—'—’—‘— _—_——1 _— —.— —— ____# —— __— — H WHUH‘W "—4 :83 wow \iiiiiijgg \\\i\\\\.~\\\\\\ \\\\\\‘ ijii This is to certify that the thesis entitled Avian Breeding Use of Constructed and Established Wetlands in Chippewa County, Michigan presented by Michael Joseph Monfils has been accepted towards fulfillment of the requirements for Master of Science degree in Fish. & Wildl. b M§ (M, V' Major professor \ Date March 7, 1996 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution LIBRARY ‘7 Michigan State University PLACE IN RETURN BOX to remove We checkout from your record. TO AVOID FINES return on or beiore dete due. DATE DUE DATE DUE DATE DUE AVIAN BREEDING USE OF CONSTRUCTED AND ESTABLISHED WETLANDS IN CI-IIPPEWA COUNTY, MICHIGAN By Michael Joseph Monfils 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 1 996 ABSTRACT AVIAN BREEDING USE OF CONSTRUCTED AND ESTABLISHED WETLANDS IN CHIPPEWA COUNTY, MICHIGAN By Michael Joseph Monfils In order to determine the response of breeding waterfowl and other avian species, breeding bird use was compared between three constructed wetlands and three established wetlands in Chippewa County, Michigan. Waterfowl use of the open water/aquatic bed zone was monitored using weekly breeding pair and three brood surveys each year in 1994 and 1995. Other breeding birds were surveyed by vegetation zone using circular plot surveys and nest searches during early, mid, and late season periods. Breeding pair densities of dabbling duck species at constructed wetlands were higher than nearby established wetlands. A similar pattern of use on constructed wetlands by waterfowl broods was observed. Nineteen breeding bird species used the study sites, and ten species were common to constructed and established wetlands. Avian use of the scrub-shrub and flooded emergent zones was similar. Densities of birds in the semb- shrub, wet meadow, and flooded emergent zones were similar. Three species were unique to constructed wetlands, and six were unique to established wetlands. Species unique to constructed wetlands were associated with scrub-shrub communities, while species unique to the established wetlands were associated with wet meadow or flooded emergent zones. ACKNOWLEDGMENTS This study is firnded through the Wildlife Conservation and Restoration Act (Pittman-Robertson) Project Number W-127-R Additional funding and field assistance was provided by Resource Management Group, Incorporated (RMG). Special thanks to Harold Prince, my major professor, for all of his guidance and assistance during this project. I would like to thank my committee members Donald Beaver and Thomas Burton for their advice and assistance. Thanks to Charles Wolverton and the RMG staff for the financial support for field expenses and their cooperation with this project. William Moritz of the Michigan Department of Natural Resources (MDNR) was important in getting this project started. I would like to thank Gregory Soulliere, wildlife biologist with the MDNR, for his cooperation and assistance. MDNR workers Timothy Maples and David Wilson provided some field assistance. Thanks go to Zachary Olson for his assistance during the 1995 field season. Ronald Brown and his staff at the RMG Escanaba omce also provided financial support and field assistance. Daniel Hayes and James Bence of the Department of Fisheries and Wildlife provided statistical advice. John Niewoonder provided advice with GIS map production. I would like to thank Jennifer Derby, Cathy Flegel, John Niewoonder, Michael Whitt, and Charlotte Young for their support. Thanks to Anne Vaara for her work with plant identification. The permission of local landowners ' to investigate wetlands on their property is appreciated. iii TABLE OF CONTENTS . Page LIST OF TABLES .............................................................................................. v. LIST OF FIGURES ............................................................................................. vi INTRODUCTION .............................................................................................. 1 DESCRIPTION OF STUDY AREAS ................................................................. 3 Constructed Wetlands .............................................................................. 3 Established Wetlands ............................................................................... 10 METHODS ......................................................................................................... 15 Experimental Design ................................................................................ 15 Habitat Zones .......................................................................................... 15 Waterfowl Breeding Pair Surveys ............................................................. 16 Waterfowl Brood Surveys ........................................................................ l6 Avian Surveys of Vegetation Zones .......................................................... 18 Analysis ................................................................................................... 19 RESULTS ........................................................................................................... 21 Habitat Zones ........................................................................................... 21 Waterfowl Breeding Pair Surveys ............................................................. 23 Waterfowl Brood Surveys ........................................................................ 26 Avian Surveys of Vegetation Zones .......................................................... 28 DISCUSSION .................................................................................................... 33 APPENDIX ........................................................................................................ 4O LITERATURE CITED ....................................................................................... 41 iv Table LIST OF TABLES Comparison of size (ha) and proportion (%) of habitat zones distributed on constructed and established wetland study sites ................ Estimated waterfowl breeding pair densities (i 1 SE) by week and wetland type for 1994 and 1995 ............................................................. Estimated waterfowl breeding pair densities (i 1 SE) by wetland type and year ................................................................................................. Estimated waterfowl breeding pair densities (x 1 SE) by species and wetland type ........................................................................................... Significant (P < 0.05) correlations of mean waterfowl breeding pair densities with several physical features of the study sites (N = 6) ............ Estimated density ()2 t SE) of Class II ducklings by period, site, and wetland type in 1995 .............................................................................. Estimated densities (x 1 SE) of Class II ducklings by species and wetland type in 1995 .............................................................................. Avian species determined to be breeding based on 1994 and 1995 surveys, listed by wetland type, study site, and vegetation zone (SS = Scrub-shrub, WM = Wet Meadow, and FE = Flooded Emergent) .............................................................................................. Density of birds per hectare (i 1 SE) (number of plots sampled in parentheses) and coemcient of similarity indices by vegetation zone ....... A list of common and scientific names of breeding birds identified during 1994 and 1995 surveys (AOU 1982) ............................................ Page 22 23 24 25 26 27 27 29 32 40 Figure LIST OF FIGURES Chippewa County, Michigan area map. Numbers indicate location of study sites ........................................................................................... Adamo wetland site map ......................................................................... LaMantia mitigation project site map ...................................................... Pothole mitigation project site map ......................................................... Thirteen Mile Road beaver pond site map ....... > ........................................ Munuscong Bay coastal irnpoundments site map ..................................... St. Mary’s River coastal wetland site map ............................................... ll 12 14 INTRODUCTION The formation of both state and federal wetland protection laws and a policy of ”no net loss" accentuates a number of important wetland values. When unavoidable impacts to wetlands are permitted, these laws often require the construction of mitigation wetlands to offset the loss of natural areas. Although constructed wetlands have been used to ofl‘set some of the lost wildlife value in the form of waterfowl breeding habitat, there have been few efi‘orts to evaluate waterfowl use of such wetlands (Leschisin et al. 1992). By relying on constructed or restored wetlands to replace natural wetlands, it is assumed that these wetlands will provide the same functions as the natural wetlands. Even in the prairie pothole region, where restoration has occurred since the 1930s, the question of whether formerly drained and cultivated wetlands can fully recover as habitats for breeding birds is unclear (Delphey and Dinsmore 1993). Although the construction/restoration of wetlands is being encouraged by programs such as the North American Waterfowl Management Plan (N AWMP), little is known as to the extent that these areas will be used by breeding waterfowl. Eastern Chippewa County, Michigan, is very conducive to wetland restoration and construction. A portion of the eastern Upper Peninsula has also been identified in the NAWMP as a grassland/wetland area of importance. Pickford and Rudyard soils are common throughout this region (USDA 1992). These soils are somewhat poorly drained 1 2 and poorly drained, respectively. Both exhibit slow permeability, which makes these soils very suitable to wetland formation. Mitigation projects involving wetlands constructed for the purpose of providing habitat for waterfowl production have been implemented in this region. This provides an opportunity to compare the fimction of constructed wetlands with established wetlands, and to determine how well constructed areas function in providing avian breeding habitat. It is expected that mitigation wetlands which develop vegetation zones similar to established areas, should also obtain similar densities and diversities of breeding waterfowl and other avian species. Delphey and Dinsmore (1993) believe that systematic comparisons between restored and natural wetlands are needed to facilitate evaluation of restoration. I set out to evaluate waterfowl and other avian breeding use of three constructed and three “establish ” wetland sites. The term “establish ” refers to areas with wetland plant communities of similar structure and hydrological conditions which have been in existence for a minimum of twenty-eight years. At established sites much of the nearby upland areas contain woody plants. All of the constructed wetlands were two or three years old, with wetland plant communities becoming developed and nearby recently disturbed upland areas dominated by herbaceous plants. DESCRIPTION OF STUDY AREAS Wands W This wetland is located adjacent to the beaver pond site, directly north of 13 Mile Rd. (T45N, RlE) (Fig. 1). The area is approximately 4.1 hectares (ha) in size and was created by a private landowner in the fall of 1993 through the construction of a small berm which blocked a small stream (Fig. 2). There was some existing willow (SaIix spp.) and speckled alder (Alnus rugosa) scrub-shrub wetland in this area which is now flooded. Common grasses at this site included canary reedgrass (Phalaris arundinacea), Agrostis spp., and F estuca spp.. Emergents such as woolgrass (Scirpus qperinus), sofi rush (Juncus efirsus), and sedges (Carex spp.) are found in the wet meadow zone. Some small areas of cattail (Typha spp.) are found. Ill' ... '[IIIEI This area was created as mitigation for wetlands lost in the construction of a mall project in Sault Ste. Marie. It is located on Killackey Rd. between Six and Seven Mile Rds. (T46N, RlE) (Fig. 1). Final construction was completed in the summer of 1993. This area was designed to create approximately 27 ha of seasonally, semipermanently flooded marsh attractive to breeding and migrating waterfowl (Fig. 3). At this time there 3 4 T l Adamo Wetland 2 LaMantia Mitigation Project 3 Pothole Mitigation Project 4 13 Mile Rd. Beaver Pond 5 Munuscong Bay Impoundments 6 St. Mary's Coastal Wetland ..i-[ro I-75 -129 6O . 3: I Riv ideDr. Figure l. Site map of Chippewa County, Michigan. Numbers indicate location of study sites. 9 e . #9 0“. . I O ‘0. I O O ’9 9 . 9:9'9z9z9:9.9:9‘ '9:9:9‘9:9:9;9:9:9.9: . - ......IOII. ..‘O‘ ....OOIIQOO I Q 0.0.0‘9. '0‘ I OOO‘GIOVOO . ......5 I O. ......O..‘ l.. O. I .9. .9 .. ......OQ.‘ 0.. ‘ 0.... 9 . Q 0 O. ......O .' g 9 9 9 9 9 9 9 9 ' O I . I . \ 9 I 9 C 9 09999999 9 I 9‘9‘9’9‘9’9’9‘9’9’9. 9 .‘ .. \ 9‘9z9t9z9:9’9:9‘9' O‘fl’.‘ ‘ O . ‘ O 9 ‘9‘9’9 ' 9 9 9 ”9'9 9'9 9‘9’9‘9’9 ‘ - '9’9'9 9'9‘ 9'9‘9'9‘9'9' '9’9‘9 , 99 .0. 9.0 9999‘ \ ' 9 9 ' ~ ' _ 9 9 9 9 O", . . . v 9 9 9 A I ' ‘9' 9'9 9 9’9 9 - .'9'9‘9'9’9‘I . 9 9 9 9 9 9 . ‘ . ...... "" ‘ . . 9 9 9 9 9 ‘9 I . I . I . I . I . I I . I I I I I I - ------ BreedingPair Survey Route X Breeding Bird Survey Area Figure 2. Adamo wetland site map. 9 99 9 99 0'9 99 IOIVOOOCO‘IIIO ’aII'OOOfioIQ\% 9 9 959.. {929293929 ‘ v 9’9'9 9 9 .. .‘I'..:‘:'3°1‘3'9 . . 99.9.9999999 ' 9 9 '9 9 ‘ 9 9 ..., .9.9.9‘ ’ 99999 “‘9’. 9 ’ ‘0‘ 5 a e 9 9 9 9 9 O . . ‘9’ 9.9 9‘. 0.9.0.9. 9 9.9 . DOD. 9'9 9‘9'9'9‘9’9’9’9’9‘9‘ '9'9 999909999909 "O".IO..C I .90.... O. 99 9999 99 9 999 9 9999 9 9 999 9'9 ‘9‘9'9’9‘9’9’9 99.999999999 ”. 9 ‘0‘. 9:9:9:9'9'9 9' . . . ‘9 9 9 Q 0 49 999 9‘9’9 '..'.9...' "%"”‘hfl Q \ I . ‘ :9 9 ..V. .9 2949.9 ,,:1.:9:9:9:9; " .‘9’9.9,9.9,9. 9’ 99 9‘9 9:9: 99 939:9:9’9’ 9 9 9 9 9 Q ' O 9 9 9 9 9 9 '9‘9’9‘9‘9 '9 6‘. . / Scale: 1" = 40 m Open Water/Aquatic Bed Emergent Vegetation @ Scrub-Shrub Vegetation . Brood Observation Point 6 .-.-e-IOI—l-O-e- 9 '0‘. 9'9’ . D . . O . O O C O O . Q . ......O. O. % O ‘ . 0 3’3“ ’99 O. '1 fl 9 f .9 e 909 flefifi 09 99 9 ”fee 999§VV593J§33§ e 999999999 999 99999999 9 O 9 99 9999999 9 99 999 9 999 9 99 999 .° . 9 9 9 9 9 9 9’9‘9':°:'9 9.9.9 99999999999999 9 99999999 909999 9999999 999999 99999 9999 99999 999 9 9 §v 9 O O O .9 '9' O O 0 9292929' 92 0 6 {$5 9 9 O. O O O O. 0.90 é”? ... gay q’e 9 O V“ ‘1‘759 9.0.0.9: ‘ 0'9 0 O e; :939 a. 5" O ......I. %%%%% O. . O Scale: 1" = 100 m a Open Water/Aquatic Bed . Brood Observation P011“ Emergent Vegetation -----.