The URL can be used to link to this page

Home My WebLink About19991423 Ver 1_Tulula Creek Wetlands Mitigation Site_20070627� MICHAEL F. EASLEY GOVEI�NOR Mr. John F. Sullivan III, P.E. Division Administrator Federal Highway Administration 310 New Bern Avenue, Suite,410 Raleigh, NC 27601 Dear Mr. Sullivan, m 4S d��4� K � � � � '� �� •� puw �� STATE OF NORTH CAROLINA DEPARTMENT OF T'RANSPORTATION June 27, 2007 ___--�_ °�.�_� 1��3 LYNDO TIPPETT SECRETARY Enclosed please find two final reports on the research Ecological Assessment of a Wetlands Mitigation Bank--Phase III: Restoration Efforts (FHWA/NC/2004-09) and Phase IV.� Post-Restoration (FHWA/NG2006-16). The overall objective for the Tulula Wetlands Mitigation Bank was to restore the functional and structural characteristics of a mountain stream and the adjacent alluvial wetlands. Specific objectives of this study included 1) determining the success of stream realignment by evaluating stream geomorphology; 2) evaluating the responses of vegetation and soil microfauna to different hydrologic re�imes and to the disturbance created by resforation; and, 3) evaluating wildlife use of the site in response to changing hydrologic conditions (amphibians) and plant community succession (birds). A meandering channel (8,500 linear feet in length) was constructed across the floodplain. The channel influenced the hydrology of Tulula, with most changes occurring near the stream channel. However,, disturbance of the soil by earth-moving equipment during restoration disrupted the soil profile. Measurements of soil pH and organic carbon indicated that for soils, decomposition rates and high total microarthropod numbers are functions of an older, less disturbed, ecosystem. The researchers posit that soil disturbance decreases the abundance of litter microarthropods, a critical ecosystem factor. Ten ponds were constructed in 1995-1996 to replace natural breeding sites destroyed during golf course construction. Data coilected indicate that these constructed ponds are of higher quality than reference ponds. Amphibians rapidly colonized the constructed ponds, and the number of species utilizing these as breeding sites averaged about 40% higher than that of reference ponds. Overall, the composition of bird species has changed from pre-restoration years; generalist species and species associated with water have become more abundant, while species associated with early-successional habitats have become less abundant. The changes observed in the overall species composition and the decrease in relative abundance of birds are attributed primarily to the loss of early-successional habitats due to flooding by beaver and natural succession. The researchers state that management is needed to control flooding by beaver dams and to maintain early-successional habitats throughout the site. The enclosed CD will auto-launch upon placement in the drive and contains contact information for the researchers and our staff, as well as the final report in .pdf format. Additionally, the enclosed information is on the NCDOT Research web site at: www.ncdot.or�/�research. We appreciate the interest and support of your staff throughout the course of the project.�If you have any questions, please feel free to contact me at (919) 508-1865. Enclosures MAILING ADDRESS: NC DEPARTMENT OFTRANSPORTATION RESEARCH S� ANALYSIS 1549 MAIL SERVICE CENTER Rn�eiGH NC 27699-1549 Sincerely, . D . rinmay "Moy" Biswas, P.E. te Research & Analysis Engineer BISWAS@DOT. STATE. NC. US TELEPHON E: 919-715-2465 FAX: 919-715-0137 WWW.DOT.STATE.NC.US/ RESEARCH ��,�� ,�� . � � �� JlJL � , ... : �'s�i � �.;.1 — E.� ! f tNETLqNDSqNI� 5��„ `�; .. � ` r.. �L�C�t'T36�N.;n;.C: i TRANSPORTATION �BUILDING 1 SOUTH WILMINGTON STREET RALEIGH NC 27601 DISTRIBUTION: AASHTO RAC members State Publications Clearinghouse NTL USDOT Library Dan DeVane (w/o) Bill Rosser (w/o) Debbie Barbour (w/o) Greg Thorpe (w/o) Art McMillan Don Lee Mike Sanderson Phil Harris David Henderson Gordon Cashin Bruce Ellis Kevin Moorhead (UNC-A) Irene Rossell (LJNC-A) C. Reed Rossell (LJNC-A) Barbara Reynolds (LJNC-A) Jim Petranka (UNC-A) Marella Buncick (USFWS) Volpe Center Library U.C. Berkeley Library Northwestern Library Mike Bruff (w/o) Calvin Leggett (w/o) Jan Bryant (w/o) Logan Williams Dennis Herman Randy Griffin Bryon Moore Jamie Lancaster Jason Elliot Rob Hanson Charles Cox Denise Moldenhauer (USFWS) Angie Pennock (USACE) Steve Lund (USACE) Scott McLendon (USACE) Marla Chambers (NCWRC) �. ;, �.�. ., a _.;.