, and E. A. Gluesing Abundance and selection of invertebrates by northern bobwhite chicks. Proceedings Annual Conference Southeast Association Fish and Wildlife Agencies 41: Johnson, D. H Estimating nest success: the Mayfield method and an alternative. Auk 96: , and L. D. Igl Area requirements of grassland birds: a regional perspective. Auk 118:24-34., and M. D. Schwartz. 1993a. The Conservation Reserve Program and grassland birds. Conservation Biology 7: , and M. D. Schwartz. 1993b. The Conservation Reserve Program: habitat for grassland birds. Great Plains Research 3: Johnson, R. G., and S. A. Temple Assessing habitat quality for birds nesting in fragmented tallgrass prairies. Pages in J. Verner, M.L. Morrison, and C. 51

61 J. Ralph, editors. Modeling habitat relationships of terrestrial vertebrates. University of Wisconsin Press, Madison, USA., and S. A. Temple Nest predation and brood parasitism of tallgrass prairie birds. Journal of Wildlife Management 54: Kaspari, M., and A. Joern Prey choice by three insectivorous grassland birds: reevaluating opportunism. Oikos 68: King, J. W., and J. A. Savidge Effects of the Conservation Reserve Program on wildlife in southeast Nebraska. Wildlife Society Bulletin 23: Knopf, F. L., and F. B. Samson Conservation of grassland vertebrates. Pages in F. L. Knopf and F. B. Samson, editors. Ecology and conservation of Great Plains vertebrates. Springer-Verlag, New York, USA. Kobal, S. N., N. F. Payne, and D. R. Ludwig Nestling food habits of 7 grassland bird species and insect abundance in grassland habitats in northern Illinois. Transactions of the Illinois State Academy of Science 91: Koford, R. R., B. S. Bowen, J. T. Lokemoen, and A. D. Kruse Cowbird parasitism in grassland and cropland in the northern Great Plains. Pages in J. N. M. Smith, T. L. Cook, S. I. Rothstein, S. K. Robison, and S. G. Sealy, editors. Ecology and management of cowbirds and their hosts. University of Texas Press, Austin, USA. Krebs, C. J Ecological methodology, second edition. Addison-Welsey Educational Publishers, Inc., Menlo Park, California, USA. 52

62 Lanyon, W. E Western Meadowlark (Sturnella neglecta). No. 104 in A. Poole and F. Gill, editors, The birds of North America, The Academy of Natural Sciences, Philadelphia, Pennsylvania, and The American Ornithologists Union, Washington, D.C., USA. Leathers, R. J Relative invertebrate availability in Nebraska s Conservation Reserve Program Management Access Program. Master s Thesis. South Dakota State University, Brookings. Luttschwager, K. A., and K. F. Higgins Nongame bird, game bird, and deer use of Conservation Reserve Program fields in eastern South Dakota. Proceedings of South Dakota Academy of Science 71: Madden, E. M., R. K. Murphy, A. J. Hansen, and L. Murray Models for guiding management of prairie bird habitat in northwestern North Dakota. American Midland Naturalist 144: Manley, S. W., R. S. Fuller, J. M. Lee, and L. E. Brennan Arthropod response to strip disking in old fields managed for bobwhites. Proceedings Annual Conference Southeast Association Fish and Wildlife Agencies 48: Martin, S. G., and T. A. Gavin Bobolink (Dolichonyx oryzivorus). No. 176 in A. Poole and F. Gill, editors, The birds of North America, The Academy of Natural Sciences, Philadelphia, Pennsylvania, and The American Ornithologists Union, Washington, D.C., USA. Martin, T. E., and G. R. Geupel Nest-monitoring plots: methods for locating nests and monitoring success. Journal of Field Ornithology 64:

