.3 I}... $33. :1 or .5?“ m3 5%..3 r «.3. ; x :. .f , ...x¥sn..r> icy... . ’fi 15...? ll... 5.:- and 5.": 8:: .r .2... .w: .21”: c fiai4€~ RM...” - “a, . , , . ab.» an. . . . . “£2.35 #53. . , . 2 K , a; . . , u . .35... .. 9 ~ . (at? .. a unuuiurwffi . , . _ «nun. 35.3.3. r . . . at‘.t..u‘...cJ~.i . . r" :t 71 .7 Jr. H i. : 2 .ayafiru: r. . . . V 1.; . u. . . . . . . in“ V . 5.0; .- 1...: 7 :75.r\a 311‘ 31:53.22. Nun“: . . . . . 1!)! . . , .2... 3.3...» . _ . . . , , i. (K:- . . . 4 .153 1 .5 ‘ Rh 1. ....: y. . . a... . .L. . 3 3.: . L33... :3! waggmg§ufi4f “33.33.33.333“... 333:, 332- .3 33... .3. .. .. ,. .5 UNIVERSITY LIBRARIES WWIWWW | Hill 1mm 3 1293 010200 This is to certify that the thesis entitled The Influence of Field Age on Mammalian Relative Abundance, Diversity, and Distribution on Conservation Reserve Program Lands in Michigan presented by Ly Thi Furrow has been accepted towards fulfillment of the requirements for Master of Science degree in Fish. & Wildl. 'or profess r Date Nov. 17, 1994 0.7639 MS U is an Affirmative Action/Equal Opportunity Institution LIBRARY E Mlchigan State; Unlverslty‘] PLACE II RETURN BOX to romovothb chockouttrorn your record. To AVOID FINES Mum on or More data duo. DATE“ DATE DUE DATE DUE FEB 0.4 1999 I!“ KB ’5 van A 'n 0%! 1 6 2000 2 7 2010 0302!): MSUI-An‘“ ' .. -1 -- I:- .. . THE INFLUENCE OF FIELD AGE ON MAMMALIAN RELATIVE ABUNDANCE, DIVERSITY, AND DISTRIBUTION ON CONSERVATION RESERVE PROGRAM LANDS IN MICHIGAN BY Ly Thi Furrow A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Fisheries and Wildlife 1994 Henry Campa,III, Ph.D. ABSTRACT INFLUENCE OF FIELD AGE ON MAMMALIAN RELATIVE ABUNDANCE, DIVERSITY, AND DISTRIBUTION ON CONSERVATION RESERVE PROGRAM LANDS IN MICHIGAN BY Ly Thi Furrow Past research evaluating wildlife use of Conservation Reserve Program (CRP) lands have focused primarily on avian populations as indicators of wildlife habitat quality. In addition to avian species, mammals may also serve as indicators of wildlife habitat quality and have not been adequately evaluated on CRP lands. Relative small mammal abundance, species composition, diversity, and vegetative characteristics were examined on replicated CP1 fields of 6 age classes and on agricultural fields in Gratiot County, Michigan in 1992 and 1993. Additionally, predator scent stations were used to monitor medium sized mammals associated with CRP fields. Results suggest that the structure and composition of various age classes of CRP fields influenced mammal abundance, richness, and diversity. Reverting CRP lands to cropland may have significant impacts on a diversity of mammal species that depend on habitat conditions provided by these grasslands. ACKNOWLEDGEMENTS I owe my deepest gratitude to many individuals and organizations for making this research and completion of my thesis possible. First, I would like to thank the Michigan Agricultural Experiment Station, the Michigan Department of Natural Resources, Wildlife Division, and Michigan State University for funding this project. To my committee members, Dr. Henry Campa III, Dr. Scott Winterstein, and Dr. Donald Straney, I am forever grateful for your advice and valuable time. I would like to extend my appreciation and respect to my advisor, Dr. Henry Campa,III, for his guidance and encouragement throughout my research and thesis work. The knowledge I have gained from you is immeasurable. I would like to thank K. Millenbah, R. Minnis, A. Pearks, and interns K. O'Brien and M. Reynolds for their dedicated efforts in data collection. To K. Millenbah, R. Minnis, and A. Pearks, I thank you immensely for your patience, dedication, and above all your friendship. To M. Beirne, thank you for always being there. I extend my appreciation to the Department of Fisheries and Wildlife. There were so many people in the department who made it possible for me to complete this degree. iii Lastly, I want to acknowledge my family. To my parents, Tim and Tam, I extend my deepest admiration and love. I would not be where I am today without your unconditional love, guidance, and friendship. Thank you for always inspiring me to learn. Words cannot express how proud I am to be your daughter. To John, my best friend, I am forever grateful for your support, encouragement and patience. You have always believed in me and you helped me believe in myself. I would also like to thank Joao and Maria for allowing me to be a part of the family and giving me a place to call home. iv TABLE OF CONTENTS EQQE LIST OF TABLES.......................................... Vii LIST OF FIGURES............... ........ . ....... .......... xii INTRODUCTION......... ....... .......... .......... . ........ 1 OBJECTIVES....... ............ . ........................... 7 STUDY AREA...... ......... ................................ 8 METHODS................. ..... . ......... ........... ....... 13 Vegetative Sampling................................. 13 Small Mammal Trapping............................... 14 Predator Scent Stations............................. 16 Soil Moisture....................................... 17 Data Analysis... ...... . ............................. 18 RESULTSOOOOOOOOOOOOOOCOOOOOOOOOOOOOOOOOOOOOOOOOOO00...... 20 Vegetative Responses Comparisons of vegetative composition among age classes............................... 20 Comparisons of vegetative structural characteristics among age classes ......... 33 Comparisons of vegetative structural characteristics between months............ 42 Principle components analysis.................. 43 Small Mammal Responses.............................. 51 Small mammal relative abundance................ 51 Species richness............................... 54 Small mammal diversity......................... 56 Capture rates.................................. 60 Soil Moisture....................................... 62 Relationship of small mammal populations to vegetative characteristics and soil moisture... 64 Predator Scent Stations............................. 64 DISCUSSION...’OOOOOOOOOIOOOOO0...... .......... 0.0.0.0.... 69 Vegetative Sampling.. .......... . ...... . ............. 69 Plant species richness and diversity........... 69 V Page Patterns in the structure and composition of vegetative species among age classes...... 70 Small Mammal Populations............................ 74 Small mammal populations on agricultural fields.................................... 74 Patterns of small mammal populations and species distribution among age classes of CRP fields................................ 74 Predator Scent Stations............................. 84 RECOMMENDATIONS. .......... .. ............................ . 86 APPENDICES........ ..................................... .. 90 LIST OF REFERENCES ....................................... 113 vi LIST OF TABLES 2399 Seed mixture for Conservation Reserve Program fields in Gratiot County, Michigan, planted in 1987, 1989 and 1991 ......... . ..................... 11 Mean (SE) number of plant species on 1-, 3-, and 5-year-old Conservation Reserve Program fields in May and July, 1992, in Gratiot County, Michigan. No significant differences (Kruskal-Wallis one- way analysis-of—variance, P>0.10) were detected among age classes for either month................ 21 Mean (SE) number of plant species on 2-, 4-, and 6-year-old Conservation Reserve Program fields in May and July, 1993, in Gratiot County, Michigan. No significant differences (Kruskal-Wallis one- way analysis-of—variance, P>0.10) were detected among age classes for either month... ..... ........ 21 Mean (SE) percent absolute frequencies of plant species on 1-, 3-, and 5-year-old Conservation Reserve Program fields in May 1992 in Gratiot County, Michigan............. ............ . ...... .. 23 Mean (SE) relative (%) frequencies of plant species on 1-, 3-, and 5-year-old Conservation Reserve Program fields in May 1992 in Gratiot County, Michigan.......................................... 24 Mean (SE) percent absolute frequencies of plant species on 1-, 3-, and 5-year-old Conservation Reserve Program fields in July 1992 in Gratiot County, Michigan..... ........ ..................... 25 Mean (SE) relative (%) frequencies of plant species on 1-, 3-, and 5-year-old Conservation Reserve Program fields in July 1992 in Gratiot County, Michigan.......................................... 27 vii Table 10 11 12 13 14 15 16 17 Mean (SE) percent absolute frequencies of plant species on 2-, 4-, and 6-year-old Conservation Reserve Program fields in May 1993 in Gratiot County, Michigan.................................. 28 Mean (SE) relative (%) frequencies of plant species on 2-, 4-, and 6-year-old Conservation Reserve Program fields in May 1993 in Gratiot County, Michigan.......................................... 29 Mean (SE) percent absolute frequencies of plant species on 2-, 4-, and 6-year-old Conservation Reserve Program fields in July 1993 in Gratiot County, Michigan .......... ........ ...... .......... 31 Mean (SE) relative (%) frequencies of plant species on 2-, 4-, and 6-year-old Conservation Reserve Program fields in July 1993 in Gratiot County, Michigan.......................................... 32 Mean (SE) plant species diversities (Shannon-Weaver index) on 1-, 3-, and 5-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1992. No significant differences (Kruskal-Wallis one-way analysis-of—variance, P>0.10) were detected among age classes for either month................ 33 Mean (SE) plant species diversities (Shannon-Weaver index) on 2-, 4-, and 6-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1993.................................... 34 Mean (SE) vegetative characteristics on 1-, 3-, and 5-year-old Conservation Reserve Program fields in May 1992 in Gratiot County, Michigan........ ...... 35 Mean (SE) vegetative characteristics on 1-, 3-, and 5-year-old Conservation Reserve Program fields in July 1992 in Gratiot County, Michigan............. 37 Mean (SE) vegetative characteristics on 2-, 4-, and 6-year-old Conservation Reserve Program fields in May 1993 in Gratiot County, Michigan.............. 38 Mean (SE) vegetative characteristics on 2-, 4-, and 6-year-old Conservation Reserve Program fields in July 1993 in Gratiot County, Michigan............. 40 viii 19 20 21 22 23 24 25 26 Mean (SE) percent horizontal cover in 4 height strata on various age classes of Conservation Reserve Program fields in July 1992 and 1993 in Gratiot County, Michigan.......................... 41 Mean (SE) relative abundance of small mammals captured on 1-, 3-, and 5-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1992. No significant differences (Kruskal-Wallis one-way analysis-of-variance, P>0.10) were detected among age classes for all months.................. 52 Mean (SE) relative abundance of small mammals captured on 2-, 4-, and 6-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1993.................................... 53 Small mammal species on Conservation Reserve Program (CRP) fields in Gratiot County, Michigan, spring-summer, 1992 and 1993...................... 55 Mean (SE) number of small mammal species captured on 1-, 3-, and 5-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1992.. 56 Mean (SE) number of small mammal species captured on 2-, 4-, and 6-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1993.. 57 Mean (SE) small mammal species diversities (Shannon-Weaver index) on 1-, 3-, and 5-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1992............................ 58 Mean (SE) small mammal species diversities (Shannon-Weaver index) on 2-, 4-, and 6-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1993............................ 59 Mean (SE) percent small mammal capture rates (captures/trapnight) on 1-, 3-, and 5-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1992. No significant differences (Kruskal-Wallis one-way analysis-of-variance, P>0.10) were detected among age classes for all months............................................ 61 ix Table 27 28 29 30 31 32 Mean (SE) percent small mammal capture rates (captures/trapnight) on 2-, 4-, and 6-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1993............................ 61 Mean (SE) relative soil moisture levels (0 = dry; 10 = saturated) on 1-, 3-, and 5-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1992. No significant differences (Kruskal-Wallis one-way analysis-of—variance, P>0.10) were detected among age classes for either month........ ...... ............................... 63 Mean (SE) relative soil moisture levels (0 = dry; 10 = saturated) on 2-, 4-, and 6-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1993. No significant differences (Kruskal-Wallis one-way analysis-of-variance, P>0.10) were detected among age classes for all months............................................ 63 Mean monthly scent-station visitation rates (% operable stations visited by a species/month) for species associated with 1-, 3-, and 5-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1992. No significant differences (Kruskal-Wallis one-way analysis-of-variance, P>0.10) were observed among age classes........... 65 Mean monthly scent-station visitation rates (% operable stations visited by a species/month) for species associated with 2-, 4-, and 6-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1993. No significant differences (Kruskal-Wallis one-way analysis-of—variance, P>0.10) were observed among age classes........... 66 Mean (SE) percent species visitation (% operable stations visited by any species/month) on 1-, 3-, and 5-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1992. No significant differences (Kruskal-Wallis one-way analysis-of- variance, P>0.10) were detected among age classes.................................... ....... 68 Table 33 Page Mean (SE) percent species visitation (% operable stations visited by any species/month) on 2-, 4-, and 6-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1993. No significant differences (Kruskal-Wallis one-way analysis-of- variance, P>0.10) were observed among age classes.......... ................. ........ ........ 68 Mean monthly temperatures (C) and total precipitation (cm) for Alma, Michigan, 1992 and1993.00.00.00000000000000000000000000000000000 90 Plant species encountered on Conservation Reserve Program fields in 1992 and 1993 Gratiot countY' MiChigan’OOOOOOOOOOOOOOOOOOOCCC... ........ 92 xi Figgre LIST OF FIGURES Page Study area location, Gratiot County, Michigan.... 9 Mean factor scores of the first 3 principle components that emphasized major patterns of variation on 3 age classes of Conservation Reserve Program fields in May 1992 in Gratiot County, Michigan. Age classes are represented by numbers located at the top of each spike...... 44 Mean factor scores of the first 3 principle components that emphasized major patterns of variation on 3 age classes of Conservation Reserve Program fields in July 1992 in Gratiot County, Michigan. Age classes are represented by numbers located at the top of each spike...... 45 Mean factor scores of the first 3 principle components that emphasized major patterns of variation on 3 age classes of Conservation Reserve Program fields in May 1993 in Gratiot County, Michigan. Age classes are represented by numbers located at the top of each spike ...... 46 Mean factor scores of the first 3 principle components that emphasized major patterns of variation on 3 age classes of Conservation Reserve Program fields in July 1993 in Gratiot County, Michigan. Age classes are represented by numbers located at the top of each spike...... 47 Study site locations by township in Gratiot County, Michigan, 1992 and 1993 .................. 91 Small mammal species distribution on various age classes of Conservation Reserve Program fields in July 1992 and 1993 in Gratiot County, Michigan......................................... 95 xii Introduction Human alterations of the landscape have occurred throughout history. Since the development of tools and the discovery of fire; from the first known humans in Africa to the colonization of the Americas; people have transformed the environment to meet their needs and desires. In the United States, landscape alterations leading to the alteration of ecosystems such as grasslands have occurred since the arrival of European explorers (Stubbendieck 1987). For thousands of years prior to early settlers, many parts of the United States were dominated by perennial grasslands (Stubbendieck 1987). The North American prairie alone embraced 3.6 million square kilometers or 15% of the continent (Constable 1985) with 40% of the grasslands in the Midwest (Ryan 1986). As human populations increased in the U.S. so did extensive changes in land use practices which resulted in the loss of grasslands and wildlife populations dependent on habitat provided by these ecosystems (Harmon and Nelson 1973, USDA 1987). Altering native vegetative communities for the production of agricultural crops and other socio-economic uses have contributed to declines in wildlife populations by 2 decreasing the diversity of cover types available to wildlife (Berner 1988). Thus, the development of extensive croplands has contributed to some wildlife populations being isolated onto fragmented perennial grasslands (Higgins et a1. 