___..,<. ,. r � TRB Library N.T.I.S. NCDOT — Public Info. Julie Hunkins (w/o) Joe Geigle (FHWA) Robert Foyle (ITRE) Derek Smith Matthew Green Mark Davis Owen Anderson Wesley Grindstaff Rob Ayers Missy Dickens � Brian Wrenn (NCDENR) . John Dorney (NCDENR) Mac Haupt (NCEEP) James Martin (CTE) Katie McDermott (CTE) � � � � � • � � � � � � � � � � � � � � � � � � � � • • • • • • • • � � • � • • • • � � �� . ; yk ,� � � n' � I NCDOT 8 CTE Joint Project 2003-18 FHWA/NC12004-09 August 2004 ��0� NORTH C,qq `i O '`� � � � . °�9\�, � >�o �P 0 ��OFTRANSQ Ecological Assessment of a Wetlands Mitigation Bank (Phase III: Restorafion Efforts) Kevin Moorhead, Ph.D., Chair and Professor Irene M. Rossell, Ph.D., Associate Professor Barbara (Kitti) Reynolds, Ph.D, Associate Professor University of North Carolina at Asheville Department of Environmental Studies Jim Petranka, Ph.D., Professor University of North Carolina at Asheville Department of Biology � � CTE / NCDOT � Joint Environmental Research Program Research Project No. 2003-18 FHWA/NG2004-09 Final Report Ecological Assessment of a Wetlands Mitigation Bank (Phase III: Restoration Efforts) Prepared By: Kevin K. Moorhead, Irene M. Rossell, Barbara C. Reynolds, C. Reed Rossell, Jr. Department of Environmental Studies University of North Carolina at Asheville Asheville, NC 28804 and James W. Petranka Department of Biology University of North Carolina at Asheville Asheville, NC 28804 August 2004 The contents of this report reflect the views of the author(s), who are responsible for the facts and the accuracy of the information presented herein. This document is disseminated under the sponsorship of the U.S. Department of Transportation, University Transportati�n Centers Program, in the interest of information exchange. The U.S. Government assumes no liability for the contents or use thereof. �a Technical Report Documentation Page 1. Report No. 2. Government Accession No. 3. Recipient's Catalog No. FHWA/NG2004-09 4. Title and Subtitle 5. Report Date Ecological Assessment of a Wetlands Mitigation Bank August 2004 (Phase III: Restoration Efforts) 6. Performin Or anization Code 7. Author(s) 8. Performing Organization Report No. Kevin K. Moorhead, Irene M. Rossell, Barbara C. Reynolds, C. Reed Rossell Jr. and James W. Petranka 9. Performing Organization Name and Address 10. Work Unit No. (TRAIS) Departments of Environmental Studies and Biology University of North Carolina at Asheville I 1. Contract or Grant No. Asheville NC 28804 12. Sponsoring Agency Name and Address 13. Type of Report and Period Covered US Department of Transportation, Final Report Research and Special Programs Administration Jul 2002 — June 2004 400 7t° Street, SW 14. Sponsoring Agency Code Washington, DC 20590-0001 2003-18 Supplementary Notes: • Supported by a grant from the US Department of Transportation and the North Carolina Department of Transportation through the Center for Transportation and the Environment, NC State University. 16. Abstract - The overall objective for the Tulula Wetlands Mitigation Bank has been to restore the functional and structural characteristics of a mountain stream and the adjacent alluvial wedands. Specific restoration objectives of this study included: 1) determining the success of stream realignment by evaluating the geomorphology of a new channel before and after water release, 2) evaluating c6anges in ecosystem structure and fundion associated with plant communiry succession in planted and unplanted portions of the floodplain in response to restored hydrology, and 3) evaluating wildlife use of the site in response to changing hydrologic conditions (amph�bians) and plant community succession (birds). A meandering channel (8,500 linear feet in lengih) was consiructed across the floodplain and water was released into the new channel in 2001 and 2002. Fight random channel segments were used for measurements of stream geomorphology and atter two years of water flow few differences were noted for channel pattern, although changes were observed for cross-sectional areas of riffles and pools. Isolated azeas of bank and bed erosion were noted. The hydrology of Tulula has been influenced by t6e stream restocation, with most notable differences occurring for water-table wells located near the channei. Although the hydrology of Tulula fen was not influenced by stream restoration, the composition of wetland plant communities in the fen was influenced by natural succession. Restoration did influence the composition of some plant communities. For example, restored wetland azeas contained fewer species than unrestored areas or restored dry areas, and the species that dominated the restored wet areas were OBL and FACW plants. In addition, production of both vegetative and reproductive stems of a common rush was influenced by restorntion and hydrologic change. Recendy disturbed areas at Tulula had lower decomposidon rates and fewer litter microarthropods compared to older plant communities. Data collected from 1996-2004 mdicate that constructed ponds are of higher quality than reference ponds based on physiochemical characteristics, seasonal hydroperiod, and use by resident amphbians. Amph�bians rapidly colonized constructed vernal ponds, and the number of species that utilize these as breeding sites averaged about 50% 6igher than that of reference ponds. 