-.BreedingPair E . Survey Route Scrub-Shrub Vegetation X Breeding Bird Survey Area Figure 3. LaMantia mitigation project site map. 7 is approximately 19.4 ha of wetland present. The wetland was created via the construction of clay berms with concrete and rip-rapped spillways to maintain water levels (Fig 3). Concrete culvert nesting structures, loafing logs, and wood duck nest boxes were also incorporated into the design. White spruce (Picea glauca), bulrush (Scirpus spp.), buttonbush (Cephalanthus occidentalis), lake sedge (Carex locustris), and sage pondweed (Potamogeton pectinatus) were planted in the fall of 1993. Some areas of willow scrub-shrub vegetation remained afier construction. Grasses such as panic grass (Panicum spp.) and timothy (Phleum pratense) are common in the surrounding uplands. Emergents such as woolgrass, soft rush, and sedges such as Carex praegracilis, C. stipata, and C. Iimosa are found in the wet meadow zone. A small band of flooded cattail is found in the upper pool (west side of wetland). Ellll” . 2' ENE] This complex is located east of M—129 on Riverside Drive (T44N, RIE) (Fig. 1). Eighteen small wetlands ranging in size fi'om 0.05 to 1.0 ha, were constructed as mitigation for wetlands lost in the US 2 road expansion (Fig. 4). This area was designed by the Michigan Department of Natural Resources (MDNR) in conjunction with the Michigan Department of Transportation (MDOT) and Ducks Unlimited for the purpose of creating a waterfowl breeding and hunting area. These wetlands were created, through excavation and berm construction, in approximately 72 ha of fallow farm field and total approximately 8.0 ha of wetland area. These areas were constructed in the summer of 1992. Spoil banks were seeded with a mixture of timothy, orchard grass (Dactylis glomerata), treefoil (Lotus comiculatus), and redtop grass (Panicum rigidulum) to '0 ... ... §.. -- -- - Breeding Pair Survey Route I Open Water/Aquatic Bed Brood Survey Group Emergent Vegetation x Breeding Bird Survey Area Cl Small Potholes N Scale: 1" = 135 m Figure 4. Pothole mitigation project site map. 9 promote dense nesting cover (Soulliere 1993). The natural seed bank was relied upon for the colonization of wetland plants. This complex is located within the Munuscong Bay Wildlife Management Area (MBWA). Fecha spp. grasses are common in the surrounding uplands. Wet meadow zones are dominated by woolgrass, sofi rush, Carex vulpinoidea, and C. sn'pata. Small stands of flooded cattail are present in some potholes, and hardstem bulrush (Scirpus acutus) is found sporadically. 10 Establishedflctlands W A beaver pond 12.1 ha in size is located directly south of 13 Mile Rd (T4SN, RIE) (Fig. 1). This area has been in existence for at least 40 years according to the landowner (Fig. 5). It is dominated by emergent and aquatic bed vegetation zones interspersed with dead shrubs. The margin of this wetland is surrounded by scrub-shrub vegetation dominated by willow and speckled alder. Cattail and burreed were the most common emergents at this site. This area provides an example of long-term vegetative succession afier an area is impounded and stabilized. Three areas of coastal wetland on Munuscong Bay (T44N, RlE) (Fig. 1) were stabilized in 1965-66 via a system of dikes (Fig. 6). This allowed water levels to be controlled independently of the natural water level fluctuations of the St. Mary’s River and Lake Huron. Zones of scrub shrub, emergent, and open water/aquatic bed vegetation are present within the 348.9 ha diked areas. Scrub—shrub zones contain willow, speckled alder, and Spiraea spp.. Sedges such as Carex lacustris, C. Mama, and C. aquatilis, and grass (Calamagrostis canadensis) are found in the wet meadow zone. Flooded emergent zones found at this site are dominated by cattail and hard-stem bulrush. fififig 9 %%VVVVVI¢§ % 9999999 9 9%ggvvv % .¢ 9999999 9999999 . . ‘ O . O . O O O . 9 %%%%%fl%% . ,99999 9 99 9999 9 9 f¢§vv 9&595%& 3’3§"99§5$V’ff9§ 99999999999 999999 99999999999 3.9.9.9.9.9.9.9.9.9.9.9. 9 .39. . .93 9 9 9 9.9.9.9.9.9.9.9,9.9.9 9.9.9.999: 9 99 9999 99991 O 000. §5%9%&&’ 99“ 9 .9.9.9 9 9.9.9. 9 9 9 9 9 .000 Q’ . ‘OO’ 99999 I 99%puflp . J5 ”¢”;” 9999 9 99999999: 99 9999 9 999999999 999999999 999 999999 999999999 99 999999 9999999 99 99 9 99999999 9999 9 999 9999‘ 9 ......OQOVOC 9 9 9 9999339 ‘%&h‘&’ 9’ ’ 9.9.99 999: 99 9 ..9‘ . '9 9 9'9’9'9’9’9'9 9 9‘ 9 9 9. ’9’9'9’9 9 9 9 '9'9’9 b 99 99999 9999 9999999 9999 5%‘3‘u5‘5‘fihfi‘ %%%&%%%%%%~&% .......99..... 99 99 9999999 ygn 3 3 g 9 %~&&99¢ 99 9 9 ° 9% 9 9 99%9 ’ 06. .0000 999999999 999999999 9 9 999999999 9 9 9909999999999 9 999999999999 9 9 99999999 9 99 9999999 9 99999999 9999999 9 999 999999999 999999999999999999 997‘99999999999999990999999 " 55". Wfifi"‘%fifl&flfihfi&%§‘ 999% 9 9 9%}999999 \ O . Q 0...... . fi”?§?’fiafi§3¢§ 9999999999999. ”haflgggvvwa¢3 9.9.9.9; 9 9 9 9 9 99 9999 9 9 3V%%¢¢Q54€<€{g¢4‘ 999999999999999 99 999 99 9 9’; . Q 9&959599Q’9 99999999 9 909 .....'.......O.. 999999999999999999 ”5%flflfififlfl%%&%flflfifi% 99999999999999999 999$3§$59 99999999999999999 9'99“." °~°9'9'9‘ '9'9'9’9’ 9 9 9 9 9 9 9 9 ’ ”9,99fifi’999’9fiffl’9 99 9 9 . C " 99,3939! 9999999 999 §¢33I§3fl 999999999999999 9999900999009w %%%%%%%%%flfiflfi’v 9999999 99 99 9 99:35 3. ”A 99 9999999 ‘9 o99999999< 9 99999999 9 "fififl”? ' 9999 4 O 0.» 00.9.. 0. 90. 00...... 900.00. {V’VVWV'5 0:...0 \ O 0 ..go 0% O § 999 99999999 9%flflflfl3fi 99999 99 9 99999%& ‘ \\\\~\ \\ . . v 9 9' '9' 9 . v.v.v... ...... 0.0 0. 99999999 9 9999999 9 '%&&fl9° 9 ’ 35 O O O 9 . a; 9’ 99 99 9 O .0 Scale: 1" = 80 m Open Water/Aquatic Bed 7A Emergent Vegetation Scrub-Shrub Vegetation . Brood Observation Point -- -- Breeding Pair Survey X Breeding Bird Survey Route Area Figure 5. Thirteen Mile Road beaver pond site map. 12 Scale: 1" = 450 m Open Water/Aquatic Bed / / Emergent Vegetation Scrub-Shrub Vegetation . Brood Observation Point - - - - Breeding Pair Survey Route X Breeding Bird Survey Area Figure 6. Munuscong Bay coastal impoundments site map. 13 W A 1.2 km zone ofcoastal wetland on the St. Mary's River between 12 and 13 Mile roads (T46N, RZE) (Fig. 1) was monitored. The 31.9 ha area surveyed in this study contains scrub-shrub, emergent, and open water/aquatic bed vegetation zones (Fig. 7). This area provides an example of avian use of a natural, uncontrolled rivetine wetland. Scrub-shrub vegetation is dominated by willow, speckled alder, and Spiraea spp. Wet meadow zones contain grasses (Calamagrostis canadensis, Phalaris arundinacea), woolgrass, and sedges (Carex vesicaria). The flooded emergent zone is dominated by hard-stem bulrush. cattail, horsetail (Equisetum spp.), and Juncus canadensis. N Scale: 1" = 195 In Open Water/Aquatic Bed 'l/A Emergent Vegetation @ Scrub-Shrub Vegetation . Brood Observation Point ----- Breeding Pair Survey Route x Breeding Bird Survey Area Figure 7. St. Mary's River coastal wetland site map. METHODS E . l D . Study sites were selected to provide a contrast of avian use of open water/aquatic bed, emergent, and scrub-shrub zones of vegetation between constructed and nearby established wetlands. Surveys for waterfowl breeding pairs and broods in the open water/aquatic bed zone, and breeding bird plot surveys and nest searches of emergent (wet meadow, flooded emergent) and scrub-shrub plant zones were conducted on all study sites. Wet meadow zones were dominated by sedges, grasses, and woolgrass. Flooded emergent zones were dominated by cattail and bulrush species. Willow and speckled alder were the most common shrub species. Weekly waterfowl breeding pair and three yearly brood surveys were conducted at all study sites. Breeding bird plot surveys and nest searches were conducted three times during each field season at each site. HahitatZones We developed cover maps for each established study site based on on-site measurements, National Wetland Inventory (NW1) maps, and aerial photographs. On-site measurements were made from a series of systematic transects using a meter tape and/or ‘ rangefinder. Cover maps of constructed sites were based on site plans and on-site measurements. Cover maps were then digitized using the ARC-INFO geographic 15 16 information system (GIS). The area and proportion (%) of scrub-shrub, emergent (wet meadow and flooded emergent), and open water/aquatic bed were estimated using ARC- INFO. Due to their small size (8.6% of total area), the small potholes (Fig. 4) were not used in estimating the proportions of emergent and open water/aquatic bed cover at the PMP site. To characterize the potential upland nesting habitat for waterfowl, the amount of grassland and forest found within 500 m of each wetland was estimated from aerial . photos using a dot grid. 1!! E I E l E . 