1987). In addition to declining wildlife populations, changing land use practices have also been associated with increased soil erosion and poor water quality, both of which have negative impacts on the quality of wildlife habitat (Miranowski and Bender 1982). In an attempt to establish grassland cover to curtail soil erosion, improve water quality, reduce surplus agricultural commodities, and provide wildlife habitat, the federal government initiated a series of land retirement programs in the mid-1930's (Berner 1984, 1988). Under early land retirement programs, cropland was taken out of production and left idle or planted to cover crops. Although early land set-aside programs provided varying amounts and qualities of wildlife habitat (Harmon and Nelson 1973, Berner 1984, Cutler 1984, Edwards 1984, Berner 1988, Isaacs and Howell 1988), agricultural practices continued to have detrimental effects on soil erosion and wildlife habitat (Berner 1984). According to Harmon and Nelson (1973), a reform combining the best features of past annual set-aside programs with new multi-year set-aside programs that emphasized undisturbed, permanent grass cover could maximize benefits for both farmers and wildlife species. 3 The most recent set-aside program is the Conservation Reserve Program (CRP), a provision of the 1985 Food Security Act (Farm Bill). Economic incentives provided by the CRP encourages landowners to remove highly erodible and environmentally sensitive cropland from production for 10 years. Cropland enrolled in CRP can be planted to a variety of approved cover crops. Nationwide, an estimated 18 million ha will be taken out of production by 1995, and planted primarily to a grass-legume perennial cover (Burger et a1. 1990). According to Berner (1988), the CRP has potential to be the most beneficial land retirement program for wildlife to date because landowners enroll for multiple years. CRP lands may provide nesting, thermal, and escape cover for a diversity of wildlife species. Historically, studies have shown that multi-year set- aside farm programs, such as the CRP, are generally better for wildlife than annual set-aside farm programs (Harmon and Nelson 1973). Multi-year programs have improved wildlife habitat by providing minimal disturbance and promoting permanent cover for wildlife use. In addition, multi-year programs such as CRP can provide quality wildlife habitat for a diversity of wildlife species as changes occur in the structure and composition of vegetative communities during the enrollment period. Past research evaluating wildlife use of earlier land retirement programs, as well as the CRP, have focused on 4 avian species as indicators of habitat quality (Joselyn and Warnock 1964, Duebbert 1969, Berner 1984, Edwards 1984, Berner 1988, Farmer et al. 1988, Millenbah 1993). However, mammalian species also are associated with grasslands and open fields (Linduska 1950, Getz 1961a and 1961b, Batzli 1968, Hansson 1968, Batzli and Pitelka 1970, French 1971, Golley et al 1975, Birney et a1. 1976, Morris 1984) and have a diversity of roles in maintaining the stability and integrity of grassland ecosystems. For example, mammals transport minerals and their interactions with other consumers and predators, can impact ecosystem structure and function. In addition, small mammals play an important role in the abundance and distribution of vegetative species (Golley et a1. 1975). Medium sized mammal species such as raccoons (Procyon lotor), skunks (Nephetis mephitis), opossums (Didelphis virginianus) and weasels (Mustela nivalis) may also be common in grasslands since they prey on small mammals and may have a direct effect on small mammal populations (Linduska 1950). Because the presence of mammals in grassland ecosystems are important, they may be used to assess the impact of land management practices on wildlife populations and serve as indicators of habitat quality. The presence of diverse successional stages of grasslands has been shown to influence mammal abundance and diversity (Getz 1961a and 1961b, Batzli 1968, Hansson 1968, 5 Batzli and Pitelka 1970, Golley et al. 1975, Birney et al. 1976, Morris 1984). Changes in the structure and composition of vegetative species due to succession may result in a shift in the composition and density of small mammal populations. A study comparing wildlife use of Payment in Kind (PIK) fields to conventional corn fields showed that successional changes from corn fields to PIK fields, dominated by grasses and forbs, increased the diversity and density of avian and small mammal species (Castrale 1983). Castrale (1983) documented that in fields set aside for more than 1 year, vertebrate communities continued to change in response to vegetative succession. The importance of cover in habitat selection by small mammals has been well documented (Eadie 1953, Beckwith 1954, Getz 1961a and 1961b, Hansson 1968, Birney et a1 1976). The structure and composition of vegetative species have been correlated with the distribution, abundance, and diversity of small mammals in grassland ecosystems. According to Eadie (1953), Getz (1961a), and Batzli (1968), establishment of permanent vegetative cover may provide food and shelter for small mammals. For example, Eadie (1953) found increased populations of Microtus spp. associated with greater vegetative cover in a timothy hay field than earlier successional fields. Conversely, some species such as Peromyscus spp. prefer less dense vegetative cover (Batzli 1968) and are often associated with early successional 6 stages and agricultural croplands (Beckwith 1954, Castrale 1983). Thus, a multi-year retirement program such as CRP that promotes undisturbed, permanent vegetative cover can provide the necessary habitat components for mammalian species preferring non-cultivated habitats. Since the CRP is a relatively new program, there is very little information on the value of the program for mammal species. Establishment of the CRP offers a unique opportunity to document the impact of land retirement programs on mammal populations in agricultural landscapes. In addition, undisturbed CRP fields can be evaluated to determine changes in the relative abundance and composition of mammalian species in response to succession. Studying structurally diverse CRP fields, such as fields enrolled in different years, can provide information on the habitat conditions needed to sustain a diversity of mammal populations throughout an agricultural landscape. In addition, data from such research may help managers make recommendations to landowners for managing agricultural lands to maintain a diversity of species. OBJECTIVES The objectives of this study were to: 1). 2). 3). 4). 5). quantify the composition of small mammal populations on CRP and agricultural fields; determine the influence of CRP field age on the relative abundance, composition and diversity of small mammals associated with CRP fields; quantify the structure and composition of vegetative species on 6 age classes of CRP fields to determine their influence on small mammal abundance, composition and diversity in response to succession; identify medium sized mammals associated with CRP fields and; make recommendations for managing CRP fields for mammals. Study Area This study was conducted in Gratiot County, Michigan which is located in the central portion of the state's lower peninsula (Figure 1) and covers approximately 1106 kmz. Approximately 83.4% of Gratiot County is devoted to agriculture. Major crops in the county are corn, field beans, soybeans, and wheat (USDA 1975). The climate of the region is characterized by cold winters and mild summers with mean annual temperatures of -4 C in winter and 21 C in summer. Mean monthly maximum temperatures from May - August 1992, ranged from 21.3 C to 25.1 C (Table A-l). In 1993 from May - August mean monthly maximum temperatures ranged from 20-7 C to 28.5 C. Precipitation is well distributed throughout the year but peaks in the summer. The mean total annual precipitation (rainfall) is 75.4 cm, 62% of which is received from April to September. Total monthly precipitation from May - August, 1992 and 1993, are listed in Table A-1. Mean annual snowfall is 104.9 cm with an average of at least 2.54 cm of snow on the ground for 2 months (USDA 1975). Nine, 8 - 15 ha CRP fields were delineated in Gratiot County, Michigan in 1992 and 1993 (Figure A-1). Each field -_ L____ Figure 1. Study area location, Gratiot County, Michigan. 10 was in the 1st-6th growing season (age classes), and was planted to a mixture of introduced vegetative species (Table 1). CRP fields in Gratiot County are characterized primarily by alfalfa, common dandelion (Taraxicum officinale), orchard grass (Dactylis glomerata), timothy grass, golden rod (Solidago spp.), and Canadian thistle (Cirsium cervense). In 1992, 3 replicates of 3 initial age classes (1st, 3rd, and 5th growing seasons) were evaluated to determine differences in mammal abundance and composition in response to succession on CRP fields. The same fields were examined in 1993 (in the 2nd, 4th, and 6th growing seasons). In addition, small mammal populations were evaluated on 3 agricultural fields each year (soybean in 1992 and corn fields in 1993). Selection of agricultural fields was based upon landowner participation and proximity to other study sites. No manipulations (mowing or burning) occurred on CRP fields since initiation into the program. Study sites, situated in the western half of the county, consisted of a series of glacial moraines, nearly level till and sandy outwash plains and channels. Capac and Parkhill soils are common on the till plains while Plainfield, Riverdale, and Vestaburg soils compose sandy outwash plains and channels. Capac and Parkhill soils are poorly drained loams on uplands and lowlands, respectively, and are well suited for agriculture. Topography is gently 11 Table 1. Seed mixture for Conservation Reserve Program fields planted in 1987, 1989, and 1991 in Gratiot County, Michigan. Size Enrollment Field (ha) Year Seed Mixture 1A 14.3 1987 2.2 kg/ha timothy, 4.5 kg/ha orchard grass, 2.2 kg/ha sweet clover, 2.2 kg/ha alfalfa 8A 14.4 1987 Same as 1A 9A3 11. 7 1987 Same as 1A 12A 11.2 1987 3.4 kg/ha timothy, 2.2 kg/ha alsike, 2.2 kg/ha sweet clover 89A 12.2 1989 3.4 kg/ha orchard grass, 3.4 kg/ha alfalfa 898 8.6 1989 Same as 89A 89C 10.1 1989 Same as 1A 91A 15.1 1991 Same as 89A 918 10.4 1991 Same as 89A 91C 12.1 1991 Same as 89A a Sampled only in 1992 due to withdrawal from CRP and replaced by 12A in 1993 12 rolling with slopes ranging from 0 - 2%. Depending on annual precipitation received, some areas retained standing water from early spring to mid-June. Sandier soils may be moderately wet in the spring, but, tend to be excessively dry in late summer (USDA 1975). METHODS Vegetative Sampling The structure and composition of vegetative species were quantified along 6 permanent 100 m transects established on each CRP field. Data were collected at 6 sampling points per transect in May and July of 1992 and 1993. No vegetative data were collected on agricultural fields except for notation of crop height during each small mammal trapping period. Maximum height of living and standing dead vegetation were measured at each sampling point. Percent canopy cover of live and dead vegetation, grasses, forbs, and woody vegetation, and litter cover were measured at each point using a 50 x 50 cm Daubenmire frame (Daubenmire 1959). Percentage of bare ground within the frame was also recorded. Percent forb canopy cover included all species other than grasses, woody vegetation and moss. Frequency of occurrence of plant species was measured by identifying and recording all species occurring within the frame. Hiding cover for small mammals was quantified using a profile board (M’Closkey and Fieldwick 1975). Profile board sampling points were located along the same permanent transects used 13 14 to assess canopy cover. At each sampling point, the board was placed vertically in the vegetation and percent cover within each of 4 strata (0 - 30 cm, 31 - 60 cm, 61 - 100 cm, and 101 - 150 cm) covered by vegetation was recorded by an observer standing 15 m from the board in a randomly selected direction (Nudds 1977). Strata intervals used were selected to quantify hiding cover for small mammal species commonly occurring in grasslands (M'Closkey and Fieldwick 1975). Percent cover was recorded as 0, 25, 50, 75, or 100% cover (Gysel and Lyon 1980). Small Mammal Trapping Live trapping was used to evaluate the relative abundance, composition and species diversity of small mammal populations on all study sites. Small mammals were trapped on CRP fields using large (9 cm x 9 cm x 23 cm) Sherman live-traps (H.B. Sherman, Co., Tallahassee, Fla.) for 5 consecutive days each month from May - August in 1992 and 1993. Agricultural fields were also trapped to compare the composition of small mammal populations on agricultural fields to those on CRP fields. Farming practices limited trapping on agricultural fields to June through August. A 6 x 6 grid with traps spaced 25 m (Smith et al. 1975) apart was centered on each field. Two traps were placed at each station and covered with vegetation to maximize captures and minimize heat stress to animals in traps. 15 Traps were set and baited on the first day of the trapping period and remained set for 4 nights. The bait mixture consisted of whole oats, lard, and anise extract. Bait was added as needed during the trapping period. Traps were checked each morning and newly captured animals were identified and ear tagged. Ear tag number, species identification, gender, and location on grid were recorded. All individuals were released at point of capture. Severely injured or weak species were euthanized by cervical dislocation as suggested by the All University Committee on Animal Use and Care (AUCAUC)(Dr. Sally Walshaw, M.A., V.M.D., Michigan State University Laboratory Animal Resources, Pers. Commun.). No attempt was made to positively identify individuals of the genus Peromyscus, therefore, white-footed mice (Peromyscus leucopus) and deer mice (Peromyscus maniculatus) may have been included in the Peromyscus spp. category. Small mammals were also trapped on assessment lines on CRP fields to determine the distribution of small mammals as influenced by changes in the structure and composition of vegetation. Each 6 x 6 grid of 72 traps was intersected by 4 assessment lines. Each assessment line extended from the 3rd grid line in the 4 cardinal directions to a point 75 m outside the boundary of fields onto adjacent cover types. Station spacing of 25 m was used along each assessment line with each station containing 1 trap. Assessment line length 16 and number of trap stations varied on each study site depending on field size. Assessment lines were trapped for 3 consecutive nights concurrently with the grid trapping period. Traps were baited as previously mentioned and the same data were recorded for individuals captured on assessment lines. Predator scent stations Large mammals associated with CRP fields were monitored each month using scent stations as described by Linhart and Knowlton (1975). Four to 6 stations were spaced at 0.30 km intervals along the perimeter of each study site. A 10 m buffer was maintained between stations and adjacent roads (Lindzey et al. 1977). Each station consisted of a circle cleared of vegetation, 1 m in diameter, with a centrally placed cotton swab saturated with a mixture of cod liver oil and fermented egg. A footprint was made by the observer in the margin of each station. If the observer's footprint was still visible the next morning, the station was considered operable. Stations were activated in the afternoon and checked the following morning. A station was considered visited if at least 1 identifiable animal track was present in the cleared area. Species identification and station number were recorded. If stations were inoperable due to rain, the procedure was repeated until 