'Ihe survivorship and output of wood frog and spotted salamandec juveniles have declined since pond construction, in part due W the accumulation of predators in ponds, the outbreak of a virus pathogen, and premature pond drying associated with drought In 2004, b¢d species richness and relative bird abundarice decreased significantly from 2002 levels. Bird species richness decreased 15%and relative bird abund�ce decreased 52%. Generalist species, such as Song Spazrow and Rufous- Sided Towhee, continued to be the most abundant species, while many Neotropical migrants of conservation concem, including the Golden-winged Wazbler, Chestnut-sided Warbler, Hooded Warbler, and Yellow-breasted C6at, declined substantially. The significant declines in bird species richness and relative bird abundaz�ce are attributed to habitat changes associated with the flooding of a lazge proportion of the floodplain as a result of beaver activity. Management intervention is recommended to control the flooding caused by beaver, and to maintain a variety of early- . successional habitat es throu out the site. 17. Key Words 18. Distribution Statement Wetlands, wedand conservation, mitigation measures, restoration ecology, site surveys, geomorp6ology; hydrology, water table, plant location, amphibians, birds 19. Security Classif. (of this report) 20. Security Classif. (of this page) 21. No. of Pages 22. Price Unclassified Unclassified 88 Form DOT F 1700.7 (8-72) Reproduction of completed page suthorized 3 DISCLAIMER The contents of this report reflect the views of the authors, who are responsible for the facts and the accuracy of the data presented herein. This document is disseminated under the sponsorship of the U.S. Department of Transportation and North Carolina Department of Transportation in the interest of information exchange. This report does not constitute a standard, specification, or regulation. The US Government assumes no liability for the contents or use thereof. ACKNOWLEDGMENTS Support for this project was provided by the U. S: Department of Transportation and the North Carolina Deparhnent of Transportation through the Center for Transportation and the Environment, NC State University. The authors thank Victor Agraz, Robert Warren, Duncan Quinn, and Dr. Dan Pittillo for their contributions to this research. We also thank the numerous undergraduate students of LJNCA for their efforts. TABLE OF CONTENTS LISTOF TABLES .................................................................................. 5 LIST OF FIGURES......... ........ .......................................................... 6 EXECUTNE SUMMARY ............................................................................................. 7 I. INTRODUCTION ................................................................................................. 9 II. RESEARCH METHODS AND RESULTS ........................................................... 10 A. Stream Restoration and Hydrology ................................................... 10 B. Vegetation Responses to Restoration ................................................. 23 C. Decomposition and Soil Microfauna .................�....:............................ 34 D. Amphibian Use of Tulula . ....... ..... .. .. .. .... ... .. ... ... ... ... ... . .. . .. ... ... . . . . .. . . 42 E. Bird Use of Tulula ............................................... .................... 54 III. DISCUSSION .........................................................................................:.............. 60 1V. RECOMMENDATIONS ........................................................................:.............. 62 V. LITERATURE CITED ........................................................................................... 63 APPENDIX A. (Cross sections of riffles and pools in eight stream segments) ............... 67 APPENDIX B. (Pre- and post-restoration water-table data from electronic wells)........... 71 APPENDIX C. (Pre- and post-restoration water-table data from manual wells) .. ........... 80 APPENDIX D. (Amphibian and reptile species of Tulula) .............................:........ 85 APPENDIX G. (Bird Species at Tulula Wetland (1994-2004) ......................... ...... 86 4 � LIST OF TABLES Table 1. Design criteria for the restored Tulula Creek .... .... ..... ... ... .... ... ... . .. .. .. .. .. ......... 12 Table 2. Bankfull width and cross-sectional area of riffles and pools .. ......... ... ... ..... ........ 15 Table 3. Percent change in cross-sectional area of riffles and pools . .................. ........... 