5 Weekly waterfowl breeding pair surveys were conducted at each site from early- May through early-June. Surveys were conducted by walking along set observation routes, and counts were varied so that some were done in the morning, afiernoon, and evening. The number of pairs and indicated pairs (lone males and groups of males of five or less) (Hammond 1969, Leschisin et al. 1992, Delphey and Dinsmore 1993) using Open water/aquatic bed zones were counted. All birds observed within 100 m of the survey route were counted by species and vegetation zone being used. Wm Brood surveys were conducted at each site in mid-June, mid-July, and early- August. Observations of open water/aquatic bed zones were made from fixed observation points, a method similar to counts made by Beard (1964). Surveys lasted for four hours beginning at sunrise. Observation points were varied in 1994. In 1994 we determined that brood observations at the PMP site were not possible to conduct from one l7 observation point. A sample of the potholes were observed from a tree stand, but this proved impractical as growing vegetation impaired vision. Surveys were modified by walking to each wetland and observing for 5 to 15 minutes. In 1995 a standardized method for brood surveys was developed. We conducted three brood surveys at each site: early (mid June-wa July), mid (early-mid July), and late (late July-early August) season. Each survey began at sunrise and lasted for four hours. Two observers surveyed broods from two separate observation points at each site (Figs. 2, 3, and 5-7). Species, number of ducklings, age (Class I, II, and III) (Gollop and Marshall 1954), and time of observation was recorded for each brood observed within a specific area. Class I is the downy stage, which covers a period of about three weeks, fi'om hatching until body feathers begin to appear. Class II stage covers weeks four through six, and is the period during which body feathers replace down plumage. Class III is the stage during which the young are fully feathered by juvenile plumage, prior to flight (Bellrose 1980), and this stage usually lasts about ten days (Gollop and Marshall 1954). At the PMP site we divided the 18 potholes into four groups of three to five (Fig. 4). Two consecutive mornings were used to survey the four pothole groups (two groups per morning, one observer per group ). Ten minutes were spent at each pothole, and the observer rotated between all potholes within a group for a four hour survey period. When compiling the results we only counted the number of Class II broods spaced approximately two weeks apart. This was done to avoid recounting young broods which stayed within an area for more than one survey. 18 E . S [1! . Z Breeding birds using emergent (wet meadow and flooded emergent) and scrub- shrub plant zones were surveyed early (mid-late May), mid (early-mid June), and late season (late June-early July). Surveys were conducted between sunrise and 1000 hours. One or two plots (0.1 ha) were placed in each zone of vegetation at each site, resulting in three to seven plots total per site, depending on the number of zones present and the size of the study site. After arrival at the survey plot five minutes were waited before beginning the survey, so that normal bird activity could resume. All birds observed and/or heard within an 18 m radius were recorded during eight-minute observation periods (Delphey and Dinsmore 1993; Brown and Dinsmore 1986). Taped calls played during the final four minutes of the survey period, were used to elicit responses fiom secretive birds such as the least bittern (Ixobrychus exilis), american bittern (Botaurus Ientiginosus), virginia rail (Rallus Iimicola), yellow rail (Cotumicops noveboracensis), and sora (Porzana carolina) (Brown and Dinsmore 1986, Delphey and Dinsmore 1993). The number of responses and the estimated distance of the response from the center of the plot was recorded. Birds observed outside of the survey plot, but within the vegetation zone being sampled, were recorded to aid in the identification of the breeding bird species found in each zone. After the count period, the area within a 13 m radius plot (0.05 ha) around the same observation point was searched for nests (Brown and Dinsmore 1986, Delphey and Dinsmore 1993). A species was classified as breeding if we observed an active nest, flightless young, or adults during 2 of 3 visits (Brown and Dinsmore 1986). On two occasions, I also considered a species as breeding when nests with identifiable eggshells were found. 19 Analysis The Statistical Analysis Software (SAS) system was used. The general linear model (GLM) procedure was used to conduct an unbalanced analysis of variance (ANOVA) on waterfowl breeding pair use of open water/aquatic bed plant zones, using the following multiple factor model: mar-p+cz,+[1(,,,.+6.,+6.+cz6,.+42:6“[35(,,,,+e,,,l Where Ya: = natural log (pair density); n = true population mean; a, = wetland type (constructed or established); Bar = study site (nested within wetland type: {AW, LMP, PMP} and {BP, MBW, SMCW}); 6, = week of survey; 6. = year of survey; (16,, a6.” and 95m are interactions of these factors; and E“, = the error associated with the model. Preliminary tests indicated that there were not significant [36m or 66. interactions, so they were eliminated from the final model. Pearson product-moment correlation coefficients were used aid in determining the physical factors associated with the breeding pair use of the wetlands. The following physical factors were used in the correlation analysis: surface area; ha of scrub-shrub; ha of emergents; ha of open water; emergentzopen water ratio; edge lengthzsurface area ratio; ha of surrounding grassland (within 500 m); and ha of surrounding forest (within 500 m). Waterfowl brood use of open water/aquatic bed plant zones were analyzed using the GLM procedure within the SAS system to conduct an unbalanced ANOVA. The following model was used: 20 Yes =|1+ar+pa§+6k+efik Where Y“ = natural log (Class II duckling density); p = true population mean; ati = wetland type; Bari = study site (nested within wetland type: {AW, LMP, PMP} and {BP, MBW, SMCW}); 6., = survey period (early, mid, or late season), and ea = the error associated with the model. The coeficient of similarity (Oosting 1956) was calculated to compare the avian use of the scrub-shrub and emergent wetland plant zones between constructed and established wetland types. The equation used was 2w / a + b, where w = the sum of the lowest of each pair of percentages for species occurring in each wetland type, a = the sum of percentages in constructed wetlands, and b = the sum of percentages in established wetlands. A value greater than 0.5 indicates that bird use of the vegetation zone is similar between wetland types. RESULTS Baum Constructed wetlands had a greater proportion of open water/aquatic bed habitat zones than established wetland sites (Table l). Constructed wetlands had less total emergent cover (21.0 - 32.6%) when compared to established wetlands (30.3 - 70.0%). Although small amounts (0.9 - 3.8 ha) of wet meadow had developed at constructed wetlands, 135.8 and 7.1 ha of wet meadow were present on two established sites (MBW and SMCW). More flooded emergent (cattail/bulrush spp.) cover was present on established (3.7 - 76.1 ha) compared to constructed sites (0.1 - 0.3 ha). Two constructed wetlands had zones of scrub-shrub cover present before and not destroyed during construction (Table 1). Emergentzopen water ratios ranged from 0.30 to 0.48 at constructed wetlands, and 0.59 to 6.97 at established. Ratios of edge lengthzsurface area (Km/ha) tended to be higher at constructed (0.22 - 0.47) compared to established wetlands (0.07 - 0.23). The constructed wetlands had a greater proportion of grassland surrounding them (61.3 - 87.2%), whereas the established wetlands had more adjacent forest (38.7 - 92.9 %). 21 Table 1. Comparison of size (ha) and proportion (%) of habitat zones distributed on constructed and established wetland study sites. Constructed Established Habitat Zone AW LMP PMP BP MBW SMCW Wetland Habitat Scrub-shrub 0.4 1.6 -- 2.1 24.4 6.3 (9.7) (8.5) (17.2) (7.0) (19.8) Emergent Wet Meadow 0.9 3.8 2.3 --- 135.8 7.1 (22.0) (19.6) (28.8) (38.9) (22.3) Flooded 0.1 0.3 0.3 3.7 76.1 15.2 Emergent (2.4) (1.4) (3.8) (30.3) (21.8) (47.8) Total 1.0 4.1 2.6 3.7 211.9 22.3 (24.4) (21.0) (32.6) (30.3) (60.7) (70.0) Open Water/ 2.7 13.7 5.4 6.3 112.6 3.2 Aquatic Bed (65.9) (70.5) (67.4) (52.5) (32.3) (10.2) Total Area 4.1 19.4 8.0 12.1 348.9 31.9 Emergent: Open Water Ratio 0.37 0.30 0.48 0.59 1.88 6.97 Edge length: Surface Area Ratio (Km/ha) 0.35 0.22 0.47 0.23 0.07 0.11 Upland Habitat Grassland 75.5 132.1 199.5 84.6 8.9 9.1 (61.3) (69.1) (87.2) (55.3) (2.8) (9.9) Forest 31.9 57.4 19.3 59.3 295.8 71.5 (25.9) (30.0) (8.4) (38.7) (92.9) (77.7) 23 1M EWIE 1° 2.5 Waterfowl breeding pair densities were two to five times greater at constructed compared to established wetlands (P = 0.0001) (Table 2). Breeding pair densities were difi‘erent between years (P = 0.014) and weeks (P = 0.0001). Mean breeding pair densities ranged from 0.9 to 2.2 pairs/ha on constructed sites, and 0.1 to 0.7 pairs/ha on established sites (Table 3). Pair densities varied by study site (P = 0.0001). There were three significant interactions; «16., = treatment-week (P = 0.0001), (16. = treatment-year (P = 0.0001), and [36m = site-year (P = 0.0009). Table 2. Estimated waterfowl breeding pair densities (i 1 SE) by week and wetland type for 1994 and 1995. Wetland Type Week Constructed Established 1 1.61011 0,710.11 2 1.5 10.13 0.4 :013 3 1.11014 0.5 ions 4 1.2 i 0.12 0.3 t 0.12 5 1.7 i 0.15 0.2 t 0.14 Mean 1.4 #106 0.4 10.06 24 Table 3. Estimated waterfowl breeding pair densities (i 1 SE) by wetland type and year. Wetland Type Constructed Sites Established Sites AW LMP PMP BP MBW SMCW 1994 1.3 :0.22 0.61014 2,110.14 0.8 :015 0110.15 0.7 i018 1995 1.1 10.12 1210.12 2.410.12 0,710.12 0,110.12 o.2:o.12 Mean 12:11.13 (3:009 2.2;009 0.7+0.10 0.1 +0.09 0.5 +0.11 Mallard (Anas plaryrhynchos) was the most common waterfowl species on Chippewa County wetlands (Table 4). Mallard, blue-winged teal (A. discors), gadwall (A. strepera), american wigeon (A. americana), and green-winged teal (A. crecca) were present in higher densities on constructed wetland types (P = 0.0001). Mallard breeding pair densities on established wetland types were similar to those of ring-necked duck (Aythya collaris), blue-winged teal, Canada goose (Branra canadensis), and american wigeon. Ring-necked duck pair densities were significantly greater on established wetland types (P = 0.0001). There was no significant difference in breeding pair densities of Canada geese between wetland types (P = 0.50). 25 Table 4. Estimated waterfowl breeding pair densities (x 1- SE) by species and wetland type. Wetland Type Species Constructed Established Canada goose 0.05 i- 0.011 0.06 i 0.010 American wigeon 0.16 i 0.018' 0.02 i 0.017 Gadwall 0.21 i 0.017' 0.01 i 0.017 Green-winged teal 0.13 1 0.016' -- Mallard 0.41 1 0.030' 0.13 1» 0.029 Blue-winged teal 0.34 i 0.031' 0.07 1- 0.030 Ring-necked duck 0.01 i- 0.014 0.09 : 0.013‘I Common goldeneye 0.06 3; 0.009' -- ‘ Significantly greater (P = 0.0001). Dabbling duck pair densities were negatively correlated with surface area, scrub-shrub cover, emergent cover, and amount of adjacent forest cover (Table 5). Positive correlations with edge lengthzsurface area ratio and area of adjacent upland grassland cover were observed. American wigeon, mallard, and blue-winged teal exhibited similar habitat associations, except mallard was negatively correlated with area of open water. 26 Table 5. Significant (P<0.05) correlations of mean waterfowl breeding pair densities with several physical features of the study sites (N = 6). Breeding Pair Densities Physical Parameters Dabblers AMWI MALL BW'I'E Wetland Habitat Surface Area -0.925‘ -0.876 -0.944‘ «0964' Scrub-Shrub -0.942' -0.927' -0.875 -0.932" Emergent -0.933' -0.891 -0.928' -0.977" Open Water -- .0344 Emergent:0pen Water - --- -- -- Edge Length:Surface 0.950‘ 0.919‘ 0.895 0969' Area Ratio Upland Habitat , Grassland 0.866 0.844 --- 0.887 Forest -0.973' -0.947' - 0.966' -0.962‘I ' P < 0.01 W More Class II ducklings were observed on constructed compared to established wetland types (P = 0.008). Duckling densities were similar between sites within treatment (P = 0.61) and between survey periods (P = 0.73) (Table 6). Breeding pair and Class II duckling densities were positively correlated (r = 0.75, P = 0.086). Although mallard (P = 0.05) and blue-winged teal (P = 0.02) had significantly greater densities of Class II ducklings on constructed compared to established wetlands, other species were not present in great enough densities to provide a contrast (Table 7). Table 6. Estimated density (a? 1 SE) of Class II ducklings 27 by period, site, and wetland type in 1995. Constructed Established AW IMP PMP BP MBW SMCW Early 1.5 2.6 0.8 0.5 0.5 2.5 Mid 5.6 3.9 4.7 0.5 1.9 0.2 Late 3.0 2.3 3.9 0.9 1.2 0 Combined 3.3 10.81 2.91-0.81 3.110.81 0.610.81 1.210.81 0.910.81 Treatment Combined 3.1 10.41 0.9;; 0.41 Table 7. Estimated densities (i 1 SE) of Class II ducklings by species and wetland type in 1995. Species Constructed Established Canada goose 0.30 1 0.18 0.20 1 0.18 Wood duck 0.05 1 0.04 0 Gadwall 0.40 1 0.17 0.02 1 0.17 American wigeon 0.70 1 0.21 0.20 1 0.21 Mallard 0.90 1 0.24‘ 0.24 1 0.24 Blue-winged teal 0.60 1 0.13" 0.09 1 0.13 Ring-necked duck 0 0.10 1 0.06 Hooded mgginw 0.10 1 0.14 0 ' Significantly greater (P = 0.05). " Significantly greater (P = 0.02). 