4 operable days were achieved. Stations were smoothed each day so new 17 tracks could be distinguished. Visitation rates were expressed as the percentage of operable stations visited by a species per month. Soil Moisture A moisture meter (Forestry Suppliers, Inc., Jackson, Miss.) was used to determined relative soil moisture on CRP fields since moisture can influence the vegetative characteristics of an area (Stubbendieck 1987) and ultimately small mammal populations (Batzli and Pitelka 1970). Samples were taken between 0800 and 1100 hours at the end of each small mammal trapping period. Moisture levels were taken each month from May - August and scaled from 0 - 10 indicating very dry to saturated, respectively. Samples were not taken within 24 hours of rainfall. Statistically adequate sample sizes for soil moisture readings were determined using Freese's (1978) sample size estimator: n = (t2)(sz)/(E2), where: t = tabulated t value at the 90% probability level 2 sample variance s E = mean x allowable error A maximum of 30 samples per study site (n=9) were taken when moisture samples were extremely variable such as when fields contained low lying areas that received excess run- off from rain. These areas had a tendency to be saturated 18 when higher sections of fields remained dry to average. Data Analysis Paired t-tests (Ott 1988) were used to evaluate differences in vegetative characteristics within an age class vegetation variables between the May and July sampling periods. The Kruskal-Wallis (KW) one-way analysis-of- variance (Siegel 1956) was used to determine differences in vegetation variables among age classes within each month. The Kruskal-Wallis one-way analysis-of-variance was also used to determine differences in small mammal populations, soil moisture data and predator scent station data within an age class among months and among age classes within a month. The KW rank statistic (Miller 1980) was used to identify specific significant differences (a=0.10) in vegetative characteristics and mammal populations between specific age classes. Differences in soil moisture and mammalian populations among age classes over all months were compared using Friedman's two-way analysis-of-variance (Siegel 1956). Blocking occurred on months. Specific significant differences (a=0.10) were identified using Friedman’s rank statistic (Miller 1980). Associations between vegetative characteristics and small mammal populations on CRP fields were described using the Spearman Rank Correlation (r8, Siegel 1956). 19 Small mammal species diversity was calculated using the Shannon-Weaver diversity index (Shannon and Weaver 1949): H = -2 pi (lnpi), where P1 = proportion of species 1th in sample of N species. Principle components analysis (PCA) (Morrison 1976) was used to evaluate the relationship between field age and vegetative variables. Principle components analysis assigned factor scores to vegetation variables and combined variables which may have been related to a few independent variables. The new variables were used to identify the original variables that best described a particular age class of CRP fields. RESULTS Vegetative Responses Comparisons of vegetative composition among age classes A total of 64 plant species were encountered on 6 age classes of CRP fields in 1992 and 1993 (Table A-2). The mean number of plant species on CRP fields increased from May to July within each year, but tended to decrease as fields aged (Tables 2 and 3). No significant differences (P>0.10), however, were observed. One-year-old CRP fields tended to have more plant species in May and July, 1992 (i215.00 and i=17.00, respectively) than 3- and 5-year-old fields. Three-year-old fields had fewer plant species (i=11.33) than 5-year-old fields (Y=12.00) in May, however, the mean number of plant species was greater on 3-year-old fields (i213.00) than 5-year-old fields (£311.33) in July. The same trends observed in plant species richness in 1992 were also observed in 1993 (Table 3). The mean number of plant species tended to increase from May to July, however, decreased within months across age classes in May. Two-year-old fields (i512.33) had a greater mean number of plant species than 4- and 6-year-old fields (i=10.33 and i=9.00, respectively). In July, 2-year-old fields had 20 21 Table 2. Mean (SE) number of plant species on 1-, 3-, and 5-year-old Conservation Reserve Program fields in May and July, 1992, in Gratiot County, Michigan. No significant differences (Kruskal-Wallis one-way analysis-of—variance, P>0.10) were detected among age classes for either month. Age Classes Sampling 1 3 5 Period May 15.00 11.33 12.00 (0.58) (2.18) (1.00) July 17.00 13.00 11.33 (3.79) (1.53) (0.88) Table 3. Mean (SE) number of plant species on 2-, 4-, and 6-year-old Conservation Reserve Program fields in May and July, 1993, in Gratiot County, Michigan. No significant differences (Kruskal-Wallis one-way analysis-of-variance, P>0.10) were detected among age classes for either month. Age Classes Sampling 2 4 6 Period ' May 12.33 10.33 9.00 (1.76) (0.67) (1.73) July 13.67 14.33 14.33 (1.67) (0.33) (0.33) 22 slightly fewer plant species (i=13.67) than 4- (i=14.33) and 6-year-old (i314.33) fields. The mean number of plant species encountered in July on 4- and 6-year-old fields were the same. 0f the 10 most common plant species (absolute frequencies 2 5%) encountered in May 1992, (Tables 4 and 5) orchard grass (Dactylis glomerata) was the most frequent in each age class (1-, 3-, and 5-year-old fields). In addition, orchard grass exhibited significantly lower mean relative frequency on 1-year-old fields than 3-year-old fields (Table 5). The relative frequency of orchard grass on 1- and 3-year-old fields, however, did not differ from 5-year-old fields. 0f the 10 most common plant species, 6 were more abundant on 1-year-old fields (Table 4) than 3- and 5-year-old fields. Field pennycress (Thalaspi arvense) was more frequent on l-year-old fields than 3- and 5-year-old fields. Red clover (Trifolium pretense) was significantly more frequent on 1-year-old fields than 3-year-old fields (Table 4). Field pennycress and whitlow grass (Draba verna) were only found on 1-year-old fields while timothy grass (Phleum pretense) was only encountered on 3- and 5-year-old fields. 0f the 10 most common plant species found in May, 1992, 7 also occurred in July (Table 6). Field pennycress, moss (Bryophyta), and whitlow grass were less abundant (< 5%) while common groundsel (Senecio vulgaris), quackgrass 23 Table 4. Mean (SE) percent absolute frequencies of plant species on 1-, 3-, and 5-year-old Conservation Reserve Program fields in May 1992 in Gratiot County, Michigan. Age Classes Species 1 3 5 Alfalfa 59.26 49.09 29.61 (0.11) (0.11) (0.04) Common Dandelion 52.88 40.76 53.73 (0.06) (0.15) (0.13) Field Pennycressa 10.17A 0.00B 0.00B (0.01) (0.00) (0.00) Goldenrod 20.39 1.87 0.00 (0.10) (0.02) (0.00) Queen Ann’s Lace 14.74 8.27 15.74 (0.04) (0.07) (0.13) Moss 25.86 7.44 15.75 (0.11) (0.03) (0.10) Orchard Grass 97.21 97.24 97.21 (0.02) (0.03) (0.02) Red Clovera 25.05A 0.878 5.60AB (0.16) (0.01) (0.00) Timothy Grass 0.00 .12.03 1.87 (0.00) (0.01) (0.02) Whitlow Grass 21.31 0.00 0.00 (0.21) (0.00) (0.00) 3 Significantly different among age classes (Kruskal-Wallis one-way analysis-of-variance, P<0.10). Frequencies of plant species having different letters, within a row, are significantly different (Kruskal-Wallis Rank Statistic a=0.10, Miller 1980). 24 Table 5. Mean (SE) relative (%) frequencies of plant species on 1-, 3-, and 5-year-old Conservation Reserve Program fields in May 1992 in Gratiot County, Michigan. Age Classes Species 1 3 5 Alfalfa 15.11 20.26 11.53 (0.05) (0.05) (0.02) Common Dandelion 12.71 16.57 19.80 (0.01) (0.08) (0.04) Field Pennycressa 5.60A 0.003 0.003 (0.03) (0.00) (0.00) Goldenrod 0.00 0.69 6.92 (0.00) (0.01) (0.04) Queen Ann's Lace 3.87 3.43 4.22 (0.01) (0.03) (0.03) Moss 6.23 3.01 5.68 (0.02) (0.01) (0.03) Orchard Grassa 23.79A 39.81B 36.78AB (0.02) (0.01) (0.04) Red Clover 5.60 0.42 2.13 (0.03) (0.00) (0.00) Timothy Grass 0.00 . 4.82 0.94 (0.00) (0.05) (0.01) Whitlow Grass 5.12 0.00 0.00 (0.05) (0.00) (0.00) 3 Significantly different among age classes (Kruskal-Wallis one-way analysis-of-variance, P<0.10). Frequencies of plant species having different letters, within a row, are significantly different (Kruskal-Wallis Rank Statistic a=0.10, Miller 1980). 25 Table 6. Mean (SE) percent absolute frequencies of plant species on 1-, 3-, and 5-year-old Conservation Reserve Program fields in July 1992 in Gratiot County, Michigan. Age Classes Species 1 3 5 Alfalfa 53.71 64.80 48.17 (0.17) (0.09) (0.10) Common Dandelion 22.20 27.83 25.93 (0.00) (0.20) (0.07) Common Groundsel 6.52 4.63 20.37 (0.04) (0.03) (0.09) Goldenrod 21.30 10.21 23.10 (0.13) (0.03) (0.15) Orchard Grassa 62.78A 87.00B 74.97B (0.01) (0.07) (0.21) Quackgrass 68.54 38.90 28.72 (0.30) (0.09) (0.19) Queen Ann's Lace 22.23 9.27 19.47 (0.05) (0.02) (0.18) Red Clovera 36.18A 1.87B 23.33AB (0.18) (0.01) (0.02) Timothy Grass 0.00 , 28.73 42.62 (0.00) (0.27) (0.09) Wheat 31.47 0.00 0.00 (0.32) (0.00) (0.00) a Significantly different among age classes (Kruskal-Wallis one-way analysis-of-variance, P<0.10). Frequencies of vegetative species having different letters, within a row, are significantly different (Kruskal-Wallis Rank Statistic a=0.10, Miller 1980). 26 (Agropyron repens), and wheat were more abundant. Orchard grass was the most frequent species on 3- and 5-year-old fields while quackgrass had the greatest absolute frequency on 1-year-old fields. Red clover was significantly more frequent on 1-year-old fields than 3-year-old fields (Table 6). The mean relative frequencies of red clover were different (KW, P<0.10) among age classes, however, use of a KW rank statistic could not detect significant differences among age classes (Table 7). All plant species were found in each age class with the exception of wheat (only on 1- year-old fields) and timothy grass (only on 3- and 5-year- old fields). Forty-four plant species were encountered on 2-, 4-, and 6-year-old CRP fields in 1993. Twenty-five of these species had absolute frequencies greater than 5%. Of the 10 most common plant species occurring in May, 1993, all were present in May 1992 with the exception of sandwort (Table 8). Eight of the 10 most common plant species tended to be more common on 2-year-old fields than the other age classes (Table 8). Plant species significantly more frequent on 2-year-old fields than other age classes included alfalfa, moss and queen ann's lace. Red clover was significantly more frequent on 2-year-old fields than 6-year-old fields, and 2- and 6-year-old fields were not different from 4-year- old fields (Tables 8 and 9). Common groundsel and timothy grass were significantly more frequent on 6-year-old fields 27 Table 7. Mean (SE) relative (%) frequencies of plant species on 1-, 3-, and 5-year-old Conservation Reserve Program fields in July 1992 in Gratiot County, Michigan. Age Classes Species 1 3 5 Alfalfa 16.07 23.83 16.33 (0.05) (0.06) (0.04) Common Dandelion 6.23 8.10 8.13 (0.01) (0.05) (0.01) Common Groundsel 1.73 1.33 6.20 (0.02) (0.01) (0.02) Goldenrod 5.43 3.33 6.67 (0.03) (0.01) (0.04) Orchard Grassa 20.00A 30.33B 26.23B (0.00) (0.04) (0.09) Quackgrass 20.63 14.63 8.23 (0.09) (0.05) (0.05) Queen Ann's Lace 6.50 3.50 5.43 (0.02) (0.01) (0.05) Red Clovera 9.23A 0.60B 0.73AB (0.04) (0.00) (0.01) Timothy Grass 0.00 7.67 14.63 (0.00) '(0.07) (0.04) Wheat 7.30 0.00 0.00 (0.07) (0.00) (0.00) a Significantly different among age classes (Kruskal-Wallis one-way analysis-of-variance, P<0.10). Frequencies of plant species having different letters, within a row, are significantly different (Kruskal-Wallis Rank Statistic a=0.10, Miller 1980). 28 Table 8. Mean (SE) percent absolute frequencies of vegetative species on 2-, 4-, and 6-year-old Conservation Reserve Program fields in May 1993 in Gratiot County, Michigan. Age Classes Species 2 4 6 Alfalfaa 69.43A 44.43AB 10.208 (0.13) (0.16) (0.10) Common Dandelion 74.07 45.37 50.90 (0.06) (0.23) (0.13) Common Groundsela 1.87A 3.70AB 11.37B (0.02) (0.02) (0.03) Goldenrod 17.61 4.63 8.33 (0.15) (0.03) (0.08) Mossa 70.37A 32.41AB 15.57B (0.13) (0.08) (0.08) Orchard Grass 99.07 97.20 53.74 (0.01) (0.02) (0.29) Queen Ann's Lacea 35.17A 2.77B 1.87B (0.12) (0.03) (0.01) Red Clovera 39.82A 7.40A 4.63A (0.05) (0.04) (0.02) Sandwort 25.01 0.00 0.00 (0.21) (0.00) (0.00) Timothy Grassa 0.00A 32.40AB 60.17B (0.00) (0.26) (0.21) a Significantly different among age classes (Kruskal-Wallis one-way analysis-of-variance, P<0.10). Frequencies of plant species having different letters, within a row, are significantly different (Kruskal-Wallis Rank Statistic a=0.10, Miller 1980). 29 Table 9. Mean (SE) relative (%) frequencies of plant species on 2-, 4-, and 6-year-old Conservation Reserve Program fields in May 1993 in Gratiot County, Michigan. Age Classes Species 2 4 6 Alfalfa 14.97 15.30 3.53 (0.04) (0.06) (0.04) Common Dandelion 15.63 13.57 20.83 (0.02) (0.06) (0.03) Common Groundsela 0.37A 1.53AB 5.23B (0.00) (0.01) (0.01) Goldenrod 3.74 1.31 2.90 (0.03) (0.01) (0.03) Moss 15.12 11.03 7.44 (0.04) (0.02) (0.03) Orchard Grass 20.91 35.27 23.03 (0.01) (0.06) (0.12) Queen Ann's Lace 7.16 1.01 0.70 (0.02) (0.01) (0.00) Red Clovera 8.38A 2.50AB 1.778 (0.01) (0.01) (0.01) Sandwort 4.76 , 0.00 0.00 (0.04) (0.00) (0.00) Timothy Grassa 0.00A 9.22AB 26.778 (0.00) (0.06) (0.09) 3 Significantly different among age classes (Kruskal-Wallis one-way analysis-of—variance, P<0.10). Frequencies of plant species having different letters, within a row, are significantly different (Kruskal-Wallis Rank Statistic a=0.10, Miller 1980). 30 than 2-year-old fields (Table 9). Occurrence of these 2 species on 4-year-old fields, however, did not differ from 2- and 6-year-old fields. All 10 species occurred in each age class with the exception of sandwort (Arenaria spp.) on 2-year-old fields and timothy grass on 4- and 6-year-old fields (Table 9). Of the 10 most common plant species occurring in May, 1993, 8 were also common in July with the exception of common groundsel and sandwort (Tables 8 and 10). All 10 plant species were found in each age class. Six of the 10 plant species were most common on 2-year-old fields, though not all differences were statistically significant (Table 10). Quackgrass was significantly more frequent on 2-year-old fields than 6-year-old fields. Red clover was significantly more frequent on 2-year-old fields than 4-year-old fields. Canadian thistle (Cirsium arvense) was significantly more common on 6-year-old fields than 4-year-old fields (Table 10) and had relative frequencies that differed between 4- and 6-year-old fields (Table 11). The diversity of plant species on CRP fields did not differ (KW, P>0.10) among age classes in 1992 (Table 12). One-year-old fields, however, did exhibit slightly greater mean plant species diversity than 3- and 5-year-old fields in May and July. In addition, plant species diversities were lowest on 3-year-old fields in May and July. In 1993, significant differences (KW, P,0.10) in plant 31 Table 10. Mean (SE) percent absolute frequencies of plant species on 2-, 4-, and 6-year-old Conservation Reserve Program fields in July 1993 in Gratiot County, Michigan. Age Classes Species 2 4 6 Alfalfa 76.87 34.97 31.53 (0.11) (0.14) (0.17) Canadian Thistlea 6.50AB 3.71A 28.348 (0.04) (0.04) (0.07) Common Dandelion 30.58 35.20 75.97 (0.06) (0.23) (0.06) Goldenrod 25.06 12.99 21.31 (0.25) (0.07) (0.09) Moss 29.67 26.83 4.64 (0.27) (0.09) (0.02) Orchard Grass 60.35 94.44 65.75 (0.19) (0.06) (0.33) Quackgrassa 62.72A 26.90AB 0.938 (0.23) (0.14) (0.01) Queen Ann's Lace 33.34 14.84 2.81 (0.19) (0.12) (0.02) Red Clovera 40.74A .1.87B 9.27AB (0.21) (0.01) (0.05) Timothy Grass 1.87 32.43 72.20 (0.02) (0.30) (0.13) a Significantly different among age classes (Kruskal-Wallis one-way analysis-of-variance, P<0.10). Frequencies of plant species having different letters, within a row, are significantly different (Kruskal-Wallis Rank Statistic a=0.10, Miller 1980). 32 Table 11. Mean (SE) relative (%) frequencies of plant species on 2-, 4-, and 6-year-old Conservation Reserve Program fields in July 1993 in Gratiot County, Michigan. Age Classes Species 2 4 6 Alfalfa 18.31 11.91 8.40 (0.05) (0.06) (0.04) Canadian Thistlea 7.47AB 1.71A 14.078 (0.01) (0.01) (0.02) Common Dandeliona'b 9.92A 9.08A 19.67A (0.02) (0.05) (0.01) Goldenrod 3.97 3.92 5.37 (0.04) (0.02) (0.02) Moss 6.86 7.88 1.14 (0.07) (0.01) (0.01) Orchard Grass 13.38 31.41 17.76 (0.04) (0.05) (0.09) Quackgrassa 17.42A 9.82AB 0.27B (0.07) (0.06) (0.00) Queen Ann’s Lace 63.17 0.78 0.71 (0.03) (0.00) (0.00) Red Clovera 7.97A .0.76B 2.34AB (0.03) (0.00) (0.01) Timothy Grassa 0.32A 6.37AB 18.78B (0.00) (0.05) (0.02) 3 Significantly different among age classes (Kruskal-Wallis one-way analysis-of—variance, P<0.10). Frequencies of plant species having different letters, within a row, are significantly different (Kruskal-Wallis Rank Statistic a=0.10, Miller 1980). b Significant differences among age classes could not be detected using the Kruskal-Wallis Rank Statistic a=0.10, Miller 1980). 