17 Table 4. Sinuosity and slope of the water surface over time ........... ......... ........... ........... 17 Table 5. Width/depth ratio and maximum depth of riffles and pools .. ... ...... ... ... .. ........... 18 Table 6. Other physical characteristics of selected meanders in each stream segment. ..:....... 18 Table 7. Erosion of channel banks after two years of water flow ........ ......... .:. ............... 20 Table 8. Taxa and wetland indicator status of plants occurring in in four study areas........... 26 Table 9. Contribution of each wetland indicator status in four study areas at Tulula. ......... .. 28 Table 10. Effects of restoration on vegetative growth and reproduction of Juncus effusus. .... 29 Table 11. Effects of restoration on biomass of plants occurring with Juncus effusus............ 29 Table 12. Effects of hydrology on vegetative growth and reproduction of Juncus effusus.. .... 29 Table 13. Effects of hydrology on biomass of plants occurring with Juncus effuses . .... .. .. .. .. 30 Table 14. Total number of overstory trees of each species in 20, l Ox10-m2 plots ................. 31 Table 15. Importance values for overstory trees in 1Ox10-m2 plots ................................. 31 Table 16. Total number of understory trees of each species in 20, 4x4-m2 plots . ... :. .... .. ..... . 32 Table 17. Importance values for understory trees in 4x4-m2 plots .................................. 33 Table 18. Mean percent cover of each plant type in lxl-m2 quadrats ............................... 33 Table 19. Importance values for plant types in lxl-m2 quadrats .................................... 34 Table 20. Survival of commercial red maple seedlings planted in 1995 .......:.................... 35 Table 21. Microarthropod responses to date and site ................................. ...... ..... 37 Table 22. Relative abundance and migratory status of birds ......................................... 57 Table 23. Means of bird richness, relative bird abundance, and habitat variables ................ 59 @ LIST OF FIGURES Fig. 1. Restored channels sections of Tulula Creek ....... ....... ..... .... ... ... . ... .. .... .. .......... 14 Fig. 2. Approximate locations of stream segments used for channel evaluations ................. 15 Fig. 3. Cumulative pebble counts of seven stream segments ......................... ............. 19 Fig 4. Transects and individual electronic wells used to assess site hydrology .................... 21 Fig. 5. Location of manual wells at Tulula ............. .............................. .......... ........ 22 Fig. 6. The daily water table and monthly averages for electronic well X1 ........................ 24 Fig. 7. Percent litter remaining in litterbags after 17 months in the field ....... ... .. .. ..... ........ 37 Fig. 8. Average number of microarthropods for three collection dates ............................. 39 Fig. 9. Average number of total microarthropods for March, 2003 ................................. 40 Fig: 10. Average percent organic carbon for soil from five plant communities .................. 41 Fig. 11. Average pH for soil from five plant communities ........................................... 42 Fig. 12. Location of standing water habitats within the study site (spring 2004) .................. 44 Fig. 13. Physiochemical characteristics of reference and constructed ponds ....................... 45 Fig. 14. Mean number of species that bred in reference and constructed ponds ................... 47 Fig. 15. Response of female wood frog and spotted salamanders to pond construction......... 49 Fig. 16. Estimates of the percentage of ponds that produced juveniles, and total juvenile recruitment from constructed and reference ponds during 1996-2003 .............................. 50 Fig. 17. Annual variation in the percentage of constructed and reference ponds that either did not fill or that dried before larvae could initiate metamorphosis ........................:................. 51 Fig. 18. Changes in the percentage of reference and constructed ponds in which catastrophic die- offs of larvae occurred from Ranavirus infections ..................................................... 52 Fig. 19. Changes in adult breeding population size on the eastern sector based on annual egg mass counts in all breeding sites ........................................................................... 53 Fig. 20. Yearly changes in the proportion of the ten constructed ponds that contained fish... 53 Fig. 21. Location of bird survey arid habitat plots .......... .... ... ... ... ... . .. . .. . . . . . .. ... .. . .... 