28 ! . S [If . Z Nineteen avian species (13 at constructed and 16 at established) were considered to be breeding during the two years of surveys (Table 8). Ten species were found at both constructed and established wetlands. Three species were unique to constructed wetlands, and six were unique to established wetlands (Table 8). At the constructed wetlands no species were observed breeding exclusively in flooded emergent vegetation. Four species were present only in wet meadow, and seven species exclusively in scrub-shrub. Red- winged blackbirds bred in scrub-shrub, wet meadow, and flooded emergent zones at constructed sites. Seven species were present in flooded emergent vegetation, one species in wet meadow, and four species in scrub-shrub vegetation at established sites. The american bittern and Virginia rail were present in wet meadow and flooded emergent zones at established sites. Swamp sparrows (Melospiza georgr‘ana) and red-winged blackbirds used the scrub-shrub, wet meadow, and flooded emergent zones at established sites. The eastern kingbird (T yrannus tyramms), yellow warbler (Dendror‘ca petechia), common yellowthroat (Georhljpr‘s rrr'chas), and swamp sparrow used scrub-shrub habitat at both constructed and established wetlands. The song sparrow (Melospiza melodia) was observed in the scrub-shrub zone at both wetland types, however, it was not seen often enough to be considered a breeder at established sites. The american robin (T urdus migratorr'us) and clay-colored sparrow (Spizella pallida) were found only in scrub-shrub zones at constructed sites, while the gray catbird (Dumerella carolinensis) was present only in scrub-shrub at the established BP site. 29 x 253.35 ex ex :38 5% ex =23 552 x eaten—8m x x 38.5. pagan X X ES 385 x x x x x x nee g8 ex 055 5.65m x «000 50g N X ex ex 8cm x ex ex k x x :2 same, x on x x 685 fig x x x e a. e a 32» 33.3: 8 m... 23 8 E 23 mm m... E 23 mm 23 mm 23 .383 302% >52 mm aim m2; 3< 8383 88:5an Afiombfim 3683 H mm 23 .3888 63 n 23.2w: .l. mmv Son 835%? v5 .37. 33m .25 3863 .3 8a: {ease 33 e8 32 8 e88 3.685 B 9 gene 88% 83. .m ease 30 _ 603950 .382 e dag—ma: fl 3580 wagon 05:? Son :5 .653»? veg—So 95. an antenna ob? was?» mac—Ewan e .888 05:28 he gouge. com . ex .x x .x .x .x x x .x x 39.83 Esteem N N ex N ex N N “on 395% g X 395% mnom .x Eis$so x .x x x 185%.? 8:58 x x x ex ex 8.33 .83, mm E 23 mm m... 23 mm w... E .23 mm 23 mm 23 38am 302m >52 mm m2m m2..— 3< con—mag 363800 .9958 w 038. 3 1 Although the pied-billed grebe (Podilymbus podiceps) was not observed on any of the survey plots, it was ofien seen/heard in flooded emergent stands at the established wetlands. It was considered a breeder at two constructed sites (AW and LMP) as well, due to the observance of adults and flightless young. Species richness was greater (seven vs. zero species) at established wetlands for the flooded emergent zone compared to constnlcted wetlands. Wet meadow zones were exclusively used by more species at constructed compared to established sites (four vs. one species). Sedge wren (Cisrothorus platensis) was the only species exclusively using wet meadow zones of established wetlands, and was not present at constructed sites. All of the species using the wet meadow zone at constructed wetlands were present in flooded emergent zones, or both wet meadow and flooded emergent zones at established wetlands (Table 8). The american coot (F ulr'ca americana), sandhill crane (Grus canadensr’s), black tern (Chlidonias niger), and marsh wren (Cistothorus palustrr's) used flooded emergent stands at established sites, but were absent fi'om constructed wetlands. Thirty- one nests were located on 26 (13 %) of 198 plots searched. Sixteen of the nests found were those of red-winged blackbirds (Agelaius phoeniceus). Densities of breeding birds per ha were similar between constructed and established wetlands for the scrub-shrub, wet meadow, and flooded emergent zones (Table 9). Densities of breeding birds were greater for the cattail flooded emergent zone at established wetlands. No bulrush stands large enough to sample had developed at constructed sites. Coeficients of similarity indicated that species use of wet meadow zones was not similar, while bird use of the scrub-shrub and flooded emergent zones was similar on constructed and established wetlands (Table 9). 32 Table 9. Density of birds per hectare (x 1 SE) (number of plots sampled in parantheses) and coeflicient similarity indices by vegetation zone. Density Vegetation Zone Constructed Established Similarity Index Scrub-shrub 35 1 5 (24) 36 1 3 (50) ' 0.605 Emergent Wet Meadow 15 1 6 (11) 15 1 2 (49) 0.191 Flooded Emergent 27 1 6 (7) 34 1 3 (57) 0.640 Cattail 27 1 6 (7) 39 1 3 (37) 0.633 Bulrush - 27 + 5 (20) - L DISCUSSION Long-term ground studies of mallards indicate that breeding pair populations fluctuate yearly between some high and low values (Dzubin 1969). Since the open water/aquatic bed zones on study sites resulted from activities with different management objectives, difl‘erences in pair densities between sites was expected. The wetland type- year and site-year interactions appear to stem fiom the changes in pair densities from 1994 to 1995. Densities tended to increase or stabilize at constructed wetlands over this period, while they tended to decrease at established wetland sites. Since the mitigated wetland sites were designed to appeal to proximal cues used by dabbling duck pairs (Kaminski and Prince 1981a, 1981b), a difl‘erential response to the habitat could be expected. A constant density was observed at constructed wetlands over the five weeks surveyed, while density decreased at established wetlands. Declining densities of dabbling duck breeding pairs over time has been reported by others (Jackson et al. 1985, Humburg et al. 1978). The importance of open water/aquatic bed zones on constructed wetlands to dabbling duck breeding pairs has been noted by Kaminski and Prince (198 1 a,b), Leschisin et a1. (1992), Ruwaldt et al. (1979), and Lokemoen (1973). This study demonstrates that constructed wetlands in early plant successional stages result in habitat with higher densities of dabbling duck breeding pairs compared to established wetland communities in. the eastern Upper Peninsula of Michigan. Wigeon, gadwall, green-winged teal, mallard, 33 34 and blue-winged teal pairs all responded to the newly created wetland habitats. Similar densities of wigeon, gadwall, green-winged teal, and mallard were found on constructed wetlands in northwestern Minnesota (Leschisin et al. 1992). Ring-necked duck (Ayrhya collaris) was the only waterfowl species with higher breeding pair densities on established wetlands. Nesting ring-necked ducks favor marshes at least partially surrounded by wooded vegetation (Bellrose 1980). All of the established wetland types were surrounded by scrub-shrub vegetation and upland forests. The waterfowl breeding pair densities estimated for the constructed wetlands in this study were intermediate between the relatively low densities of the established wetlands and the high densities found by other workers in the prairie potholes (Kantrud and Stewart 1977) and large prairie marshes (Karninski and Prince 1981a, b). Waterfowl breeding pairs occupy open water/aquatic bed zones of wetlands based on proximate cues provided by the open water and emergent habitat configuration (Karninski and Prince 1981a, b). The findings of Leschisin et al. (1992) suggested that dabbling duck breeding pairs select constructed wetlands primarily on the basis of proximate cues. Kaminski and Prince (1984) surveyed quarter sections by helicopter for waterfowl breeding pairs, and found that even though vegetation-water interspersion was not the strongest habitat correlate with dabbling duck use of prairie marshes, its frequent individual occurrence coupled with that of shoreline development suggested