33 Table 12. Mean (SE) plant species diversities (Shannon- Weaver index) on 1-, 3-, and 5-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1992. No significant differences (Kruskal-Wallis, P>0.10) were detected among age classes for either month. Age Classes Months 1 3 5 May 2.23 1.68 1.84 (0.09) (0.12) (0.07) July 2.24 1.83 1.96 (0.18) (0.11) (0.15) species diversity were detected among age classes in May, but not in July (Table 13). Two-year-old fields had significantly greater mean plant species diversity than 6-year-old fields. Two- and 6-year-old fields, however, did not differ from 4-year-old fields. Although no significant differences were found in July, 2-year-old fields had slightly greater species diversity than 4- and 6-year-old fields. Comparisons of vegetative structural characteristics among age classes Several differences (KW, P<0.10) were detected among age classes for vegetative structural characteristics measured in 1992 and 1993. In May 1992, differences (KW, P<0.10) among age classes were observed for percent canopy cover of grasses, forbs, litter cover, bare ground, and litter depth (Table 14). Percent grass canopy cover was greater on 5-year-old fields than 1-year-old fields, 34 Table 13. Mean (SE) plant species diversities (Shannon- Weaver index) on 2-, 4-, and 6-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1993. Age Classes Months 2 4 6 Maya 2.10A 1.78A8 1.588 (0.15) (0.05) (0.14) July 2.14 2.03 2.12 (0.17) (0.10) (0.06) 3 Significantly different among age classes (Kruskal-Wallis one-way analysis-of-variance, P<0.10). Mean species diversity having different letters, within a row, are significantly different (Kruskal-Wallis Rank Statistic a=0.10, Miller 1980). 35 Table 14. Mean (SE) vegetative structural characteristics on 1-, 3-, and 5-year-old Conservation Reserve Program fields in May 1992 in Gratiot County, Michigan. Age Classes Characteristic 1 3 5 Live height (dm) 2.87 4.40 3.97 (0.46) (0.45) (0.39) Dead height (dm) 7.45 9.20 11.23 (0.65) (1.11) (0.90) 2 Total canopy 55.13 57.23 62.63 (4.56) (5.29) (2.80) % Live canopy 51.43 53.03 58.10 (3.28) (5.57) (4.83) % Dead canopy 3.97 4.63 5.73 (0.94) (1.12) (1.16) % Grass canopya 27.17A 45.40AB 47.008 (4.22) (7.76) (2.55) % Forb canopya 29.07A 12.438 15.80AB (0.44) (2.20) (4.41) % Woody canopy 0.00 0.03 0.09 (0.00) (0.03) (0.09) % Litter canopya 25.57A ,57.478 51.33AB (12.15) (3.82) (3.52) % Bare grounda 33.13A 2.43AB 1.988 (11.84) (1.64) (0.72) Litter depth (cm)a 2.30A 11.278 5.87AB (0.86) (3.22) (1.32) a Significantly different among age classes (Kruskal—Wallis one-way analysis-of—variance, P<0.10). Vegetative structural characteristics having different letters, within a row, are significantly different (Kruskal-Wallis Rank Statistic a=0.10, Miller 1980). 36 however, 1- and 5-year-old fields were not significantly different from 3-year-old fields. Percent forb canopy cover on 1-year-old fields was significantly greater than 3-year- old fields, however, 1- and 3-year-old fields did not differ from 5-year-old fields. Percent bare ground was significantly less on 5-year-old fields than 1-year-old fields. Percent bare ground on 1- and 5-year-old fields did not differ from 3-year-old fields. One-year-old fields had significantly less litter than 3-year-old fields, though 1- and 3-year-old fields were not different from 5-year-o1d fields. Trends observed in vegetative structural characteristics in July, 1992 were similar to those observed in May (Tables 14 and 15). Percent total canopy cover and grass canopy cover were significantly greater (KW, P<0.10) on 3-year-old fields than 1-year-old fields, but 1- and 3- year-old fields exhibited no differences from 5-year-old fields. Percent bare ground of CRP fields differed among age classes, however, statistically significant differences between age classes could not be detected with the KW rank statistic. Differences in vegetative structural characteristics (KW, P<0.10) detected among age classes in May, 1993, included percent canopy cover of forbs, live vegetation, dead vegetation, and litter depth (Table 16). Two-year-old fields had significantly greater percent canopy cover of Table 15. 37 Mean (SE) vegetative structural characteristics on 1-, 3-, and 5-year—old Conservation Reserve Program fields in July 1992 in Gratiot County, Michigan. Age Classes Characteristic 1 3 5 Live height (dm) 9.10 8.20 9.40 (1.35) (1.56) (1.44) Dead height (dm) 3.97 7.47 9.83 (1.84) (1.52) (0.30) % Total canopya 74.70A 92.778 86.37AB (6.78) (2.37) (5.82) % Live canopy 68.77 86.03 79.87 (6.59) (2.45) (6.24) % Dead canopy 5.97 6.57 6.10 (0.35) (1.94) (0.35) % Grass canopya 27.63A 62.00B 54.63AB (6.94) (6.03) (5.82) % Forb canopy 50.47 36.60 37.77 (7.55) (6.18) (1.85) % Woody canopy 0.00 0.00 0.00 (0.00) (0.00) (0.00) % Litter canopy 43.33 ,86.87 72.00 (15.45) (10.01) (13.01) % Bare ground‘i'“b 33.37A 2.47A 2.47A , (14.11) (0.99) (1.20) Litter depth (cm) 1.27 5.60 3.47 . (0.70) (2.08) (1.52) 3 Significantly different among age classes (Kruskal-Wallis one-way analysis-of-variance, P<0.10). Vegetative structural characteristics having different letters, within a row, are significantly different (Kruskal-Wallis Rank Statistic a=0.10, Miller 1980). b Significant differences among age classes could not be detected using the Kruskal—Wallis Rank Statistic a=0.10, Miller 1980). 38 Table 16. Mean (SE) vegetative structural characteristics on 2-, 4-, and 6-year-old Conservation Reserve Program fields in May 1993 in Gratiot County, Michigan. Age Classes Characteristic 2 4 6 Live height (dm) 4.30 3.87 2.60 (0.49) (0.75) (0.63) Dead height (dm) 4.60 4.00 6.27 (1.46) (1.02) (1.23) % Total canopy 90.97 96.83 96.60 (4.73) (0.99) (1.23) % Live canopya 79.70A 58.70AB 40.778 (9.20) (5.80) (7.73) % Dead canopya 6.47A 31.73AB 33.978 (1.88) (9.82) (2.49) % Grass canopy 52.80 52.20 37.37 (10.56) (6.35) (6.56) % Forb canopya 38.47A 12.87AB 9.678 (10.85) (4.65) (3.29) % Woody canopy 0.00 0.00 0.03 (0.00) (0.00) (0.03) % Litter canopy 16.00 .28.80 28.27 (7.49) (4.80) (2.87) % Bare ground 8.87 3.17 2.97 (4.67) (0.99) (1.42) Litter depth 1.67A 4.03A 6.23A (cm)a'b (0.38) (0.78) (2.56) 3 Significantly different among age classes (Kruskal-Wallis one-way analysis-of-variance, P<0.10). Vegetative structural characteristics having different letters, within a row, are significantly different (Kruskal-Wallis Rank Statistic a=0.10, Miller 1980). b Significant differences among age classes could not be detected using the Kruskal-Wallis Rank Statistic a=0.10, Miller 1980). 39 live vegetation and forbs than 6-year-old fields. Two- and 6-year-old fields, however, did not differ from 4-year-old fields. Conversely, 2-year-old fields had significantly less percent dead canopy cover than 6-year-old fields and dead canopy cover on 2- and 6-year-old fields did not differ from 4-year-old fields. Litter depth did differ (KW, P<0.10) among age classes, however, use of a KW rank statistic could not detect differences between age classes. No differences (KW, P>0.10) were observed among age classes for vegetative structural characteristics in July 1993 with the exception of litter depth (Table 17). Two- year-old fields had significantly less litter depth than 6- year-old fields, though 2- and 6-year-old fields were not different from 4-year-old fields. Many vegetative variables in July continued to show similar trends as encountered in May, 1993, though differences were not statistically significant. Percent canopy cover of forbs, and bare ground were greater on 2-year-old fields than older fields while percent dead canopy cover, percent litter cover, and litter depth were greater on older fields (4- and 6-year-old) than 2-year-old fields. Percent horizontal cover in each of 4 height strata did not differ (KW, P>0.10) among age classes on CRP fields in 1992 and 1993 (Table 18). The lowest height stratum in each age class tended to have the greatest percent cover. 40 Table 17. Mean (SE) vegetative structural characteristics on 2-, 4-, and 6-year-old Conservation Reserve Program fields in July 1993 in Gratiot County, Michigan. Age Classes Characteristic 2 4 6 Live height (dm) 10.67 7.67 12.13 (0.95) (3.47) (0.68) Dead height (dm) 3.67 5.63 4.50 (2.27) (3.91) (2.15) % Total canopy 92.40 99.97 99.63 (6.42) (0.03) (0.37) % Live canopy 71.57 80.50 63.60 (12.83) (5.14) (6.12) % Dead canopy 10.23 6.90 20.97 (3.07) (1.65) (5.22) % Grass canopy 35.90 75.17 46.17 (4.45) (9.84) (10.03) % Forb canopy 48.83 23.17 33.63 (9.27) (9.08) (5.18) % Woody canopy 0.05 0.00 0.05 (0.05) (0.00) (0.05) % Litter canopy 15.03 _ 21.53 19.07 (4.58) (4.07) (1.17) % Bare ground 7.59 0.05 0.37 (6.42) (0.05) (0.37) Litter depth (cm)a 1.38A 2.63AB 4.108 (0.71) (0.27) (0.35) a Significantly different among age classes (Kruskal-Wallis one-way analysis-of-variance, P<0.10). Vegetative structural characteristics having different letters, within a row, are significantly different (Kruskal-Wallis Rank Statistic a=0.10, Miller 1980). 41 Am.oc Ae.mav Am.omv Ae.ev A~.ec “a.~c o.a¢ a.aa 0.5m N.a m.m o.ma Eoomauaoa Am.ov 18.6av Am.mav Aa.~V Am.mav Am.oav o.ma h.~¢ b.0b o.ma o.hm o.am Eoooalam 56.03 A¢.aav Am.mc “n.4ac Ao.mav Am.ec o.ooa m.am m.ma m.¢b m.¢a m.¢w EOOmlam Ao.oc Am.ov 16.63 Ao.~v Am.oc Ao.ov o.ooa m.am o.ooa o.ma a.aa o.ooa EvomIo w v m m m a Edumuum unaamm mama 2a omamsmm mommmao mad mama ca 60amawm mommoau mam. .Esumuum £000 you mommoao mam macaw omuomumc who; Aoa.oAm .moc0aum>uuo Imamaaocm hmsumco maaamsuamxmsnxv moosmumumao usmoamasaam oz .coaanoaz .hucsoo acauwuo ca mama can mama wash sa mcamam amuaoum m>ummmm soaum>ummcoo mo mmmmmao mam msoaum> so muouum usaaon v :a um>oo awucouauos unmoumm Ammv com: .ma manna 42 Comparisons of vegetative structural characteristics between months Vegetative characteristics that significantly increased (paired t-test, P<0.10) between May and July, 1992 (percent of change between May and July is indicated in parentheses for each age class), included percent total canopy cover on 1-, 3-, and 5-year-old fields (35%, 62%, 38%), respectively; percent forb canopy cover (37%), percent live canopy cover (62%), and percent dead canopy cover (42%) on 3-year-old fields. Litter depth (69%) on 5-year-old fields decreased significantly from May to July. All other vegetative variables in each age class were not different (paired t- test, P>0.10) between months. However, all structural variables with the exception of height of dead vegetation, and percent canopy cover of dead vegetation, woody canopy (3- and 5-year-old fields), and litter canopy tended to increase from May to July. I In 1993, vegetative structural characteristics that significantly increased (paired t-test, P<0.10) between May and July included height of live vegetation (148%, 367%) on 2- and 6-year-old fields, respectively; and percent forb canopy cover (248%) on 6-year-old fields. Percent litter canopy cover (34%) and percent bare ground (624%) on 4-year- old fields decreased significantly (paired t-test, P<0.10) from May to July. All other vegetative variables were not significantly different (paired t-test, P>0.10) between months. However, of these, all with the exception of height 43 of dead vegetation, percent canopy cover of dead vegetation (4- and 6-year—old fields), percent canopy cover of grasses (2-year-old fields), percent litter canopy cover, percent bare ground, and litter depth tended to increase in each age class from May to July. Principle components analysis The relationships between field age and vegetative characteristics in May and July, 1992 and 1993, were examined using principle components analysis (Morrison 1976) (Figures 2-5). In May, 1992, 92% of the total variance was explained in the first 3 principle components. The first principle component, which explained 52% of the total variance, illustrated the relationship between percent cover of forb canopy and grass canopy (Fig. 2). Therefore, fields having greater percent cover of forb canopy would have less grass canopy and vice versa. Twenty—eight percent of the total variance is explained in the second principle component. This component illustrated the relationship between percent total canopy cover and canopy cover of live vegetation to litter depth (Fig. 2). The percent cover of live vegetation and standing dead vegetation comprised percent total canopy cover. The entire area within the Daubenmire frame was comprised of percent total canopy cover, litter cover, and bare ground. Therefore, fields having greater percent total and live canopy cover would have less percent litter cover and thus, less litter depth. \ / / o \ \ 3 . \ \ Cano / / 3 \o'\.\ x 1 ’ / .°0o.. /\:\ .40.. \ // / .‘eo.‘..E..0o°. \ i\\ O PC3 ” / 5.63%”... \\\ / 0°..‘ E NR." 00. \\ -1 ,/ / O. P\\ // 01.... .5..‘>’.\\ \\ 0 P03 / / ....‘ Eeoq.....0 4\ // ”w. E \%1‘° \\\ -1 , / ° < . \ / ‘ \\ . 3 . \.2 Utter A/ < - Depth "l’o “3 Camp)! 2 3 3 2 Total P01 368d Canopy Pcz 8”" Figure 3. Mean factor scores of the first 3 principle components that em hasized major patterns of variation on 3 age classes of onservation Reserve Program fields in July 1992, in Gratiot County, Michigan. Age classes are re resented by numbers located at the top of each spi e. 46 \ /// RN // //\\ \N\ \ ~ Bar ///////\§\\ \\ 3 Groan // / \§\ \\ \\ 2 // // : \ \\ ,/ /< /\3\ :§\ \ 1 / “24'"... E .fllk.\ \\\ PC3 2/ flfik,‘ \\ 0 , / ‘ . Q \\ // o >\> \\ .2 2.32:1in ’ < ’4» - O \ "3 2 -1 Dead .2 '1 v: 0 Canopy 3 3 Live PC1 Figure 4. Mean factor scores of the first 3 principle components that emphasized major patterns of variation on 3 age classes of Conservation Reserve Program fields in May 1993, in Gratiot County, Michigan. Age classes are re resented by numbers located at the top of each spi e. / \ Dead 3 Canopy 2 A \\\\K\\\ /m//// \\\\ \\\\j\\\\ \\\\\ ////7 O O C j / .1 I o .A ./..../..../..../... / / 7 / /:°7./ e V v A O l /.:"/ 0 P03 002 ] t I I .1 \. \ \ ....,/ //////// ///[//// 7 lb Grass _3 Canopy //f///// / \\\\\\\\ ‘3 i: i it .3. - “V -- - ‘2 Bare Forb 2 -1 0 $1 0 1 Ground Utter Total PC1 Figure 5. Mean factor scores of the first 3 principle components that em hasized major patterns of variation on 3 age classes of onservation Reserve Program fields in July 1993, in Gratiot County, Michigan. Age classes are re resented by numbers located at the top of each spi e. 48 The third principle component, which explained 12% of the total variance, described the relationship between percent dead canopy cover to percent litter canopy cover (Fig. 2). In May, 1-year-old fields were characterized by greater percent canopy cover of forbs than grasses (PC1), greater percent total cover and live vegetation cover than litter depth (PC2), and slightly greater percent canopy cover of dead vegetation than litter cover (PC3) (Fig. 2). Three- year-old fields were characterized by greater percent canopy cover of grasses than forbs (PC1), greater litter depth than percent total canopy cover and live vegetation (PC2), and moderate weighing toward percent cover of litter canopy to dead canopy (PC3)(Fig. 2). Five-year-old fields were characterized by greater percent canopy cover of grasses to forbs (PC1), moderate weighing toward percent total canopy cover and live vegetation to litter depth (PC2), and moderate weighing toward percent litter canopy cover to dead vegetation (PC3)(Fig. 2). In July, 1992, 79% of the total variance was explained in the first 3 principle components. The first principle component, explaining 41% of the total variance, illustrated the relationship between percent cover of bare ground and total canopy cover (Fig. 3). Twenty-three percent of the total variance is explained in the second principle component. This component illustrated the relationship between percent canopy cover of forbs and dead vegetation 49 (Fig. 3). The third principle component, explaining 15% of the total variance, illustrated the relationship between percent forb canopy cover and litter depth (Fig. 3). One-year-old fields were characterized by greater percent cover of bare ground to total canopy cover (PC1), no weight toward either gradient on PC2, and moderate weighing toward forb canopy cover to litter depth on PC3 (Fig. 