56 0 � EXECUTIVE SUMMARY. Our goal is to document the ecological success of the wetlands at the Tulula Wetlands Mitigation Bank (Graham County) in response to restored hydrology, soils, and vegetation. Our data should provide NCDOT an ecological assessment that may be useful far evaluating other wetland restoration projects located throughout the state. The following objectives provide the framework for a comprehensive ecological assessment of the restored wetlands of Tulula: 1) determine the success of stream realignment by evaluating the geomorphology of the new channel before and after water is introduced, 2) evaluate changes in ecosystem structure and function associated with plant community succession in planted and unplanted portions of the floodplain in response to a higher water table and overbank flooding, and 3) evaluate wildlife use of the site in response to changing hydrologic conditions (amphibians) and plant community succession (birds). A primary focus of restoration at Tulula was to improve site hydrology. A meandering channel (8,SOO linear feet in length) was constructed across the floodplain in five separate sections that were connected in fa112001 and summer 2002. Eight random channel segments were used for measurements of stream geomorphology, including sinuosity, cross-sectional areas of riffles and pools, bank slope, slope of the water surface, and overall channel configuration. After two years of water flow, differences were noted in certain aspects of channel morphology, and localized areas of erosion were noted with erosion control pins and through increases in the cross-sectional areas of some riffles and pools. However, the overall configuration of the channel was maintained over the two-year period. The restoration of hydrology at Tulula was evaluated primarily by changes in water-table depth as recorded with a series of electronic and manual wells. Our assumption was that the overall water table of the site would rise after the channel was restored and the drainage ditches were plugged. We found that the hydrology of Tulula was influenced by these restoration efforts, with most changes occurring in water-table wells located near the stream channel. Restoration appeared to have little influence on the hydrology of the fen or of areas located farther from the channel. � Natural succession continues to change the composition of wetland plant communities � across Tulula. In 2003, overstory-sized trees were present in a fen that had been characterized by . an open canopy in 1994, and there was a dramatic increase in the number of understory-sized • trees. The ground layer in this part of the fen also showed an increase in woody species, and a decrease in the importance of plants that require sunlight, such as rushes. Soil disturbance attributed to restoration activities increased the taxonomic richness in dry areas. In wet areas, restoration combined with a high water table led to colonization by almost almost exclusively OBL and FACW species. Both restoration and the higher water table . increased the number and biomass of vegetative stems of Juncus effusus (soft rush), and the higher water table increased the number of reproductive stems of this species. ' � � � � f � Ten ponds were constructed in 1995-1996 to replace natural breeding sites that were • destroyed during golf course construction. Data collected from 1996-2004 indicate that � constructed ponds are of higher quality than reference ponds based on physiochemical characteristics, seasonal hydroperiod, and use by resident amphibians. The reference ponds have � progressively deteriorated between 1996-2002 with respect to seasonal hydroperiod. In 2002 the � majority either did not fill or dried prematurely, resulting in catastrophic mortality of pond • populations. In contrast, the hydroperiod of most constructed ponds appears to be ideal for most vernal pond breeders. Seven of 10 ponds underwent seasonal drying in most years, typically in � late summer or fall after larvae had metamorphosed. Fish have colonized many ponds since � 2002 in association with above normal rainfall, beaver activity, and completion of the final phase of reconstruction. Amphibians rapidly colonized the constructed ponds, and the number of species that utilize these as breeding sites averaged about 50% higher than that of reference ponds. The survivorship and output of juveniles of two focal species (wood frog; spotted salamander) have declined since pond construction, in part due to the accumulation of predators in ponds, the outbreak of a virus pathogen, and premature pond drying associated with drought. Nonctheless, a small percentage of ponds on site have successfully produced juveniles annually, and populations of both species are being maintained at viable levels. Results of breeding bird surveys in 2004 indicated that species richness and relati��e abundance decreased significantly from 20021evels. Species richness decreased 15%, ��•ith 33 species recorded. American Woodcock, Common