63 Mayfield, H. F Suggestions for calculating nest success. Wilson Bulletin 87: McCoy, T. D Avian abundance, composition, and reproductive success on Conservation Reserve Program fields in northern Missouri. Master s Thesis. University of Missouri, Columbia., M. R. Ryan, E. W. Kurzejeski, and L. W. Burger, Jr Conservation Reserve Program: Source or sink habitat for grassland birds in Missouri? Journal of Wildlife Management 63: , M. R. Ryan, L. W. Burger, Jr., and E. W. Kurzejeski. 2001a. Effects of conservation practice, mowing, and temporal changes on vegetation structure on CRP fields in northern Missouri. Wildlife Society Bulletin 29: , E. W. Kurzejeski, L. W. Burger, Jr., and M. R. Ryan. 2001b. Grassland bird conservation: CP1 vs. CP2 plantings in Conservation Reserve Program fields in Missouri. American Midland Naturalist 145:1-17. McIntyre, N. E., and T. R. Thompson A comparison of Conservation Reserve Program habitat plantings with respect to arthropod prey for grassland birds. American Midland Naturalist 150: Millenbah, K. F The effects of different age classes of fields enrolled in the Conservation Reserve Program in Michigan on avian diversity, density, and productivity. Master s Thesis, Michigan State University, East Lansing. 54

64 , S. R. Winterstein, H. Campa III, L. T. Furrow, and R. B. Minnis Effects of Conservation Reserve Program field age on avian relative abundance, diversity, and productivity. Wilson Bulletin 108: Natural Resource Conservation Service Practice specification-early successional habitat development/management-disking. Nebraska Field Office Technical Guide 525. S-647a Nebraska s major land resource areas. Accessed 29 April Patterson, M. P., and L. B. Best Bird abundance and nesting success in Iowa CRP fields: the importance of vegetation structure and composition. American Midland Naturalist 135: Petersen, K. L., and L. B. Best Design and duration of perturbation experiments: implications for data interpretation. Studies in Avian Biology 19: Pietz, P. J., and D. A. Granfors Identifying predators and fates of grassland passerine nests using miniature video cameras. Journal of Wildlife Management 64: Pruitt, L Henslow s sparrow status assessment. U. S. Fish and Wildlife Service, Bloomington Field Office, Region 3, Bloomington, Indiana, USA. Pulliam, H. R Sources, sinks, and population regulation. American Midland Naturalist 132:

65 Reynolds, R. E., T. L. Shaffer, J. R. Sauer, and B. G. Peterjohn Conservation Reserve Program: benefit for grassland birds in the northern plains. Transactions of the North American Wildlife and Natural Resources Conference 59: Robel, R. J., J. N. Briggs, A. D. Dayton, and L. C. Hulbert Relationships between visual obstruction measurements and weight of grassland vegetation. Journal of Range Management 23: Robertson, R. J Optimal niche space of the redwinged blackbird (Agelaius phoeniceus). I. Nesting success in marsh and upland habitat. Canadian Journal of Zoology 50: Rotella, J. J., E. M. Madden, and A. J. Hansen Sampling considerations for estimating density of passerines in grasslands. Studies in Avian Biology 19: Ryan, M. R., L. W. Burger, and E. W. Kurzejeski The impact of CRP on avian wildlife: a review. Journal of Production Agriculture 11: Samson, F. B., and F. L. Knopf Prairie conservation in North America. BioScience 44: SAS Institute SAS 9.1 for windows. SAS Institute Inc., Cary, North Carolina, USA. Sauer, J. R., J. E. Hines, and J. Fallon.. The North American breeding bird survey, results and analysis Version.2. USGS Patuxent Wildlife Research Center, Laurel, Maryland, USA. 56

66 Schmidt, K. A Foraging theory as a conceptual framework for studying nest predation. Oikos 85: Schneider, N. A Passerine use of grasslands managed with two grazing regimes on the Missouri Coteau in North Dakota. Masters Thesis, South Dakota State University, Brookings, South Dakota. Shaffer, J. A., C. M. Goldade, M. F. Dinkins, D. H. Johnson, L. D. Igl, and B. R. Euliss Brown-headed cowbirds in grasslands: their habitats, hosts, and response to management. Prairie Naturalist 35: Sporrong, J. M Response of nongame birds and terrestrial invertebrates to restoration of upland grasslands in the Rainwater Basin Region, Nebraska. Masters Thesis, Oklahoma State University, Stillwater. Stewart, R. E A life history study of the yellow-throat. Wilson Bulletin 65: Steinauer, E. M., and S. L. Collins Prairie ecology-the Tallgrass Prairie. Pages in F. B. Samson and F. L. Knopf, editors. Prairie conservation-preserving North America s most endangered ecosystem. Island Press, Washington DC., USA. Swanson, D. A., D. P. Scott, and D. L. Risley Wildlife benefits of the Conservation Reserve Program in Ohio. Journal of Soil and Water Conservation 54: Effects of management practices on grassland birds: savannah sparrow. Northern Prairie Wildlife Research Center, Jamestown, ND. 57