that emergent vegetation-water interspersion does influence dabbler pair densities and species richness. The constructed wetland types in this study were designed to provide an abundance of shallow water and maximum shoreline length, thus simulating the 50:50 open water to emergent, or hemi-marsh (Weller and Spatcher 1965, Weller and Fredrickson 1974) 35 configuration. Edge lengthzsurface area ratio was positively correlated with breeding pair densities, whereas emergentzopen water ratio was not correlated with pair densities. This is consistent with the small quantity of emergent vegetation present on the constructed sites. The constructed wetlands were created within substantial areas of grassland vegetation, and as Kaminski and Prince (1981a, b) demonstrated, the interspersion of water and grassland is structurally the same as would be found in a marsh. The shallow flooding of surface drains at constructed sites increased shoreline length and interspersion of water and vegetation. The PMP had the greatest amount of interspersion and the highest mean density of breeding pairs. The LMP incorporated a high degree of interspersion and loafing logs into the design. Evrard (1975) noted the highly significant correlation between waterfowl use of constructed ponds and the presence of loafing structures. The constructed wetlands were designed to attract breeding dabbling ducks and on average, had smaller surface areas, higher edge lengthzsurface area ratios, and more adjacent grassland nesting cover compared to established wetlands (Table 1). All of these factors were positively correlated (P < 0.05) with breeding pair densities. Several workers have found higher densities of breeding pairs on smaller wetlands (Evans and Black 1956, Lokemoen 1973, Leschisin et al. 1992). Leschisin et al. (1992) found greater pair use of wetlands with longer shoreline lengths for all species except the gadwall. Lokemoen (1973) found that pairs seemed to prefer stock ponds with open shorelines, and that all species combined were significantly associated with grassy shorelines. Breeding pair densities in this study were negatively correlated with the amount of emergent and scrub- shrub vegetation. Evans and Black (1956) found that potholes with excessively dense 36 vegetation were used very little by breeding pairs, and that those with sparse cover, or no vegetation at all, were clearly preferred. The higher densities of Class II ducklings on constructed compared to established wetland types are consistent with the increased density of breeding pairs and the presence of grassland nesting cover adjacent to all of the constructed wetlands. Ball et al. (1975) found a significant linear correlation between distance of overland travel completed prior to two weeks of age and number of surviving ducklings in broods. Rotella and Ratti (1992) found that broods that moved farther had lower duckling survival. “0th increased duckling survival due to close proximity of nesting cover, a greater density of Class II ducklings would be expected at the constructed wetlands. Several workers have noted that brood density in constructed wetlands is primarily a function of the physical features (Lokemoen 1973, Mack and Flake 1980, Hudson 1983, Belanger and Couture 1988). Belanger and Couture (1988) found that size, shoreline irregularity, and proportion of emergent coverage were the most important features in determining brood densities. Man-made ponds greater than 0.5 ha in size and with more irregular shorelines were used most by broods (Belanger and Couture 1988). Mack and Flake (1980) found that shoreline length was positively correlated with the number of blue-winged teal, mallard, and pintail broods using South Dakota stock ponds, and that it was a better predictor of brood use than pond size. A similar relationship between shoreline length and brood use was found on beaver ponds in Ontario (Patterson 1976). Average edge lengthzsurface area ratio was higher at constructed wetlands, which could have made them more attractive to broods. Mack and Flake (1980) found that ponds located within idle grassland tended to have a higher occurrence of blue-winged teal broods. The substantial 37 grassland areas surrounding the constructed wetlands in this study are likely making the wetlands more attractive to dabbling duck broods. Waterfowl broods select habitats in which the highly nutritious foods needed for growth are readily available (Sedinger 1992). The presence of greater Class II duckling densities at the constructed wetland sites suggests that adequate food is available. Although avian use of the flooded emergent zone was similar between constructed and established wetlands, more species and a higher density was found in cattail zones at established sites. This similarity appeared to be related to red-winged blackbird abundancy at both the constructed and established flooded emergent zones. Use of the wet meadow zone was not similar between constructed and established wetlands, and more species were found in wet meadow zones at constructed sites. However, all of the species found in constructed wet meadow were found in either the flooded emergent zone, or both the wet meadow and flooded emergent zones at established wetlands (Table 8). The greater number of species found in wet meadow at constructed sites could be attributed to the lack of flooded emergent zones. It appears that the wet meadow zone at constructed wetlands is providing breeding habitat for species that use both flooded emergent and wet meadow habitat. The wet meadow zone should continue to be important to several breeding bird species at the constructed sites until flooded emergent stands become established. Five of the six species unique to the established sites were associated with the flooded emergent or wet meadow zones. These species included american coot, sandhill crane, black tern, sedge wren, and marsh wren. Marsh-nesting species can nest only when the physical environment is suitable, and nest- site selection involves primarily physical features of the habitat which protect the nest, 38 eggs, and young from floods, heat stress, and predators (Burger 1985). The lack of development of flooded emergent cover at constructed wetlands was probably the most important factor limiting some of the marsh-nesting species. The smaller average size of the constructed wetlands could also be limiting species richness. Brown and Dinsmore (1986) found that as wetland area increased, so did species richness, however the rate of increase decreased as marshes became larger. Tyser (1983) found similar results in cattail marshes in Wisconsin. The large sizes of the flooded emergent stands at the MBW and SMCW established sites, was most likely an important factor in attracting more species to these areas. Three of the six species which were unique to established wetlands have been found to be area-dependent (black tern) or possibly area—dependent (american coot, marsh wren), meaning they were more frequently found in larger marshes (Brown and Dinsmore 1986). Wetlands within wetland complexes have also been shown to attract more breeding bird species than isolated marshes (Brown and Dinsmore 1986). Two of the established sites (MBW and SMCW) were situated within large coastal emergent wetland complexes, which could have also contributed to a greater species richness in these areas. The coemcient of similarity index indicates that bird use of the scrub-shrub zone was similar between constructed and established wetland types. This was expected since the scrub-shrub zones at constructed wetlands were established before construction took place. The presence of american robin and clay-colored sparrow in the scrub-shrub zone at constructed wetlands is explained by the presence of grassland cover adjacent to the shrubs at these sites. American robins could use this grassland cover for feeding, and clay-colored sparrows prefer breeding habitats near to water which have a combination of grasses, trees, and shrubs (Knapton 1979, Ehrlich et al. 1988). 