Grackle, and Eastern Wood-pewee were new species recorded during surveys. Relative bird abundance decreased 52%, with a tota] of 166 observations. Generalist species, such as Song Sparrow and Rufous-Sided Towhee, continued to be the most abundant species breeding at Tulula, but their numbers decreased dramatically from 20021evels. The Red-winged Blackbird also continued to be one of the most abundant species, but its numbers held steady relative to 20021evels. Many Neotropical migrants of conservation concern declined substantially in 2004 including the Golden-winged Warbler, Chesmut-sided Warbler, Hooded Warbler, and Yellow-breasted Chat. The significant declines in bird species richriess and abundance in 2004 are amibuted to habitat changes associated with the flooding of a large proportion of the floodplain as a result of beaver activity. Productivity of the habitat for birds at Tulula has decreased and correlates with an increase in the large amounts of area covered with standing water and dominated by rushes and sedges. Management intervention is needed in order to restore the productivity of the habitat for birds. Management objectives should include taking appropriate actions to control the flooding caused by beaver, and maintaining a variety of early-successional habitat types throughout the site. � � � � � I. INTRODUCTION � Surface transportation projects such as highway construction often impact wetland • resources and cause unavoidable losses of small wetland areas. Increasingly, wetland losses are being mitigated by the creation of "banks" of restored or natural wetlands that are protected from � future disturbance. Mitigation banks allow the consolidation of efforts to mitigate for small � wetland losses, facilitate advanced planning, and enhance the monitoring and evaluation of � mitigation projects (Short 1988). The Tulula Wetland Mitigation Bank was created to offset • impacts of highway projects in western North Carolina, particularly in the Little Tennessee River S basin (1,158,883 ac) located in Macon, Swain, Graham, Jackson, Clay, and Transylvania Counties. The site was ideal for establishing a mitigation bank in the mountains of North � Carolina because of its relatively large size (235 ac) and its need for large-scale restoration. � The Tulula Wetland Mitigation Bank (Tulula) (35°17'N, 83°41'W) is located in Graham � County, NC in the floodplain of Tulula Creek, 7.7 miles west of Topton. The site covers � � approximately 235 ac at an elevation ranging from 2500 to 2800 ft. It is characterized by a • relatively large, level floodplain along Tulula Creek, and is bordered by forested uplands and infrequent seepage communities on adjacent slopes: A complete description of vegetative � communities at Tulula is found in Moorhead et al. (2001a). Tulula was part of the Nantahala � National Forest and owned by the U.S. Forest Service until the mid-1980's, when it was traded to � a group of developers for commercial development of a golf course. During construction of the • golf course, the bed of Tulula Creek was dredged and channelized and several drainage ditches were dug. Spoil from the drainage ditches and from 11 small golf ponds was spread over � portions of the floodplain. A large portion of the floodplain forest was removed during the �� construction of 18 fairways. About 40% of the wetlands were disturbed by drainage and timber . harvest during golf course construction. � Tulula was purchased in 1994 by the North Carolina Department of Transportation . (NCDOT) to develop a wetlands mitigation bank. We have collected information on baseline , ecological conditions (soils, hydrology, flora, and fauna) and have evaluated restoration activities at the site since 1994 (see www:unca.edu/tulula for details and species lists). Assessing the success of wetland restoration projects requires an evaluation of ecosystem structure and function. Long-term success is rarely documented, and failure is common for a variety of reasons. Our goal was to document the ecological success of the wetlands at Tulula in response to restored hydrology, soils, and vegetation. Our data should provide NCDOT an ecological assessment that may be useful for evaluating other wetland restoration projects located throughout the state. The following objectives provide the framework for a comprehensive ecological assessment of the restored wetlands of Tulula: 1) determine the success of stream realignment by evaluating the geomorphology of the new channel before and after water is introduced, 2) following restoration of site hydrology, evaluate changes in ecosystem structure and function associated with plant community succession in the floodplain in response to a higher water table and overbank flooding, and 3) evaluate wildlife use of the site in response to changing