67 Jamestown, ND: Northern Prairie Wildlife Research Center Home Page. (Version 12DEC2003). Accessed 29 April Taylor, S Focus on pheasants- proposed activities. Nebraska Game and Parks Commission. Unpublished report. Lincoln, Nebraska, USA. Temple, S. A., B. M. Fevold, L. K. Paine, D. J. Undersander, and D. W. Sample Nesting birds and grazing cattle: accommodating both on midwestern pastures. Studies in Avian Biology 19: Dickcissel (Spiza americana). No. 703 in A. Poole and F. Gill, editors, The birds of North America, The Birds of North America, Inc., Philadelphia, Pennsylvania, USA. Thompson, F. R., III, W. Dijak, and D. E. Burhans Video identification of predators at songbird nests in old fields. Auk 116: U.S. Department of Agriculture. 2003a Farm Bill Conservation Reserve Program long-term policy; interim rule. Federal Register, 7 CFR Part b. The Conservation Reserve Program statistics. Accessed 29 April Vander Lee, B. A., R. S. Lutz, L. A. Hansen, and N. E. Mathews Effects of supplemental prey, vegetation, and time on success of artificial nests. Journal of Wildlife Management 63:

68 Vickery, P. D Grasshopper sparrow (Ammodramus savannarum). No. 239 in A. Poole and F. Gill, editors, The birds of North America, The Academy of Natural Sciences, Philadelphia, Pennsylvania, and The American Ornithologists Union, Washington, D.C., USA. Whitmore, R. C Structural characteristics of grasshopper sparrow habitat. Journal of Wildlife Management 45: Whitmore, R. W., K. P. Pruess, and R. E. Gold Insect food selection by 2-weekold ring-necked pheasant chicks. Journal of Wildlife Management 50: Wiens, J. A., and J. T. Rotenberry Habitat associations and community structure of birds in shrub steppe environments. Ecological Monographs 51: Nesting biology of dickcissels and Henslow s sparrows in southwestern Missouri prairie fragments. Wilson Bulletin 111: , and J. Faaborg Patterns of area sensitivity in grassland-nesting birds. Conservation Biology 13: , D. H. Johnson, and J. Faaborg Evidence for edge effects on multiple levels in tallgrass prairie. Condor 102: , S. E. Hawks, J. A. Shaffer, and D. H. Johnson Guidelines for finding nests of passerine birds in tallgrass prairie. Prairie Naturalist 35: , D. H. Johnson, and J. A. Shaffer.. Variability in vegetation effects on density and nesting success of grassland birds. Journal of Wildlife Management 69:

69 Yasukawa, K., and W. A. Searcy Red-winged blackbird (Agelaius phoeniceus). No. 184 in A. Poole and F. Gill, editors, The birds of North America, The Academy of Natural Sciences, Philadelphia, Pennsylvania, and The American Ornithologists Union, Washington, D.C., USA. Zar, J. H Biostatistical analysis, fourth edition. Prentice-Hall, Upper Saddle River, New Jersey, USA. Zimmerman, J. L Polygyny in the dickcissel. Auk 83: Zimmerman, J. L Nesting success of dickcissels (Spiza americana) in preferred and less preferred habitats. Auk 99: Nest predation and its relationship to habitat and nest density in dickcissels. Condor 86: Breeding season habitat selection by the Henslow s sparrow (Ammodramus henslowii) in Kansas. Wilson Bulletin 100: Density-independent factors affecting the avian diversity of the tallgrass prairie community. Wilson Bulletin 104: Avian community responses to fire, grazing, and drought in the tallgrass prairie. Pages in F. L. Knopf and F. B. Samson, editors. Ecology and conservation of Great Plains vertebrates. Springer-Verlag, New York, USA. 60