3). Three-year-old fields were characterized by greater total canopy cover (PC1), forb cover (PC2), and no weight toward either gradient on PC3 (Fig. 3). Five—year-old fields were characterized by moderate weighing toward total canopy cover to bare ground (PC1), no weighing toward either gradient on PC2 and moderate weighing toward litter depth than forb cover on PC3 (Fig. 3). In May, 1993, 87% of the total variance was explained by the first 3 principle components.. The first principle component, explaining 50% of the total variance, illustrated the relationship between percent canopy cover of dead vegetation and live vegetation (Fig. 4). The second and third principle components, explaining 26% and 11% of the total variance, respectively, illustrated the relationship between percent cover of bare ground and total canopy cover (Fig. 4). Two-yeareold fields were characterized by greater percent canopy cover of live vegetation to dead vegetation (PC1), and percent cover of bare ground to total canopy 50 cover (PC2), and moderate weighing toward bare ground than total canopy cover on PC3 (Fig. 4). Four-year-old fields were characterized by moderate weighing toward percent canopy cover of dead vegetation to live vegetative canopy cover (PC1) and total canopy cover to bare ground on PC2 and PC3 (Fig. 4). Six-year-old fields were characterized by greater percent canopy cover of dead vegetation to live vegetative canopy cover (PC1), and moderate weighing toward total canopy cover to bare ground on PC2 and PC3 (Fig. 4). In July, 1993, 79% of the total variance was explained in the first 3 principle components. The first principle component explained 34% of the total variance and illustrated the relationship between percent bare ground to total canopy cover (Fig. 5). The second principle component explained 25% of the total variance and illustrated the relationship between percent cover of litter canopy and depth to forb canopy cover (Fig. 5). The third principle component, explaining 20% of the total variance, illustrated the relationship between percent canopy cover of dead vegetation and grass canopy cover (Fig. 5). Two-year-old fields were characterized by moderate weighing toward percent bare ground to total canopy cover (PC1) and dead vegetative canopy cover to grass canopy cover (PC3), and greater percent forb canopy cover to litter canopy and depth (PC2)(Fig. 5). Four-year-old fields were characterized by no weighing toward either variable on PC1, 51 greater percent litter canopy cover and depth to forb canopy cover (PC2), and greater percent canopy cover of grasses to dead vegetation (PC3)(Fig. 5). Six-year-old fields were characterized by moderate weighing toward total canopy cover to bare ground (PC1), greater litter canopy cover and depth to forb canopy cover (PC2), and greater percent dead canopy cover to grass canopy cover (PC3)(Fig. 5). Small Mammal Responses Small mammal relative abundance In 1992, 461 small mammals were captured on 1-, 3-, and 5-year-old CRP fields. Mean small mammal relative abundance did not differ within a month among age classes or within an age class among months (May, June, July, and August) (Table 19). Mean small mammal relative abundance tended to be greater on l-year-old fields than other age classes in all months with an exception of August. A total of 175 small mammals were captured on 1-year-old fields. Of these, 169 (96%) were Peromyscus spp. On older (3- and 5-year-old) fields, meadow voles (Microtus pennsylvanicus) comprised the majority of mammal species captured. Of the 156 small mammals captured on 3-year-old fields, 120 (77%) were meadow voles. On S-year-old fields, 103 (76%) of 136 total captures were meadow voles. Mean small mammal relative abundance tended to increase from May to August on older (3- and 5-year-old) fields, but decreased on 1-year-old fields 52 Table 19. Mean (SE) relative abundance of small mammals captured on 1-, 3-, and 5-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1992. No significant differences (Kruskal-Wallis one-way analysis-of- variance, P>0.10) were detected among age classes for all months. Age Classes Trapping Period 1 3 5 May 17.33 3.33 5.00 (8.37) (0.89) (1.00) June 15.67 15.00 11.00 (7.23) (7.23) (9.02) July 12.67 12.00 11.00 (4.63) (8.19) (5.30) August 12.67 21.67 18.33 (3.71) (12.25) (8.84) (Table 19). A total of 958 small mammals were trapped on 2-, 4-, and 6-year-old CRP fields in 1993, an increase of 48% from 1992. Mean small mammal relative abundance in 1993 was greater than 1992 with an exception on 2-year-old fields in May and July (Table 20). A total of 206 small mammals were captured on 2-year-old fields in 1993. Three-hundred-forty small mammals were captured on 4-year-old fields. Of these, 292 (86%) were meadow voles. 0n 6-year-old fields, 412 small mammals were captured. Meadow voles (376 captures), again, comprised the majority (91%) of total captures. There were no significant differences (KW, P>0.10) in small mammal relative abundance within months among age 53 Table 20. Mean (SE) relative abundance of small mammals captured on 2-, 4-, and 6-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1993. Age Classes Trapping Period 2 4a 6a May 8.67 3.67A 7.00A (0.88) (2.73) (2.52) June 28.00 22.00AB 44.00AB (15.14) (10.26) (5.57) July 9.00 41.33B 39.33AB (4.16) (4.06) (12.72) August 23.00 46.33B 47.00B (5.57) (10.11) (8.00) 5 Significantly different within age classes among months (Kruskal-Wallis one-way analysis-of-variance, P<0.10). Relative abundance values having different letters, within a column, are significantly different (Kruskal-Wallis Rank Statistic a=0.10, Miller 1980). classes in 1993, however, significant differences were detected within age classes among months (Table 20). Four- year-old fields had a greater abundance of small mammals in July and August than in May. Mean relative abundance of mammals on 6-year-old fields was significantly greater in August than in May. Mean small mammal relative abundance tended to increase from May to August with an exception on 2-year-old fields. Mean small mammal relative abundance tended to be lower on agricultural fields than CRP fields in 1992 and 1993 (Tables 19 and 20). In August, 1992, mean small mammal relative abundance on soybean fields was 9.67 captures. In 54 1993, mean small mammal relative abundance on corn fields were 5.67 (June), 1.67 (July), and 1.33 (August). Species richness Throughout the study, 9 small mammal species were trapped on CRP fields (Table 21) while only Peromyscus spp. were captured on agricultural fields. All 9 species were captured on CRP fields in 1992. Thirteen-lined ground squirrels, opossums, and house mice were trapped on 1-year-old fields only in 1992 (Table 21). On 3- and 5-year-old CRP fields, however, Peromyscus spp., least weasels, masked shrews, meadow jumping mice, meadow voles, and short-tailed shrews were trapped. In 1992, the mean number of small mammal species trapped differed significantly (KW, P<0.10) among age classes in August (Table 22). The difference between specific age classes, however, could not be detected with the use of a Kruskal- Wallis rank statistic. The lowest mean number of species trapped in 1992 occurred on 1-year-old fields in May and the greatest number on 5-year-old fields in August (Table 22). Only Peromyscus spp. were trapped on 1-year-old fields in May. The mean number of small mammal species captured for the entire period differed significantly (Friedman, P<0.10) among age classes. However, differences between age classes could not be detected. Of the 9 species trapped in 1992, 7 were also captured in 1993 (Table 21). Species not trapped in 1993 included 55 Table 21. Small mammal species on Conservation Reserve Program (CRP) fields in Gratiot County, Michigan, spring- summer, 1992 and 1993. Common name Scientific name House Mouse 1‘ Mus musculus Least Weasel 2'3'4'5'6 Mustela nivalis Masked Shrew 3““5'6 Sorex cinereus Meadow Jumping Mouse 25L4'535 Zapus hudsonius Meadow Vole 1'2'3'4'5'6 Microtus pennsylvanicus Peromyscus spp. 1'2'3'4'5'6 Peromyscus spp. Opossum 1 Didelphis virginiana Short-tailed Shrew 2'3'4'5'6 Blarina brevicauda Thirteen-lined Ground 1JL4 Spermophilus tridemlineatus Squirrel ‘ Numbers following common names of small mammal species represent age classes of CRP fields where animals were captured. 56 Table 22. Mean (SE) number of small mammal species captured on 1-, 3-, and 5-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1992. Age Classes Trapping Period 1 3 5 May 1.00 1.33 1.67 (0.00) (0.33) (0.33) June 1.67 3.00 1.67 (0.33) (0.58) (0.33) July 1.33 2.33 1.67 (0.33) (0.88) (0.33) Augusta 1.67A 3.00A 3.67A (0.33) (0.58) (0.33) ‘ Significantly different among age classes within months (Kruskal-Wallis one-way analysis-of-variance, P<0.10). Significant differences among age classes could not be detected using the Kruskal-Wallis Rank Statistic a=0.10, Miller 1980). opossums and house mice. The mean number of small mammal species captured on 2-year-old fields were significantly greater than 4-year-old fields, however, the mean number of species on 2- and 4-year-old fields did not differ from 6 year-old fields (Table 23). The mean number of small mammal species trapped increased from May to August in each age class except on 2-year-old fields. The mean number of species on 2-year-old fields decreased from June to July, but increased from July to August (Table 23). Small mammal diversity One-year-old CRP fields had lower mean species diversities of small mammals than 3- and 5-year-old fields 57 Table 23. Mean (SE) number of small mammal species captured on 2-, 4-, and 6-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1993. Age Classes Trapping Period 2 4 6 Maya 2.00A 0.678 2.67AB (0.58) (0.33) (0.33) June 3.00 2.00 2.33 (0.00) (0.00) (0.33) July 1.67 2.33 3.33 (0.33) (0.88) (0.33) August 3.33 3.33 3.67 (0.67) (0.88) (0.33) 3 Significantly different among age classes (Kruskal-Wallis, P<0.10). Number of species having the same letter, within a row, are not significantly different (Kruskal-Wallis Rank Statistic, a=0.10, Miller, 1980). each month in 1992, though not all differences were significant (Table 24). In June, 3-year-old fields had significantly greater mean small mammal diversity than 1-year-old fields, however, 1- and 3-year-old fields did not differ from 5-year-old fields. Mean diversities were different among age classes in August, however, differences between age classes could not be identified. Five-year-old fields in August had the greatest overall mean small mammal diversity in 1992. In 1993, 2-year-o1d fields had a greater diversity of small mammal species than 4- and 6-year-old fields each month, though not all differences were significant 58 Table 24. Mean (SE) small mammal species diversities (Shannon-Weaver index) on 1-, 3-, and 5-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1992. Age Classes Trapping Period 1 3 5 May 0.00 0.23 0.42 (0.00) (0.23) (0.21) Junea 0.13A 0.288 0.26AB (0.13) (0.10) (0.20) July 0.06 0.53 0.39 (0.06) (0.32) (0.20) Augusta"b 0.24A 0.82A 0.98A (0.17) (0.35) (0.28) ‘ Significantly different among age classes within months (Kruskal-Wallis one-way analysis-of-variance, P<0.10). Means having different letters, within a row, are significantly different (Kruskal-Wallis Rank Statistic a=0.10, Miller 1980). b Significant differences among age classes could not be detected using the Kruskal-Wallis Rank Statistic (a=0.10, Miller 1980). (Table 25). Two-year-old fields in May had a greater (KW, P<0.10) diversity of species than 4-year-old fields, however, 2- and 4-year-old fields were not different from 6-year-old fields. Differences in mean species diversity among age classes in June could not be detected with the use of a KW rank statistic. Mean small mammal species diversity in July was significantly greater on 2-year-old fields than 6-year-old fields, however, 2- and 6-year-old fields were not different from 4-year-old fields (Table 25). 59 Table 25. Mean (SE) small mammal species diversities (Shannon-Weaver index) on 2-, 4-, and 6-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1993. Age Classes Trapping Period 2 4 6 Maya 0.7lA 0.008 0.42AB (0.13) (0.00) (0.26) Junea'b 0.71A 0.28A 0.27A (0.13) (0.10) (0.06) Julya 0.90A 0.37AB 0.148 (0.10) (0.21) (0.09) August 0.89 0.69 0.41 (0.19) (0.26) (0.23) a Means having different letters, within a row, are significantly different (Kruskal-Wallis Rank Statistic a=0.10, Miller 1980). b Significant differences among age classes could not be detected using the Kruskal-Wallis Rank Statistic a=0.10, Miller 1980). Two-year-old fields in July had the greatest overall mean small mammal diversity for 1993. Small mammals captured on assessment lines increased species diversity on all age classes of CRP fields, however, the number of captures were relatively low on these lines. All small mammal species trapped on assessment lines were also caught on grids with the exception of house mice which were only caught on assessment lines (Table 21). Traps in vegetation types adjacent to CRP fields generally had lower capture success than those on CRP 60 fields. The 3 main adjacent vegetation types were agricultural fields, woodlots, and other CRP/open grassland fields. Peromyscus spp. were the only species trapped in all 3 adjacent cover types while meadow voles, short-tailed shrews, least weasels, and masked shrews were trapped predominately on adjacent CRP/open fields. A woodland jumping mouse (Napaeozapus insignis) was captured in a woodlot. Capture rates Mean percent capture rates (captures/trapnight) did not differ (KW, P>0.10) among age classes in 1992 (Table 26). Capture rates in all months on 1-year-old fields were greater than or equal to those observed on 3- and 5-year-old fields with an exception in August. In 1993, no differences (KW, P>0,10) were detected in capture rates among age classes (KW, P>0.10). However, differences (KW, P<0.10), in capture rates were detected within 4- and 6-year-old fields among months (Table 27). Mean percent capture rates increased from May to August on 4- and 6-year-old fields, though not all differences were significant. Percent capture rates were significantly greater in August than in May on 4- and 6-year-old fields (Table 27). On 2-year-old fields, mean percent capture rates varied from month to month with the lowest capture rate in May and the greatest in June (Table 27). Mean percent capture rates 61 Table 26. Mean (SE) percent small mammal capture rates (captures/trapnight) on 1-, 3-, and 5-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1992. No significant differences (Kruskal-Wallis one-way analysis- of-variance, P>0.10) were detected among age classes for all months. Age Classes Trapping Period 1 3 5 May 4.00 0.67 1.33 (0.02) (0.00) (0.00) June 3.33 3.33 2.33 (0.02) (0.01) (0.20) July 3.00 2.67 2.67 (0.01) (0.02) (0.01) August 2.67 5.00 4.00 (0.01) (0.03) (0.02) Table 27. Mean (SE) percent small mammal capture rates (captures/trapnight) on 2-, 4-, and 6-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1993. Age Classes Trapping Period 2 4a 6a May 1.33 0.67A 1.33A (0.01) (0.01) (0.01) June 6.00 5.00AB 9.67AB (0.03) (0.02) (0.01) July 2.33 8.67AB 8.67AB (0.01) (0.01) (0.03) August 5.00 10.008 10.67B (0.01) (0.02) (0.02) 3 Significantly different within age classes among months (Kruskal-Wallis one-way analysis-of—variance, P<0.10). Capture rates having different letters, within a column, are significantly different (Kruskal-Wallis Rank Statistic a=0.10, Miller 1980). 62 for 1993 were 39% and 50% lower on 2-year-old fields than 4- and 6-year-old fields, respectively. The greatest capture rate observed on all age classes was on 6-year-old fields in August. Mean percent small mammal capture rates on agricultural fields in 1992 and 1993 were similar to capture rates observed in May on CRP fields in 1992 and 1993. Mean percent small mammal capture rates on soybean fields in August 1992 (5.00 captures/trapnight) were greater than capture rates observed on corn fields in 1993. In 1993, mean percent small mammal capture rates decreased from June to August (3.00% - June, 1.00% - July, and 1.00% — August). Soil Moisture No differences (KW, P>0.10) in mean relative soil moisture were detected within months among age classes or within an age class between months in 1992 and among months in 1993 (Tables 28 and 29). Although no differences were found, apparent trends that relate relative soil moisture to field age were observed. In 1992 and 1993, soil moisture of fields tended to increase throughout the summer and from younger to older fields (Tables 28 and 29). The lowest moisture level was observed on 2-year-old fields in May while the greatest was recorded on 6-year-old fields in August. 