39 The scrub-shrub and wet meadow zones were very important to breeding birds at the constructed sites. Eleven of the thirteen species present at constructed wetlands exclusively used either the scrub-shrub or wet meadow zones. It would be advantageous to construct wetlands adjacent to patches of various wetland plant zones to increase avian breeding use of these areas. Overall species richness of the constructed wetlands was limited by the lack of flooded emergent cover. Sumcient time has not passed since construction of these areas for cattail/bulrush stands to fully develop. The data in this study indicate that breeding waterfowl will quickly respond to constructed wetland habitats when proximate cues are considered in the wetland designs. Other avian breeding species respond to these habitats based on the physical features of the vegetation zones present, thus colonization takes longer because of the time involved with vegetation development. The constructed wetlands are already providing breeding habitat for several avian species due to the presence of patches of vegetation zones, and more species would be expected as the zones continue to develop. APPENDIX 40 Table A-1. A list of common and scientific names of breeding birds identified during 1994 and 1995 surveys (AOU 1982). Common Name Scientific Name pied-billed grebe Podib'mbus podiceps american bittern Botaurus Ientr’gr‘nosus Virginia rail Rallus limicola sora Porzana carolina american coot F ulica americana sandhill crane Grus canadensis common snipe Gallinago gallinago black tern Chlidonias niger eastern kingbird T yrannus rymmus sedge wren Cisrothorus platensis marsh wren Cisrorhorus palustrr’s american robin T urdus migratorr’us gray catbird Dumetella carolinensis yellow warbler Dendroica petechr’a common yellowthroat Geothbpis trichas clay-colored sparrow Spizella pallida song sparrow Melospr‘za melodia swamp sparrow Melospiza georgr'ana red-winged blackbird AJgelaius phoeniceus LITERATURE CITED 41 American Ornithological Union. 1982. Thirty-fourth supplement to the American Ornithological Union Check-list of North American birds. Supplement to Auk 99 (3) l6pp. Ball, I.J., D.S. Gilmer, L.M. Cowardin, and J.H. Riechmann. 1975. Survival of wood duck and mallard broods in north-central Minnesota. J. Wildl. Manage. 39(4):776-780. Beard, EB. 1964. Duck brood behavior at, the Seney National Wlldlife Refuge. J. Wildl. Manage. 28(3):492-521. Bellanger, L., and R. Couture. 1988. Use of man-made ponds by dabbling duck broods. J. Wildl. Manage. 52(4):718-723. Bellrose, RC. 1980. Ducks, geese, and swans of North America. Stackpole Books, Harrisburg, PA. 540 pp. Brown, M., and J .J . Dinsmore. 1986. Implications of marsh size and isolation for marsh bird management. J. Wildl. Manage. 50(3):392-397. Burger, J. 1985. Habitat selection in temperate marsh-nesting birds. Pages 253-281 in Habitat selection in birds, M.L. Cody (Ed). Academic Press, Inc., London, Britain. Delphey, P.J., and J .J . Dinsmore. 1993. Breeding bird communities of recently restored and natural prairie potholes. Wetlands 13(3):200-206. Dzubin, A. 1969. Comments on carrying capacity of small ponds for ducks and possible efi‘ects of density on mallard production. Pages 138-160 in Saskatoon Wetlands Seminar. Can. Wildl. Serv. Rep. 6. Ehrlich, P.R., D.S. Dobkin, and D. Wheye. 1988. The birder’s handbook: a field guide to the natural history of North American birds. Simon and Schuster, New York. Evans, CD, and KB. Black. 1956. Duck production on the prairie potholes of South Dakota. US. Fish and Wildlife Service Special Scientific Report, Wildlife No. 32. Washington, DC. Evrard, J .O. 1975. Waterfowl use of dug ponds in northwestern Wlsconsin. Wildl. Soc. Bull. 3(1):13-18. Gollop, J .B., and W.H. Marshall. 1954. A guide for aging duck broods in the field. Miss. Flyway Counc. Tech. Sect. 14 pp. Hammond, MC. 1969. Notes on conducting waterfowl breeding surveys in the north central states. pp.23 8-254 In Saskatoon Wetlands Seminar. Canadian Wildlife 42 Service Report Series 6, Ottawa, Ontario, Canada. Hudson, M. S. 1983. Waterfowl production on three age-classes of stock ponds in Montana. J. Wlldl. Manage. 47(1):]12-117. Humburg, D. D. H. H Prince, and R. A Bishop. 1978. The social organization of a mallard population 1n northern Iowa. J. Wildl. Manage. 42(1): 72-80. Jackson, D.H., H.H. Prince, and RA Bishop. 1985. Influence of drought on mallards breeding in northern Iowa. J. Wildl. Manage. 49(2):442-448. Kaminski, RM, and H.H. Prince. 1981a. Dabbling duck and aquatic macroinvertebrate responses to manipulated wetland habitat. J. Wlldl. Manage. 45(1): 1-15. , and _. 1981b. Dabbling duck activity and foraging responses to aquatic macroinvertebrates. The Auk 98:115-126. , and _. 1984. Dabbling duck-habitat associations during spring in Delta Marsh, Manitoba. J. Wildl. Manage. 48(1):37-50. Kantrud, HA, and RE. Stewart. 1977. Use of natural basin wetlands by breeding waterfowl in North Dakota. J. Wildl. Manage. 41(2):243-253. Knapton, R.W. 1979. Optimal size of territory in the clay-colored sparrow, Spr'zella pallida. Canadian Journal of Zoology. 57: 1358-1370. Leschisin, D.A., G.L. Williams, and MW. Weller. 1992. Factors affecting waterfowl use of constructed wetlands in northwestern Minnesota. Wetlands 12(3): 178-183. Lokemoen, J .T. 1973. Waterfowl production on stock-watering ponds in the northern plains. J. Range Manage. 26(3): 179-184. Mack, G.D., and L.K. Flake. 1980. Habitat relationships of waterfowl broods on South Dakota stock ponds. J. Wildl. Manage. 44(3):695-699. Oosting, HI. 1956. The study of plant communities. W. H. Freeman and Company, San Francisco, California. 440 pp. Patterson, J .H. 1976. The role of environmental heterogeneity in the regulation of duck populations. J. Wildl. Manage. 40(1):22-32. Rotella, J .J ., and J .T. Ratti. 1992. Mallard brood movements and wetland selection in southwestern Manitoba. J. Wlldl. Manage. 56:508-515. Ruwaldt, J .J ., Jr., L.D. Flake, and J.M. Gates. 1979. Waterfowl pair use of natural and 43 man-made wetlands in South Dakota. Sedinger, J .S. 1992. Ecology of prefledged waterfowl. Pages 109-127 in Ecology and management of breeding waterfowl, B.D.J. Batt, et al. (Eds). University of Minnesota Press, Minneapolis, MN ., 635 pp. Soulliere, G]. 1993. Munuscong wildlife area management plan: an ecosystem conservation and user accommodation approach. Michigan Dept. of Natural Resources Wildlife Div. Rep. 3175. 40 pp. Tyser, RW. 1983. Species-area relationships of cattail marsh avifauna. Passenger Pigeon. 45: 125-128. US. Department of Agriculture. 1992. Soil survey of Chippewa, County, Michigan. USDA-SCS, Sault Ste. Marie, MI. Weller, M.W., and LB. Fredrickson. 1974. Avian ecology of a managed glacial marsh. Living Bird 12:269-291. , and CE. Spatcher. 1965. Role of habitat in the distribution and abundance of marsh birds. Iowa Agric. and Home Econ. Exp. Stn. Spec. Rep. 43. 31pp. "‘11111111111111111