70 Table 1. Overall relative abundance a, species richness b, and species diversity c of breeding grassland birds in treatment and reference fields in Stanton County, Nebraska, Treatment (n = 8) Reference (n = 8) Year Mean S.E. Mean S.E. Treatment Year P d Overall abundance (no./transect) Species richness Species diversity < < a b Mean abundance of all bird species from 3 sampling periods during each year. Number of avian species observed in each field. c Shannon-Weiner species diversity index (Krebs 1999). d P = P-value for treatment (treatment vs. reference) and year (2004 vs. ) effects from a 2-way analysis of variance. Interaction effects (treatment x year) were observed for species richness (F 1,28 = 4.47, P = 0.044). Means for species richness are reported separately for each year. 61

71 Table 2. Relative abundance a (birds/transect) of breeding grassland birds that were observed in >1% of surveys in treatment and reference fields in Stanton County, Nebraska, Treatment (n = 8) Reference (n = 8) Species b Year Mean S.E. Mean S.E. Treatment Year P c DICK <0.001 GRSP <0.001 BOBO RWBL COYE SEWR WEME 2004 HESP < < < < a Mean abundance from 3 sampling periods each year. b DICK = dickcissel, RWBL = red-winged blackbird, BOBO = bobolink, GRSP = grasshopper sparrow, COYE = common yellowthroat, SEWR = sedge wren, WEME = western meadowlark, and HESP = Henslow s sparrow. c P = P-value for treatment (treatment vs. reference) and year (2004 vs. ) effects from a 2-way analysis of variance. Interaction (treatment x year) effects occurred for 62

72 dickcissels (F 1,260 = 5.13, P = 0.024), grasshopper sparrows (F 1,260 = 8.55, P = 0.004), and western meadowlarks (F 1,260 = 5.29, P = 0.022). 63

73 Table 3. Relative abundance a (birds/transect) of breeding grassland birds occurring in >1% of surveys in 3 portions of Conservation Reserve Program fields managed by disking/interseeding in Stanton County, Nebraska, Disked-2003 (n = 8) Disked-2004 (n = 8) Undisked (n = 8) Species c Year Mean S.E. Mean S.E. Mean S.E. All Species 2004 DICK a b 3.88b 4.25a b 6.08a 0.33b b 2.71b 0.83b b a c 0.19 RWBL a b b 0.09 BOBO GRSP a a a b b a b a b b ab a 0.20 COYE a b b b a c 0.00 a Mean abundance from 3 sampling periods each year. b Means in rows with different letters were different (P < 0.05). c DICK = dickcissel, RWBL = red-winged blackbird, BOBO = bobolink, GRSP = grasshopper sparrow, and COYE = common yellowthroat. d P = P-value for treatment (disked 2003 vs. disked 2004 vs. undisked) effects from a 2- way analysis of variance. Treatment x year interaction effects occurred in all species combined (F 2, 138 = 16.76, P < 0.001), dickcissels (F 2,138 = 47.98, P < 0.001), red-winged blackbirds (F 2,138 = 10.41, P < 0.001), grasshopper sparrows (F 2,138 = 5.77, P = 0.004), 64

74 and common yellowthroats (F 2,138 = 10.12, P < 0.001). Means are reported separately for each year. 65

75 Table 4. Number of nesting species, number of nests, and nest density (nests/ha) of grassland birds in treatment and reference fields in Stanton County, Nebraska, Treatment (n = 6) Reference (n = 6) Year Mean S.E. Mean S.E. Number of species Number of nests Nest density (nests/ha)

76 Table 5. Nest success probabilities a for incubation, nestling, and overall nesting period for all bird species in treatment and reference fields in Stanton County, Nebraska, All bird species Treatment (n = 159) Incubation Nestling (%) (%) Overall (%) Reference (n = 30) Incubation Nestling (%) (%) Overall (%) P b a Determined using Mayfield (1975) method. b P = P-value for chi-square analysis of overall nest success probability (Dow 1978). 67

77 Table 6. Nest success probabilities a for all bird species, dickcissels, and red-winged blackbirds in Conservation Reserve Program fields in Stanton County, Nebraska, Species b All bird species Incubation (%) 2004 (n = 93) Nestling (%) Overall (%) Incubation (%) (n = 96) Nestling (%) Overall (%) DICK P c RWBL a Determined using Mayfield (1975) method. b DICK = dickcissel and RWBL = red-winged blackbird. c P = P-value for chi-square analysis of overall nest success probability (Dow 1978). 68