63 Table 28. Mean (SE) relative soil moisture levels (0 = dry; 10 = saturated) on 1-, 3-and S-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1992. No significant differences (Kruskal-Wallis one-way analysis-of- variance, P>0.10) were detected among age classes for either month. Age Classes Sampling Period 1 3 5 July 1.13 2.32 3.31 (0.50) (0.78) (0.96) August 2.45 2.43 3.77 (0.63) (1.04) (0.57) Table 29. Mean (SE) relative soil moisture levels (0 = 10 = saturated) on 2-, 4-, and 6-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1993. No significant differences (Kruskal-Wallis one-way analysis- of-variance, P>0.10) were detected among age classes for all months. Age Classes Sampling Period 2 4 6 May 1.83 3.85 5.10 (1.11) (1.55) (0.70) June 3.10 4.51 5.24 (1.93) (1.34) (0.87) July 3.40 4.31 4.80 (1.61) (1.64) (0.73) August 3.38 4.41 6.28 (1.78) (0.94) (0.43) 64 Relationship of small mammal populations to vegetative characteristics and soil moisture Several relationships between small mammal populations and vegetative characteristics and small mammal populations and soil moisture were observed in 1993 (Spearman rank correlation). Percent small mammal capture rates in July and August were positively correlated with percent total canopy cover in July (r8 = 0.68, P = 0.00 and r8 = 0.61, P < 0.01, respectively). In addition, percent small mammal capture rates in August were highly associated with August soil moisture levels (r8 = 0.76 P < 0.00). Predator Scent Stations Five medium sized mammal species were associated with CRP fields in 1992 and 6 in 1993 (Tables 30 and 31). Raccoons were the most frequent visitors recorded on all age classes of CRP fields in 1992 except on 3-year-old fields in August (Table 30). All 5 medium sized mammal species recorded in 1992 were present in 1993 with the addition of the opossum, on 2-year-o1d fields in June, 1993 (Table 31). Raccoons were again the most frequent species recorded on all age classes of CRP fields throughout the summer except 4-year-old fields in June. Skunks were also recorded each month with the exception of 6-year-old fields in August. No differences (KW, P>0.10) in percent species visitation rates (% operable stations visited by a species/month) were observed within months among age classes 65 Table 30. Mean monthly scent-station visitation rates (% operable stations visited by a species/month) for species associated with 1-, 3-, and 5-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1992. No significant differences (Kruskal-Wallis, P>0.10) were observed among age classes. Age Classes Species Monitoring Period 1 3 5 Raccoon May 12.57 2.38 5.29 (Procyon lotor) June 20.00 14.29 8.33 July 16.19 9.72 9.72 August 12.04 0.00 6.67 Skunk May 6.31 0.00 3.33 (Mephitis) June 6.78 0.00 0.00 July 7.14 0.00 0.00 August 9.26 2.56 4.44 Woodchuck May 0.00 0.00 0.00 (Marmota monax) June 0.00 0.00 0.00 July 0.00 0.00 0.00 August 0.00 0.00 2.23 Cat May 0.00 0.00 1.96 (Fells catus) June 2.23 13.89 0.00 July 0.00 8.33 3.33 August 0.00 6.67 0.00 Dog May 0.00 0.00 0.00 (Canis familiaris) June 0.00 2.78 0.00 July 0.00 4.17 0.00 August 0.00 0.00 2.23 66 Table 31. Mean monthly scent-station visitation rates (% operable stations visited by a species/month) for species associated with 2-, 4-, and 6-year old Conservation Reserve Program fields in Gratiot County, Michigan, 1993. No significant differences (Kruskal-Wallis one-way analysis-of- variance, P>0.10) were observed among age classes. Age Classes Species Monitoring Period 2 4 6 Raccoon May 11.11 8.11 3.13 (Procyon lotor) June 21.05 0.00 13.89 July 20.83 7.41 7.69 August 5.41 6.25 5.26 Skunk May 2.22 5.41 6.25 (Mephitis) June 7.89 4.55 8.33 July 12.50 7.41 3.85 August 5.41 6.25 0.00 Woodchuck May 4.44 0.00 6.25 (Marmota monax) June 5.26 2.27 0.00 July 8.33 0.00 3.85 August 2.70 0.00 0.00 Cat May 4.44 0.00 3.13 (Fells catus) June 0.00 4.55 0.00 July 12.50 0.00 0.00 August 2.70 0.00 0.00 Dog May 0.00 0.00 0.00 (canis familiaris) June 0.00 2.27 0.00 July 0.00 0.00 0.00 August 0.00 0.00 0.00 Opossum May 0.00 0.00 0.00 (Didelphis June 2.63 0.00 0.00 virginiana) July 0.00 0.00 0.00 August 0.00 0.00 0.00 67 or within age classes among months (Tables 32 and 33). However, a trend of decreasing mean mammal visitations were observed from younger to older fields in 1992 and 1993. 68 Table 32. Mean (SE) percent species visitation (% operable stations visited by any species/month) on 1-, 3-, and 5-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1992. No significant differences (Kruskal-Wallis one-way analysis-of-variance, P>0.10) were observed among age classes. Age Classes Monitoring Period 1 3 5 May 32.67 4.67 10.67 (0.13) (0.05) (0.06) June 34.33 33.00 8.33 (0.17) (0.17) (0.08) July 21.00 18.33 11.67 (0.12) (0.04) (0.07) August 25.00 9.33 13.33 (0.14) (0.06) (0.13) Table 33. Mean (SE) percent species visitation (% operable stations visited by any species/month) on 2-, 4-, and 6-year-old Conservation Reserve Program fields in Gratiot County, Michigan, 1993. No significant differences (Kruskal-Wallis one-way analysis-of-variance, P>0.10) were observed among age classes. Age Classes Monitoring Period 2 4 6 May 22.67 19.33 19.67 (0.11) (0.03) (0.12) June 39.33 13.67 20.67 (0.08) (0.07) (0.11) July 28.00 16.00 10.67 (0.20) (0.11) (0.07) August 15.67 13.33 2.67 (0.16) (0.09) (0.03) DISCUSSION Vegetative Sampling Plant species richness and diversity Ecosystems are not static by nature. Plant communities interact with many factors over time to produce sequences of successional change (Haufler 1990). Factors that may influence later successional stages on CRP fields include climate, soil types, and seeding mixtures. In the case of manipulated grasslands, succession induces change in plant community structure and composition, subsequently influencing wildlife populations. In this study, younger (1- and 2-year-old) CRP fields tended to have a greater number and diversity of plant species than older (3-, 4-, 5-, and 6-year-old) age classes of fields (Table 12 and 13). This trend is probably not solely the result of the original seed mixtures planted since most fields were planted with the same species (Table 1). Most likely a combination of weather, soil types, natural invasion of plants, and competition among plant species influenced the vegetative composition of fields. Younger fields, especially recently enrolled CRP fields, may provide optimal conditions for a diversity of vegetative 69 70 species to become established. Establishment of non-planted vegetative species on CRP fields may occur in several ways. For example, germination of wind-disseminated seeds are facilitated by the presence of large areas of bare soil during the first year of CRP plantings. Since younger fields were recently disturbed leaving an open canopy, a diversity of species from adjacent vegetation types could easily colonize CRP fields. Also, dormant seeds of species that could not germinate due to prior cultivation practices could now become established (e.g. field pennycress, bouncing bet). In addition, short- lived species that do not compete successfully with other plants will most likely be numerous for several years before denser types of vegetation (e.g. grasses) become dominant (Beckwith 1954). Patterns in the structure and composition of vegetative species among age classes Most vegetative structural characteristics measured within each age class increased from May to July in 1992 and 1993, primarily due to the growth of vegetation throughout the spring and summer. In 1992 and 1993, younger (1- and 2- year-old) fields were characterized by greater percent canopy cover of forbs, live and dead vegetation, total canopy cover, and percent bare ground (only 2-year-old fields)(Fig. 2-5) than older age classes. Greater percent forb canopy cover on 1- and 2-year-old fields may be attributed to the initial seed mixtures and natural 71 invasions of other plant species which was facilitated by the presence of bare ground (Tables 14-17). Bare ground predominately occurred on 1- and 2-year-old fields as a result of recent enrollment into CRP and disturbance from the initial planting. The presence of greater dead canopy cover on younger (1- and 2-year-old) fields than older fields in May 1992, and July 1993, can be explained by the variables on each principle component. In May 1992, the 2 variables most influential in describing 1-, 3-, and 5-year-old fields included percent canopy cover of dead vegetation and litter canopy cover (PC3)(Fig. 2). On this gradient, 3- and 5- year-old fields had greater litter canopy cover while 1- year-old fields had the greatest amount of dead canopy cover. Of these 2 variables, it was expected that 1-year- old fields would have greater dead canopy cover than litter canopy cover since there had not been substantial time for accumulation of litter cover on younger fields. In July 1993, 2 of the most influential variables describing 2-, 4-, and 6-year-old fields on PC3 (Fig. 5) were dead canopy cover and grass canopy cover. Of these 2 variables, dead canopy cover best described 2-year-old fields since grasses had not yet become established. One and 2-year-old fields had greater percent live canopy cover than litter depth (Fig. 3) and dead canopy cover (Fig. 4), respectively, in July which was the peak of 72 the growing season. On these fields, greater percent canopy cover of live vegetation was comprised mostly of forbs (Tables 15 and 17). The middle aged fields each year (3- and 4-year-old fields) were best described by greater percent canopy cover of grasses, total and live vegetative canopy, and litter depth and less percent forb canopy cover and bare ground (Figs. 2-5). The frequencies of occurrence of many forbs and legumes were less on 3- to 6-year-old fields than younger (1- and 2-year-old) fields (Tables 4 and 5, and 8 and 9). Alfalfa, for example, declined with field age. The primary factor for the decline in alfalfa is most likely a result of subsequent replacement by invading grasses and weeds. The presence of grasses on older fields may be a result of different factors such as annual and biennial plants having short life cycles and being unable to compete for resources with increasing perennial species (Beckwith 1954). Basu et al. (1978) also found a decline (14%) in alfalfa over a 3 year period in a mixed hay field which was replaced by perennial grasses (timothy). In this study, forb canopy cover decreased each year as grasses became more established, thus, dead canopy cover on fields 2 3 years of age was comprised mostly of grasses. A prominent litter depth on middle aged (3- and 4-year-old) fields indicated that dead plant material can be found in the vegetative canopy. As plant species became 73 established with age, percent bare ground decreased due to increased litter cover accumulated from dead and dying vegetation. This is particularly critical since litter cover and depth peaked on 3- and 4-year-old CRP fields and, thus, reduced the amount of area available for forb species to germinate and develop. Furthermore, many forb species may be unable to succeed themselves due to inhibitory effects of decaying roots upon growth and survival (Keever 1950). Older (5- and 6-year-old) CRP fields were best described by greater percent canopy cover of grasses (5-year-old fields), dead vegetation, litter cover and depth, and moderate total canopy cover (to bare ground) than younger fields (Figs. 2-5). It should be noted that litter depth was greater on 3-year-old fields than 5-year-old fields in 1992 (Table 14). A decline in litter depth from 3- to 5-year-old fields may be attributed to the compaction and decay of a heavy litter layer over time. According to Rice and Parenti (1978) and Knapp and Seastedt (1986), the accumulation of litter cover may decrease plant productivity in undisturbed grasslands. Several factors may cause such a decline. Peet et al. (1975) suggested that increased dead vegetation and litter cover may shorten the growing season of plants by keeping the ground cooler for an extended amount of time, therefore, inhibiting germination. Furthermore, they mentioned that litter may contain toxins 74 that inhibit the growth of vegetation. Small Mammal Populations Small mammal populations on agricultural fields Soybean (1992) and corn (1993) fields had the lowest diversity of small mammal species (only Peromyscus spp.) of all types and age classes of fields trapped. These results are similar to others who have trapped small mammals in row crops. Castrale (1983, 1985) documented that agroecosytems, such as row-crop fields, offer minimal wildlife habitat for small mammal species. Often, deer mice (Peromyscus maniculatus) are the only residents of cultivated fields (Linduska 1942). On corn and soybean fields in southeastern Indiana, Castrale (1985) found small mammal populations of moderate density but low species diversity (H' < 0.50), with deer mice comprising 73% of the total captures. Results from this study found small mammal populations of low relative abundance and diversity (H’ = 0.00) with only Peromyscus spp. captured on corn and soybean fields. In addition, Castrale (1985) had higher capture rates, especially deer mice, on conventionally tilled soybean fields than conventionally tilled corn fields with a negative relationship between capture rates and percent ground cover. Patterns of small mammal populations and species distribution among age classes of CRP fields Small mammal abundance, richness, and species diversity 75 were greater across all age classes of CRP fields than agricultural fields. The distribution of small mammal species can be influenced by diet, suitable vegetation cover, and the presence or absence of other small mammal species (Moreth and Schramm 1972, Hallett et al. 1983, Moro 1991). Although no correlations were observed between small mammal abundance and diversity and changes in vegetation variables, a pattern of small mammal distributions was evident on CRP fields. Results from this study suggest that vegetative characteristics associated with CRP fields may be critical factors in determining the presence of various small mammal species in this agricultural landscape. Mean small mammal diversities on 1-year-old fields were very low (H’ = 0.00 - 0.24). The low species diversity observed on 1-year-old fields may be attributed to minimal grass cover and the greater presence of bare ground. Although vegetation cover was available, the cover provided was relatively sparse. Therefore, many small mammal species, especially diurnal species, may not have occupied these fields. It is not surprising that the most common species on 1-year-old fields, Peromyscus spp., was mostly a nocturnal species (LoBue and Darnell 1959). The presence and high relative abundance of Peromyscus spp. on younger fields, especially l-year-old fields, is consistent with the habitat preference documented in the literature for the species, in particular deer mice 76 (P. maniculatus)(8eckwith 1954, Batzli 1968, Castrale 1983, Hallett et al. 1983). Variables that seem to affect Peromyscus spp. abundance are relatively little ground cover, the presence of seeds, and fields supporting annual or biennial types of vegetation (Beckwith 1954, Castrale 1983). Additionally, Hallett et al. (1983) encountered Peromyscus spp. more frequently in drier areas. Such habitat conditions were present on younger (1- and 2-year- old) CRP fields. These sites had greater percent canopy cover of bare ground and forbs (than grasses) (Figs. 2-5) comprised of annual and biennial species (Table A-2). In addition, the lack of litter cover and depth on 1- and 2- year-old fields (Figs. 2-5) resulted in lower relative soil moisture levels on younger fields (Tables 28 and 29). As litter and total vegetation cover increased on older fields (Figs. 2-5), canopy cover subsequently increased and minimized moisture loss on these fields. Small mammal species diversity was relatively high on 2-year-old fields (H' = 0.71 - 0.90). Annuals and biennials continued to persist on 2-year-old fields where vegetation was less dense than older fields. In addition, the establishment of perennial grasses and forbs by the second enrollment year may have provided habitat conditions for a greater diversity of mammal species than 1-year-old fields. Cook (1959) found that Microtus spp. needed at least a year of litter accumulation to persist in an area. In this 77 study, Microtus spp. became abundant on CRP fields 2 2 years of age. The presence of 2 years of litter accumulation, and