78 Table 7. Vegetation characteristics of successful and unsuccessful nests of all grassland bird species in Conservation Reserve Program fields in Stanton County, Nebraska, Successful (n = 60) Unsuccessful (n = 180) Vegetation Year Mean S.E. Mean S.E. Fate Year characteristic Nest height (cm) P a <0.001 Visual obstruction (dm) <0.001 Litter depth (cm) <0.001 Maximum vegetation height (cm) <0.001 Forb (%) Grass (%) Dead material (%) Bare ground (%) < < <0.001 a P = P-value for fate (successful vs. unsuccessful) and year (2004 vs. ) effects from 2-way analysis of variance. Interaction (fate x year) effects occurred in maximum vegetation height (F 1,236 = 8.19, P = 0.005), nest height (F 1,234 = 5.55, P = 0.019), and visual obstruction (F 1,236 = 5.48, P = 0.020). Means are reported separately for each year for these variables. 69

79 Table 8. Vegetation characteristics of successful and unsuccessful dickcissel nests in Conservation Reserve Program fields in Stanton County, Nebraska, Successful (n = 20) Unsuccessful (n = 62) Vegetation Year Mean S.E. Mean S.E. Fate Year characteristic Nest height (cm) P a Visual obstruction (dm) <0.001 Litter depth (cm) <0.001 Maximum vegetation height (cm) Forb (%) Grass (%) Dead material (%) Bare ground (%) a P = P-value for fate (successful vs. unsuccessful) and year (2004 vs. ) effects from 2-way analysis of variance. Interaction (fate x year) effects occurred in maximum vegetation height (F 1,78 = 6.26, P = 0.014) and dead material cover (F 1,78 = 4.09, P = 0.047). Means are reported separately for each year for these variables. 70

80 Table 9. Vegetation characteristics of successful and unsuccessful red-winged blackbird nests in Conservation Reserve Program fields in Stanton County, Nebraska, Successful (n = 18) Unsuccessful (n = 86) Vegetation Year Mean S.E. Mean S.E. Fate Year characteristic Nest height (cm) Visual obstruction (dm) P a Litter depth (cm) <0.001 Maximum vegetation height (cm) <0.001 Forb (%) Grass (%) Dead material (%) Bare ground (%) a P= P-value for fate (successful vs. unsuccessful) and year (2004 vs. ) effects from 2-way analysis of variance. No interaction (fate x year) effects occurred. 71

81 Table 10. Vegetation characteristics of successful and unsuccessful grasshopper sparrow nests in Conservation Reserve Program fields in Stanton County, Nebraska, Successful (n = 7) Unsuccessful (n = 12) Vegetation Mean S.E. Mean S.E. Fate characteristic Nest height (cm) P a Visual obstruction (dm) Litter depth (cm) Maximum vegetation height (cm) Forb (%) Grass (%) Dead material (%) Bare ground (%) a P = P-value for nest fate effects from 1-way analysis of variance. 72

82 Table 11. Vegetation characteristics of successful and unsuccessful bobolink nests in Conservation Reserve Program fields in Stanton County, Nebraska, Successful (n = 7) Unsuccessful (n = 8) Vegetation Mean S.E. Mean S.E. Fate characteristic Nest height (cm) P a Visual obstruction (dm) Litter depth (cm) Maximum vegetation height (cm) Forb (%) Grass (%) Dead material (%) Bare ground (%) a P = P-value for nest fate effects from 1-way analysis of variance. 73

83 Table 12. Vegetation characteristics of treatment and reference fields in Conservation Reserve Program fields in Stanton County, Nebraska, Vegetation characteristic Visual obstruction (dm) Treatement (n = 8) Reference (n = 8) Year Mean S.E. Mean S.E. Treatment Year <0.001 P a Litter depth (cm) < <0.001 Maximum vegetation height (cm) <0.001 Forb (%) Grass (%) < <0.001 <0.001 Dead material (%) < <0.001 Bare ground (%) < <0.001 a P = P-value for treatment (treatment vs. reference) and year (2004 vs. ) effects from 2-way analysis of variance. Interaction (treatment x year) effects occurred in horizontal visual obstruction (F 1,1040 = 8.65, P < 0.001), litter depth (F 1,1040 = 30.17, P < 0.001), maximum vegetation height (F 1,1040 = 61.70, P < 0.001), percent dead material (F 1,1040 = 25.20, P < 0.001), and percent bare ground (F 1,1040 = 47.13, P < 0.001). 74