standing vegetation comprised of abundant forbs, may have provided sufficient cover and a diversity of plant species for a greater diversity of small mammal species than on 1-year-old fields. In addition, greater insect abundance (as a food source) on younger (1- and 2-year-old) CRP fields, documented by Millenbah (1993), may have contributed to a greater diversity of small mammals on these age classes of fields. Small mammal species diversity declined on fields 2 3- growing seasons old possibly due to changing habitat conditions. The greater percent canopy cover of grasses and litter cover on older fields provided the necessary habitat components for Microtus spp., however, not for other small mammal species such as Peromyscus spp. and thirteen-lined ground squirrels. On abandon agricultural areas in Michigan, Beckwith (1954) documented the same trend, that is, a transition from Peromyscus spp. to meadow voles as plant communities changed from being dominated by annual- biennials to perennial grasses. Five—year-old fields in August 1992 (Table 24), had a very high mean small mammal species diversity (H' = 0.98). This may be attributed to the abundance of Microtus spp. present in August which may be a food source for short-tailed shrews and least weasels (Linduska 1950, Baker 1983). Both short-tailed shrews and 78 least weasels were captured in August on 5-year-old fields. The occurrence of meadow voles on older CRP fields is consistent with the habitat preference information documented in the literature for the species. In Michigan, Linduska (1950), Getz (1961a) and Hallett et al. (1983) found meadow voles preferring moist areas with abundant grass-like vegetation. According to Getz (1961a), the presence of grasses and grass-like plants (Poa spp., Solidago spp., Aster spp.) may be a primary factor restricting meadow voles to certain habitats. Suitable vegetative structure and composition is essential in determining abundance and distribution of meadow voles since the vegetation serves as food and cover for the species (Getz 1961a, Batzli 1968, Birney et al. 1976). Klatt and Getz (1987) suggested that meadow voles occupy areas of dense vegetative cover and deeper litter cover to decrease risk of predation. Such conditions were characteristic of older CRP fields. Older (3 - 6-year-old) CRP fields were characterized by relatively greater percent canopy cover of grasses, litter cover and depth and total canopy cover than younger fields (Tables 14-17). Furthermore, Thompson (1965) found that captive meadow voles preferred introduced grasses and forbs to native species as food. The presence of short-tailed shrews, least weasels, opossums and thirteen-lined ground squirrels, most likely occurred on younger (1- and 2-year-old) fields because they 79 prey upon small mammals such as Peromyscus spp. and meadow voles (Linduska 1950, Eadie 1953, Whitaker and Mumford 1972). Linduska (1950) documented that short-tailed shrews were always associated with meadow voles and never trapped in the absence of voles. This relationship was also observed on CRP fields in this study (Figure A—2). Although 4-, 5-, and 6-year-old CRP fields generally exhibited lower species diversities than 2- and 3-year-old fields (Tables 24 and 25), they (4 - 6-year-old), however, had greater numbers of individuals (3-year-old fields also had high relative abundance)(Tables 19 and 20). These results may be attributed to older fields being comprised of mostly habitat specialists such as meadow voles, meadow jumping mice, and short-tailed shrews, that need dense cover and greater soil moisture to persist in an area (Getz 1961a and 1961b, Whitaker 1963, Whitaker and Mumford 1972). Subsequently, these species primarily occurred on older sites. The high abundance of meadow voles, especially in July and August, was probably a result of a successful breeding season in spring and early summer. Numerous Microtus nests were observed under clumps of dead grass on older CRP fields. A factor that may have affected relative abundance and distribution of small mammals across all age classes is weather. In 1992, percent capture rates on CRP fields were less than in 1993 (Tables 26 and 27). Lower capture success 80 in 1992 may have occurred due to rain during each trapping period (May, June, July and August). O'Farrell (1974) reported that cloudy weather and rain resulted in increased rodent activity, however, depending on severity of the rain. After several rain showers in 1992, it was frequently observed that traps had sprung, most likely set off by heavy rains, rendering traps inaccessible to small mammals. O'Farrell (1974) also suggested that greater intensity of moonlight had a negative affect on activity. In the field, O'Farrell (1974) observed that rodent activity was low when the moon was present. However, rodent activity increased whenever the moon was hidden by clouds. All trapping in this study occurred during the full moon. This factor could have impacted absolute capture success both years. As previously mentioned the distribution small mammal species, particularly meadow voles and Peromyscus spp., may be influenced by species interactions. The patterns of small mammal captures (Figure A-2) indicated that Peromyscus spp. were not trapped in areas having a greater relative abundance of meadow voles. When meadow voles were widely distributed on trapping grids, Peromyscus spp. were conspicuously absent or trapped primarily on assessment lines. Hallet et al. (1983) suggested that at high relative densities of meadow voles and meadow jumping mice, Peromyscus spp. were restricted to habitats not occupied by either Microtus or Zapus. Species exclusion, however, does 81 not appear likely since meadow voles are diurnal and Peromyscus spp. are mostly nocturnal (LoBue and Darnell 1959). LoBue and Darnell (1959) suggested that the differences in the response of these 2 species to vegetative cover is in part due to their activity patterns. Because many small mammals are restricted to a single, horizontal plane of the environment (Grant 1972), it is difficult to avoid interactions, especially if block size (habitat) is small. CRP fields in this study, may be considered large blocks relative to normal movements patterns of small mammals, however, home range was not evaluated in this study. If block size is large enough and appropriate habitat conditions are provided within fields, 2 or more mammal species may coexist by use of microhabitats (Morris 1984). Since most CRP fields were bordered by agricultural fields on at least 1 side, captures of Peromyscus spp. on assessment lines may be a result of movement of Peromyscus spp. from these areas. Of the 9 species captured on CRP fields, the distribution of meadow jumping mice and thirteen-lined ground squirrels only occurred under specific habitat conditions. Although meadow jumping mice were captured on all age classes of CRP fields with the exception of 1-year-old fields, they were captured most often on field 8A in 1992 and 1993 (5- and 6-year-old fields, respectively) (Figure A-2). This field had standing water and greater 82 soil moisture levels throughout the spring and early summer than other fields. Getz (1961b) and Whitaker (1963) found Zapus only in moist habitats with greater abundance in areas near standing water. In addition, older (3- and 5-year-old/4-, and 6-year-old) fields each year tended to have greater percent grass canopy for food and cover (Tables 14-17). Therefore, moisture, cover and food on this particular CRP field may have provided optimal habitat for meadow jumping mice. Although meadow jumping mice abundance was low on CRP fields, they tended to decrease temporally. This was probably due to drier conditions in late summer. Conversely, thirteen-lined ground squirrels were trapped primarily on 1- and 2-year-old fields which were characterized by greater percent bare ground and low soil moisture levels. According to Baker (1983), thirteen-lined ground squirrels prefer areas where their vision is not impaired by vegetation so they can forage more effectively. In addition, the need for burrow sites and foraging strategies probably restrict these species to drier soils with reduced cover. The woodland jumping mouse (Napaeozapus insignis) was the only species trapped in an adjacent habitat and not captured on CRP fields. Napaeozapus was captured in an adjacent woodlot, however, a geographic distribution of this species in Michigan shows that it has not been recorded in Gratiot County (Baker 1983). It is uncertain whether the 83 species was correctly identified. All other species trapped in adjacent cover types corresponded with their habitat preferences documented in the literature. For example, only Peromyscus spp. were trapped on agricultural fields, while meadow voles and Peromyscus spp. were trapped on adjacent CRP/open grassland fields. Most likely, small mammal species move between CRP fields and adjacent habitats if conditions are suitable. Subsequently, habitat size, juxtaposition, and interspersion may play critical roles in species diversity and richness on CRP fields. According to the theory of island biogeography, habitat size and degree of isolation affects species abundance and diversity on islands (Hunter 1990). Areas that are small and isolated tend to support fewer species with small populations. While small mammals may not require or have access to get to a diversity of habitat types, they may require a diversity of habitat conditions (structure and composition) to meet their habitat needs. CRP fields that remain undisturbed for a number of years may provide the necessary conditions required by a diversity of small mammal species. In addition, since the structure and composition of vegetative species on CRP fields are influenced by field age, a variety of age classes throughout the agricultural landscape may diversify small mammal populations. 84 Predator Scent Stations Although no significant differences in visitation rates for medium size mammals were found among age classes of CRP fields visitation rates tended to decrease with field age (Tables 30 and 31). This trend suggests that younger fields, characterized by greater percent cover of bare ground and less total canopy cover, provided sparse vegetation cover which may have allowed easier movement of mammals into the fields. In addition, since 2- and 3-year- old fields had greater small mammal species diversity (Tables 24 and 25), they may have provided a variety of prey species for larger mammals. All the species identified on scent stations (Tables 32 and 33) with the exception of woodchucks have been known to consume small mammals and insects (Linduska 1950, Baker 1983). Millenbah (1993) suggested that greater biomass and diversity of insects were available on younger (1 - 3-year- old) CRP fields. The remains of such animals as mice (of the genus Peromyscus), voles (of the genus Microtus), meadow jumping mice, house mice, and thirteen-lined ground squirrels have been found in the digestive tracts of Michigan striped skunks (Baker 1983). Raccoons, the most frequent visitors among all age classes of CRP fields, also may feed heavily on meadow voles and Peromyscus spp. (Stuewer 1943). Raccoons, skunks, opossums, and woodchucks were visually observed on younger CRP fields by researchers 85 as well as tracks noted at scent stations. Relatively more cats and dogs associated with older fields were most likely due to the proximity of these fields to residential areas. RECOMMENDATIONS The CRP has potential to be the most beneficial land retirement program for wildlife species to date (Berner 1988). Since the initiation of the CRP in 1985, nearly 14 million ha of farmland have been enrolled in the program causing dramatic changes in the composition of agroecosystems. Results from this study indicate that leaving CRP fields undisturbed during the first 6 years of enrollment may reduce plant and animal diversity as fields age. As a result of succession, fields become less structurally complex which may directly impact mammalian communities. Maintaining a diversity of age classes of CRP fields throughout agricultural landscapes is important to provide habitat for a diversity of mammal species that may not occur in croplands. Manipulating CRP fields after 3- to 5-growing seasons may diversify the structural complexity of individual CRP fields for small mammals and create a diversity of successional stages of grasslands across a broader landscape. It is important, however, to consider size, shape, interspersion and juxtaposition of CRP fields since 86 87 these factors will also affect wildlife use. For example, most small mammal species do not need more than 1 habitat type to meet their requirements (Grant 1972), but more mobile species like birds may require different habitat types for food, cover, and nesting. Since CRP has potential to benefit a diversity of wildlife species, in addition to mammals, other species' requirements should be considered prior to management decisions. For example, Millenbah (1993) also suggested that periodic manipulations after 3- to S-growing seasons may help maintain greater diversity, density, and productivity of avian species on CRP fields. Periodic manipulations by burning, mowing or disking could be used to alter the structural complexity of CRP fields. Before initiating manipulations, however, considerations of their effects on wildlife species should be evaluated since shifts in small mammal populations can be directly related to changes in plant communities (Moreth and Schramm 1972, Krefting and Ahlgren 1974). Disturbances to small mammal habitats will tend to have a greater effect on species composition rather than the abundance of small mammal species (Beck and Vogl 1972, Lemen and Clausen 1984). In addition to the type of disturbance and age at which manipulations will be implemented the season should also considered when planning manipulations for wildlife. For instance, in this study small mammal abundance was greatest 88 in July and August on CRP fields. If burning, mowing, or disking is initiated in May or June, species dependent on dense cover for nesting may be reduced. A disturbance initiated in late August or early September, may not allow adequate cover to develop to meet some small mammal species (e.g. Microtus) nesting requirements the following spring. However, mammal species will most likely return following the first year’s growth after manipulation. Continued enrollment of agricultural land into the CRP should be encouraged since it provides wildlife habitat for mammal species that otherwise may not be present in croplands. Although agricultural lands have their own unique vegetative composition and structure, they do not provide the complexity found in established grasslands. Reverting CRP lands to agricultural cropland may have significant impacts on a diversity of wildlife species, specifically, small mammals that depend on habitat conditions provided by these grassland ecosystems. Because the first CRP contracts expire in 1995, the future of wildlife populations in agricultural regions of the U.S. are uncertain. Awareness of the benefits of the CRP for wildlife is important if managers want to continue to use this program as a management tool. Communication among researchers, managers, landowners, and policy makers, will be essential in future planning and decision making for the CRP. Disseminating research results on the wildlife 89 values of CRP in journals and popularized magazines may help inform a diverse audience about the benefits of the program. APPENDICI ES 90 6.0a 6.4a ~.e~ m.m m.oa m.e~ umsasa 0.4 e.ma m.mm m.m «.ma a.mm sash m.aa e.oa a.mm e.m e.oa e.e~ mash e.a e.m s.o~ m.m m.m m.a~ an: coaumuamao0um Caz x02 coauouamao0um Caz x02 sumo: amuoa .9809 .QSOB amuoe .QE0B .0509 mama mama .mama can mama .smaasoas .maaa you AEUV COaucuamao0am amuou use AUV m0usumu0mfi0u manHGOE s00: .au< 0an09 9]. 1A a I: Osc 91¢ a SEVILLE PINE RIVER BETHANY WHEELER 3 91A % a 918 (:02 0C1 32 SUMNER ARCADA EMERSON LAFAYETTE “89¢ an NORTH NEW HAVEN NEWARK STAR HAMILTON 89A n a 9A 12A NORTH u SHADE FULTON WASHINGTON BLBA fl CRP fields Figure A-l. Michigan, 1992 and 1993. 