84 Table 13. Vegetation characteristics of 3 portions of Conservation Reserve Program fields managed by disking/interseeding in Stanton County, Nebraska, Vegetation characteristics Visual obstruction (dm) Disked-2003 (n = 8) Disked-2004 (n = 8) Undisked (n = 8) Year Mean S.E. Mean S.E. Mean S.E a a c b a b c 0.09 Litter depth (cm) b 2.33b b 0.71c a 2.89a Maximum vegetation height (cm) a 81.79a c 80.72a b 58.83b Forb (%) a b c b a c 0.60 Grass (%) b c a a b a 1.64 Dead material (%) c 19.48b b 14.48c a 35.21a Bare ground (%) b 3.33b a 7.40a c 1.15c a Means in rows with different letters were different (P < 0.05). b P = P-value for treatment (disked 2003 vs. disked 2004 vs. undisked) effects from a 2- way analysis of variance. Treatment x year interaction effects occurred in all analyses (horizontal visual obstruction [F 2,570 = , P < 0.001]; litter depth [F 2,570 = 81.33, P < 0.001]; maximum vegetation height [F 2,570 = , P < 0.001]; percent forb [F 2,570 = 75

85 128.05, P < 0.001], grass [F 2,570 = 46.53, P < 0.001], dead material [F 2,570 = 18.59, P < 0.001], and bare ground [F 2,570 = 56.38, P < 0.001] cover). 76

86 Table 14. Logistic regression models for vegetation variables that best predicted grassland bird presence in treatment fields in Stanton County, Nebraska, Variables were selected from a set of vegetation variables using a backward-elimination routine. Species a Fitted logistic model bc P d DICK (forb) (vo) RWBL (max ht) GRSP (forb) (litter) 0.07 (max ht) BOBO (dead) (bare) COYE (forb) (max ht) WEME (forb) 0.07 (grass) SEWR (forb) (litter) HESP No Significant Model -- a DICK = dickcissel, RWBL = red-winged blackbird, BOBO = bobolink, GRSP = grasshopper sparrow, COYE = common yellowthroat, SEWR = sedge wren, WEME = western meadowlark, HESP = Henslow s sparrow. b P(presence) = 1/ (1 + exp{ - [b o + b 1 (x)]}) where P(presence) was the probability that a bird species was present, b o and b 1 were intercept and slope coefficients, and x was the predictor variable (vegetation variable), and Absence(x) = 1 Presence(x). c forb = forb cover, grass = grass cover, dead = dead material cover, bare = bare ground, vo = horizontal visual obstruction, max ht = maximum vegetation height, and litter = litter depth. d P = P-value for Hosmer and Lemeshow goodness-of-fit test of overall model. 77

87 Table 15. Logistic regression models for vegetation variables that best predicted grassland bird presence in reference fields in Stanton County, Nebraska, Variables were selected from a set of vegetation variables using a backward-elimination routine. Species a Fitted logistic model bc P d DICK No Significant Model -- RWBL (litter) GRSP No Significant Model -- BOBO (grass) (dead) COYE (max ht) WEME No Significant Model -- SEWR No Significant Model -- HESP No Significant Model -- a DICK = dickcissel, RWBL = red-winged blackbird, BOBO = bobolink, GRSP = grasshopper sparrow, COYE = common yellowthroat, SEWR = sedge wren, WEME = western meadowlark, HESP = Henslow s sparrow. b P(presence) = 1/ (1 + exp{ - [b o + b 1 (x)]}) where P(presence) was the probability that a bird species was present, b o and b 1 were intercept and slope coefficients, and x was the predictor variable (vegetation variable), and Absence(x) = 1 Presence(x). c forb = forb cover, grass = grass cover, dead = dead material cover, bare = bare ground, vo = horizontal visual obstruction, max ht = maximum vegetation height, and litter = litter depth. d P = P-value for Hosmer and Lemeshow goodness-of-fit test of overall model. 78

88 Figure 1. Location of study area (represented by the black box) in Stanton County, Nebraska. 79

89 NEBRASKA 80