0 Agricultural fields S - Soybean fields used in 1992 C - Corn fields used in 1993 SB - Soybean in 1992 / Corn in 1993 Study site locations by township in Gratiot County, 92 Table A-2. Plant species on Conservation Reserve Program fields (CRP) in 1992 and 1993 in Gratiot County, Michigan. Common name Alfalfa 13'2'3'4'5'5 Medicago sativa Aster 25L4 Aster spp. Avens 1 Genum spp. Bittercress 1 cardamine spp. Black Medick 1 Medicago lupulina Bouncing Bet 1'2 Saponaria officinalis Box Elder 4 Acer negundo Bull Thistle 5 Cirsium vulgare Canadian Thistle 2'3““5'6 Cirsium arvense Chickory 1'2'5'6 Cichorl'um intybus Chickweed ITLS Stellaria spp. Cinquefoil 1'2'3'5 Potentilla spp. Common Dandelion 1'2'3'4'5'6 Taraxacum officinale Common Groundsel 1'2'3'4'5'6 Senecio vulgaris Common Morning Glory 2'4'6 I pomoea purpurea Common Mullenl'2 Verbascum thapsus Common Plantain 1'2'3'4'5'6 Plantago major Common Ragweed 4 Ambrosia artemisiifolia Common Smartweed 3 Polygonum hydropiper Common Winter Cress 1'2 Barbarea vulgaris Crowfoot 1'2 Ranunculus spp. Curled Dock 1'2'4'6 Rumex cripus Daisy Fleabane 1'25L4'6 Erigeron philadelphicus Deptford Pink 1 Dianthus armeria Field Binweed 1'2 Convolvulus arvensis Field Pennycress 1'6 Thalaspi arvense Field Sorrel 1 Rumex acetosella Foxtail 1'2 Alopecurus spp. Goldenrod 1'2'3'4'5'6 Solidago spp. Table A-2. Continued. 93 Common name Hawthorne 4 Horsetail 4'5 Joe-Pye Weed 3'5 Kentucky Bluegrass Lady's Thumb 5 Lance-leaved Viol Milkweed 1' 5'6 MOSS 1'2'3'4'5'6 Mustard4'6 Orchard Grass Oxeye Daisy 5 Path Rush 1 Pussytoes 1 Quacking Aspen 3 Quackgrass 1'2'3'4 Queen Ann's Lace Red Clover 1'2'3'4 Red Elm 5'5 1,2,3,4,6 et 3 1,2,3,4,5,6 [5'6 1,2,3,4,5,6 Red-osier Dogwood 6 Redtop 3'4 Reed Canary Grass Ryegrass 3'5 Sandwort.1'2 Slender Wheatgras Smooth Brome Speedwell 1'3'4 Switch Grass 4 Timothy Grass Virginia Creeper Wheat 1'4 2'3, 2,3,5 S 1 2,3,6 4,5,6 2,6 Crataegus spp. Equisetum spp. Eupatorium spp. Poa pratensis Polygonum persicaria Viola lanceolata Asclepias spp. Bryophyta Cruciferae Dactlyis glomerata Chrysanthemem leucanthemum Juncus tenuis Antennaria spp. Populus tremuloides Agropyron repens Daucus carota Trifolium pretense Ulmus rubra Cornus stolonifera Agrotis gigantea Phalaris arundinacea Lolium perenne Arenaria spp. Agropyron trachycaulum Bromus inermis Veronica spp. Panicum virgatum Phleum pretense Parthenocissus quinquefolia Triticum aestivum Table A—2. Continued. 94 Common name Whitlow Grass 1'3 Wild Lettuce 2'4 Wild Peppergrass 1'2 Wild Strawberry 3 r 5 v 5 Yellow Sweet Clover 1'4'6 Scientific name Draba verna Lactuca canadensis Lepidium virginicum Fragaria virginiana Melilotus officinalis 3 Numbers following common name represent age classes that plant species were encountered. 95 Figure A-2. Small mammal species distribution on various age classes of Conservation Reserve Program fields in July 1992 and 1993 in Gratiot County, Michigan. +23 +m +95 +m 195.1 + +95 + iysi- + + 105+ + ++f$+++rsw+fsw+f$+rs +++++ 195.1 +- -+ +"S+ 93+” ++++++ ++++++ +PS _.)_PS Field 91A ps = Peromyscus spp. July, 1992 l-year old 96 Figure A-2. Continued. + + + + + + + + n+ + + + Field 91B pa = Peromyscus spp. July, 1992 1-year old 97 Figure A-2. Continued. F if j i— +— +— +PS+PS+- V$+ +PS+PS+ + + + PSPS+ + + + + +PS+PSPS+ +PS+ + ¥5P5+ + + + + + + +J°S+ 95+ + P’s + + +°5+J’5+"5+J’5 + ++++++ Field 91C 86 = Peromyscus Spp- July, 1992 1-year old 98 Figure A-2. Continued. +j++ + +W+ +fiV+ + + +++ + +""+ + +PV+NV++ ++++++WPW+++++WW++++ + + + 44v++ + + ++ +"V+”" + ++++++ Field 91A mv = meadow vole July' 1993 ps = Peromyscus spp. 2-year old sts = short-tailed shrew mjm = meadow jumping mouse 99 Figure A-2. Continued. +YLGS+++++ ++""++++++++++++++ ++++++' +Ms+ps++++ ”Gir+++++ mv a meadow vole Field 91B July, 1993 ps 8 Peromyscus spp. 2-year old tlgs = thirteen-lined ground squirrel 100 Figure A-2. Continued. + 'i' +ps+ + + + + + +++++++ + +"V+ + +~v +sm+ -f +”V+- + + +++++++++++ 1) 4+4- 1 + + + +++++++ + -+ +— +-+ ++++++ Field 91c mv = meadow vole July, 1993 sts = short-tailed shrew 2-year old pa = Peromyscus spp. 101. Figure A-2. Continued. V +Jg++++++ 'tMV‘+”VjJ“’j‘ j— 4— 5T5+' '+”VjJ“’j- -tMV—t “t"V-t' j— '+”Vj‘ ‘FMV j” j’ 'i‘ j" '+’w‘t”vj- j- j- aw + + www '+rw'+flv‘t”V—t j- j— *— + + + + _Lmv +HV 98+ 954_ Field 89A mv = meadow vole July, 1992 p8 = Peromyscus spp. 3-year old sts a short-tailed shrew 102 Figure A-2. Continued. d- —F g— j + A w i— ‘2— —t rt- ‘4“MV‘+‘ ’t' —t' 'j— 'j— _+' '4‘ I Ps—t‘ '4“ -+- “t' -+‘ -+- —t‘ —t‘ ‘j— ‘+' *t‘ ‘t‘ ‘j—Lw‘j- '+- 'j— '4— —+— +MV+W+ + + —t' -+' —t' —t‘ -+‘ —t‘ ‘F —F -F u_Mv fi— fi— '2— 4_ Field 393 mv meadow vole Peromyscus spp. July, 1992 p8 least weasel 3-year old 1" Figure A-2. 103 Continued. g. a. fi_ g_ 3_ J. ++++++ ++++++ ++++++ j- -+- -t- 'j- '+- -+- -F -+— ++++++ +++W+++ + + + + + + + -+— Field 89C mv = meadow vole July, 1992 3-year old 104 Figure A-2. Continued. _+_ + ++++++ +NV+”“+- +MV+MVifiv +- +”“+ -HW+”V+”V ++PW+++ i-'+ +JV+W+-+JW+ +mw+ imkt + +W+W+MI j- +MV+- +— ifivifiv +MV Field 89A mv a meadow vole July, 1993 4-year old Figure A-2. Continued. 105 -+- "t‘ -t‘ 'j— '4" 'j- '4’ '4‘STSfiflE¥T+‘ Ttxssj— *4” —H“V—l-‘“&>s+ + + 'j— 'ijv'i—LJL+4§¥S+J4V'+— 'j— ‘4‘ 'j- '4‘ ’+J4V'+J4V-ty““4—STst' ‘j— 'j— ‘4—MV'+J”Vj— ‘quv V V ++fi+t++++ _LMV + —+— + + + ._+_ 4‘ J Field 393 mv = meadow vole July, 1993 p8 = Peromyscus spp. 4-year old sts = short-tailed shrew 1w = least weasel tlgs = thirteen-lined ground squirrel 106 Figure A-2. Continued. 4— —FMV-+“V +— —+MV—+r« -t“V-+“V-+“V-+- -+“V—+NV HJ“’j— *t -+MV-+5Ys+“V + + + 4540+ + +W+ STs+ -l- +“V-i- +“V +“V srs—l—“V -l-“" -l—”" +“V -l-’”s‘¥s—l— + _+mv + + ++ _+_ Field 39c mv = meadow vole July, 1993 sts = short-tailed 4-year old shrew 107 Figure A-2. Continued. l 1"“ j ’ + jmv .+ imv 1. 1. jfiv +”V'+ +' -+ -+ ifiv + -t -+ '+ i’V jfiv + jfiv‘t '+ i’V + ++ + ++ + WPVWPW + -+ +- -+ -+ +- -+ +‘ -+ +- i- '+ +- +- +- -+ +- +' Field 1A mv = meadow vole July, 1992 S-year old 108 Figure A-2. Continued. + + + + i- -+ + +- i- + + +‘ +“"+' 4' i- 't + '+ + j- -+ + i- -+ + + +- i- -+ + i- -+ + + + +- —+ + + j' ‘+ +' j’“-+ + j‘ -+ + + V’#J + + + .+ + Field 8A mjm = meadow jumping July, 1992 mouse S-year old 109 Continued. Figure A-2. +++++++ V w++++++++w+++++++ +w++am ++++aa +. V mv = meadow vole Field 12A July, 1993 6-year old 110 Figure A-2. Continued. + 1... + +4 +4 1 + + + WW WW + W W W W W W + + + W W + + WWI» +94 + WWW W +5.88 Us W + + W W + W W+ +_ W + + 7") +1. + + mv = meadow vole Field 1A July, 1993 sts a short-tailed 6-year old shrew 1w = least weasel 111 Figure A-2. Continued. f + 1 H + W+ W+ W+ W W W W W i” 95+“ 1... mm +nv M + + W+ + WWWWWW+ + + + + W+ WW+ ~F‘V+ + + + + + ++++++ Field 8A mv = meadow vole July, 1993 ps = Peromyscus spp. 6—year old 112 Figure A-2. Continued. +_ 4. 1. i J_ 1. 1. J. +- +— -+ -+ -+ -+ 4— 4- 4— i— 4— 4— -P 4— 1; 4— i— 4— i— 4- -F -: -F i— +— +- +- +- +— +— 4— 9- +— +— -+ +— +— +— +- 4.6 if 4. 3_ 4. 4. a a a Field 9A mv = meadow vole July, 1992 sts = short-tailed shrew S-year old LIST 01" REFERENCES LIST OF REFERENCES Baker, R. H. 1983. Michigan mammals. Michigan State University Press, East Lansing, Michigan. 642pp. Batzli, G. O. 1968. Dispersion patterns of mice in California annual grassland. J. Mammal. 49:239-250. . and F. A. Pitelka. 1970. Influence of meadow mouse populations on California grassland. Ecology. 51:1027-1039. Basu, P. K., H. R. Jackson, and V. R. Wallen. 1978. Alfalfa decline and its cause in mixed hay fields determined by aerial photography and ground survey. Can. J. Plant Sci. 58:1041-1048. Beck, A. M. and R. J. Vogl. 1972. The effects of spring burning on rodent populations in a brush prairie savanna. J. Mammal. 53:336-346. Beckwith, S. L. 1954. Ecological succession on abandoned farm lands and its relationship to wildlife management. Ecol. Monogr. 24:349-376. ' Berner, A. H. 1984. Federal land retirement program: a land management albatross. Trans. N. Am. Wild. Nat. Resour. Conf. 49: 118-131. . 1988. The 1985 farm act and its implications for wildlife. Pages 437-465 in W. J. Chandler, and L. Labate, eds. Audubon wildlife report 1988/1989. The National Audubon Society., New York, N.Y. Birney, E. C., W. E. Grant, and D. D. Baird. 1976. Importance of vegetative cover to cycles of Microtus populations. Ecology. 57:1043-1051. Burger, L. W., E. W. Kurzejeski, T. V. Dailey, and M. R. Ryan. 1990. Trans. N. Am. Wildl. Nat. Resour. Conf. 55:74-83. 113 114 Castrale, J. S. 1983. Wildlife use of cultivated fields set aside under the Payment in Kind (PIK) program. Proc. Indiana Acad. Sci. 93:173-180. . 1985. Responses of wildlife to various tillage conditions. Trans. N. Am. Wildl. Nat. Resour. Conf. 50:142-149. Cook, S. F. 1959. The effects of fire on a population of small rodents. Ecology 47:278-291. Constable, G. 1985. Grasslands and tundra. Time-Life Books, Inc. Alexandria, Vir. 176pp. Cutler, M. R. 1984. Integrating wildlife habitat features in agricultural programs. Trans. N. Am. Wildl. Nat. Resour. Conf. 49:132-140. Daubenmire, R. F. 1959. A canopy coverage method of vegetational analysis. Northwest Sci. 33:43-64. Duebbert, H. F. 1969. High nest density and hatching success of ducks on South Dakota cropland adjustment program land. Trans. N. Am. Wildl. and Nat. Resour. Conf. 34:218-228. Eadie, W. R. 1953. Response of Microtus to vegetative cover. J. Mammal. 34:263-264. Edwards, W. R. 1984. Early ACP and pheasant boom and bust! - a historical perspective with rational. Pages 71-83 in R.T. Dumke, R.B. Sthiel, and R.B. Kahl, eds. Proc. Perdix III. Gray partridge and ring-necked pheasant workshop proc. Wisc. Dept. Nat. Resour. Farmer, A. H., R. L. Hays, and R. P. Webb. 1988. Effects of the conservation reserve program on wildlife habitat: a cooperative monitoring study. Trans. N. Am. Wildl. and Nat. Resour. Conf. 53:232-238. Freese, F. 1978. Elementary forest sampling. U.S.D.A. Agric. Handb. No. 232. 91pp. French, N.R. 1971. Small mammal studies in the U.S.- IBP grassland biome. Ann. Zool. Fennici. 8:48-53. Getz, L. L. 1961a. Factors influencing the local distribution of Microtus and Synaptomys in southern Michigan. Ecology. 42:110-119. 115 . 1961b. Notes on the local distribution of Peromyscus leucopus and Zapus hudsonius. Amer. Midl. Nat. 65:487-499. Golley, F. B., L. Ryszkowski, and J. T. Sokur. 1975. The role of small mammals in temperate forests, grasslands, and cultivated fields. Pages 223-241 in F. B. Golley, K. Petrusewicz, and L. Ryszkowski, eds. Small mammals: Their productivity and population dynamics. Cambridge Univ. Press. Grant, P. R. 1972. Interspecific competition among rodents. Annu. Rev. Ecol. and Syst. 3:79-107. Gysel, L. W., and L. J. Lyon. 1980. Habitat analysis and evaluation. Pages 305-328 in S. D. Schemnitz, ed. Wildlife management techniques manual. The Wildl. Soc., Washington, D.C. Hallett, J. C., M. A. O’Connell, and R. L. Honeycutt. 1983. Competition and habitat selection: test of a theory using small mammals. Oikos 40:175-181. Hansson, L. 1968. Population densities of small mammals in open field habitats in South Sweden in 1964-1967. Oikos 19:53-60. Harmon, K. W., and M. W. Nelson. 1973. Wildlife and soil considerations in land retirement programs. Wildl. Soc. Bull. 1:28-38. Haufler, J. B. 1990. Static management of forest ecosystems. Pages 123-130 in J. M. Sweeney, ed. Management of dynamic ecosystems. North Cent. Sect., The Wildl. Soc., West Lafayette, Ind. Higgins, K. F., D. E. Nomsen, and W. A. Wentz. 1987. The role of the conservation reserve program in relation to wildlife enhancement, wetlands, and adjacent habitats in the northern great plains. Pages 99-104 in J. E. Mitchell, ed. Impacts of the conservation reserve program in the great plains. U.S. For. Serv. Gen. Tech. Rep. RM-158. Hunter, M. L. 1990. Wildlife, forests, and forestry: principles of managing forests for biological diversity. Prentice Hall Inc, Englewood Cliffs, N.J. 370pp. 116 Isaacs, B., and D. Howell. 1988. Opportunities for enhancing wildlife benefits through the conservation reserve program. Trans. N. Am. Wildl. and Nat. Resour. Conf. 53:222-231. Joselyn, G. B. and J. E. Warnock. 1964. Value of federal feed grain program to production of pheasants in Illinois. J. Wildl. Manag. 28:547-551. Keever, C. 1950. Causes of succession on old fields of the Piedmont, North Carolina. Ecol. Monogr. 20:229-250. Klatt, B. J. and L. L. Getz. 1987. Vegetation characteristics of Microtus orchrogaster and M. pennsylvanicus habitats in East-Central Illinois. J. Mammal. 68:569-575. Knapp, A. K. and T. R. Seastedt. 1986. Detritus accumulation limits productivity of tallgrass prairie. Bioscience 36:662-668. Krefting, L. W. and C. E. Ahlgren. 1974. Small mammals and vegetation changes after fire in a mixed conifer- hardwood forest. Ecology 55:1391-1398. Lemen, C. A. and M. K. Clausen. 1984. The effects of mowing on the rodent community of a native tall grass prairie in Eastern Nebraska. Prairie Nat. 16:5-10. Lindzey, F. G., S. K. Thompson, and J. I. Hodges. 1977. Scent station index of black bear abundance. J. Wildl. Manage. 41:151-298. Linduska, J. P. 1942. Winter rodent populations in field- shocked corn. J. Wildl. Manage. 6:353-363. . 1950. Ecology and land-use relationships of small mammals on a Michigan farm. Michigan Dept. Conserv. Game Div. 144pp. Linhart, S. B. and F. F. Knowlton. 1975. Determining the relative abundance of coyotes by scent station lines. Wildl. Soc. Bull. 3:119-124. LoBue, J. and R. M. Darnell. 1959. Effect of habitat disturbance on small mammal population. J. Mammal. 40:425-436. M'Closkey, R. T. and Fieldwick. 1975. Community structure in sympatric rodents. Ecology. 57:728-739. 117 Millenbah, K. 1993. The effects of different age classes of fields enrolled in the conservation reserve program in Michigan on avian diversity, density, and productivity. M.S. Thesis, Michigan State Univ., East Lansing. 107pp. Miller, R. P. 1980. Simultaneous statistical inferences. Second ed. Springer-Verlag, New York, N.Y. 299pp. Miranowski, J. A. and R. L. Bender. 1982. Impact of erosion control policies on wildlife habitat on private lands. J. Soil and Water Conserv. 33:289-291. Moreth, L. H. and P. Schramm. 1972. A comparative survey of small mammal populations in various grassland habitats with emphasis on restored prairie. Pages 79- 84 in L. C. Hulbert tech. coord. Proc. third mid-west prairie conference. Kansas State University, Manhattan. Moro, D. 1991. The distribution of small mammal species in relation to heath vegetation near Cape Otway, Victoria. Wildl. Res. 18:605-618. Morrison, D. F. 1976. Multivariate statistical methods. McGraw-Hill. New York, N.Y. 495pp. Morris, D. W. 1984. Patterns and scale of habitat use in two temperate-zone, small mammal faunas. Can. J. Zool. 62:1540-1547. Nudds, T. D. 1977. Quantifying the vegetative structure of wildlife cover. Wildl. Soc. Bull. 5:113-117. O'Farrell, M. J. 1974. Seasonal activity patterns of rodents in a sagebush community. J. Mammal. 55(4):809-823. Ott, L. 1988. An introduction to statistical methods and data analysis. PWS-Kent Publishing Company. Boston, Mass. 582pp. Peet, M. R., R. Anderson, and M. S. Adams. 1975. Effect of fire on big bluestem production. Amer. Midl. Nat. 94:15-26. Rice, E. L. and R. L. Parenti. 1978. Causes of decreases in productivity in undisturbed tall grass prairie. Amer. J. Botany. 65:1091-1097. 118 Ryan, M. R. 1986. Nongame management in grassland and agricultural ecosystems. Pages 117-136 in J. B. Hale, L. B. Best, and R. L. Clawason, eds. Management of nongame wildlife in the midwest: a developing art. The Wildl. Soc. Shannon, C. E. and W. Weaver. 1949. The mathematical theory of communication. Univ. Illinois Press, Urbana. 177pp. Siegel, S. 1956. Nonparametric statistics for the behavior sciences. McGraw-Hill Book Co. New York, N.Y. 312pp. Smith, M. H., R. H. Gardner, J. B. Gentry, D. W. Kaufmam, and M. H. O'Farrell. 1975. Density estimations of small mammal populations. Pages 25-53 in F. B. Golley, K. Petrusewicz, and L. Ryszkowski, eds. Small mammals: Their productivity and population dynamics. Cambridge Univ. Press. Stubbendieck, J. 1987. Historical development of native vegetation on the Great Plains. Pages 21-28 in J. E. Mitchell, ed. Impacts of the conservation reserve program in the Great Plains. U.S. For. Serv. Gen. Tech. Rep. RM-158. Stuewer, F. W. 1943. Raccoons: their habits and management in Michigan. Ecol. Monogr. 13:203-258. Thompson, D. Q. 1965. Food preferences of the meadow vole (Microtus pennsylvanicus) in relation to habitat affinities. Amer. Midl. Nat. 74:76-86. USDA. 1975. Soil survey of Gratiot County, Michigan. Soil Conservation Service. 142pp. USDA. 1987. Environmental assessment and final regulatory impact and flexibility analysis for the highly erodible land conservation provisions of the Food Security Act of 1985. Soil Cons. Serv. Washington, D.C. 55pp. Whitaker, J. O. 1963. A study of the meadow jumping mouse, Zapus hudsonius, in central New York. Ecol. Monogr. 33:215-254. . and R. E. Mumford. 1972. Food and ectoparasites of Indiana shrews. J. Mammal. 53:329-335. HICHIGRN STRTE' UNIV. LIBRRRIEs llllljllj WNWWWI”WNW‘IillHlWHl 31293010200743