hydrologic conditions (amphibians) and plant community succession (birds). io II. RESEARCH METHODS AND RESULTS Ecological conditions at Tulula have been documented for over ten years by UNCA (see www.unca.edu/tulula; North Carolina Department of Transportation 1997, Rossell et al. 1999, Moorhead et al. 2001a, Moorhead et al. 2001b). Ecological success of wetlands restoration at Tulula has been evaluated by comparing the extensive pre-restoration database to the post- restoration data. A. Stream Restoration and Hydrology 1. Stream Restoration A primary focus of restoration at Tulula was to improve site hydrology. A meandering channel (8,SOO linear feet in length) was constructed across the floodplain during the winter of 1999/2000. The design of the new channel was based partially on the physical characteristics of a relic channel found primarily at the lower end of the site. The relic channel was used, when practical, as part of the new meandering channel. The channel was re-constructed in 2001/2002 to correct problems associated with longitudinal grade. Common streambank erosion techniques, such as fiber matting, coir fiber rolls, root wads, and live stakes of willow (Salix spp.) and silky dogwood (Cornus amomum), were installed to improve the short-term stability of the new channel. Four sections of the constructed channel, in the upper and middle portions of the site, were joined together by crossing the dredged channel of Tulula Creek in fall 2001. The fifth section was connected in two stages in May (Section V) and June (Section Va) 2002. The design criteria used to construct the channel are shown in Table 1. Table 1. Design criteria* for the restared Tulula Creek. Parameter Proposed Average Value Range Cross-sectional area 18 15 — 20 ft2 Bankfull Width 8.5 ft 8-10 ft Average Depth 2.2 ft 1.6 — 2.9 ft Maximum Depth 3.6 ft 2.2 — 5.3 ft Width/Depth Ratio 4 3.1— 6.3 Meander Wavelength 70 — 80 ft 60 —100 ft Sinuosity 1.62 1.44-1.93 Arc Length 50 ft 40 — 70 ft Radius of Curvature 15 ft 10 — 25 ft Channel Slope 0.0020 0.0017-0.0022 Rosgen Stream Type** ES *North Carolina Department of Transportation (1997) **Rosgen (1996) 11 Methods A primary objective for restoration efforts at Tulula was to determine the success of stream realignment by evaluating the geomorphology of the new channel before and after water introduction. Eight random channel segments were chosen in the five stream sections that were restored in 2001/2002. Each segment included four to six riffle-pool sequences varying in length from 120 to 180 ft. Each segment began and ended at the top of a riffle and the origin and end were permanently staked with PVC pipe and rebar. These two points served as reference to partially describe the channel geomorphology. A 300-ft measuring tape was secured between the origin pin and the end pin. Beginning at 0 ft(the origin pin), the orthogonal distance from the tape to the left bank, thalweg, and right bank was measured every 6 ft on the 300-ft tape. The data were used to develop overall channel configuration (planview) and to determine sinuosity of channel segments. Data derived from this work included meander.wavelength, arc length, belt width, and the radius of curvature. In each of the eight segments, two riffles and two pools (defined as the middle of a. meander) were chosen to establish permanent cross-sections. Bankfull width was determined and channel cross-sections were determined by taking depth measurements every 8 in along a tape that was stretched from the two bank pins of a riffle or pool at the top of each bank. Bank inclination was determined with a clinometer. The cross-section data were used to calculate cross-sectional area, average depth, maximum depth, and the width/depth ratio. Erosion bank pins were installed at the toe or middle of a channel bank at a few riffle and pool cross-sections. The erosion pins were hammered 21 in into the bank walls with 3 in exposed in the channel. Pebble counts, using a modified Wolman method (Rosgen 1996), were conducted for each of the eight stream segments although consistent methodology and results were only available at year 2 of water flow. Pebble counts are used to determine the particle size distribution of channel materials. • The slope of the water surface was surveyed using standard surveying equipment. A 300- ft tape was placed in the channel along the thalweg, with a start point in the channel by the origin pin. The features of each segment (each pool and riffle) were surveyed at the top of the left and right banks and for the thalweg. The water depth was also noted for the thalweg. The top, middle, and bottom of each riffle were surveyed as well as the middle of a meander. The distance of these features were noted from the 300-ft tape lying in the thalweg of the channel. The permanent riffle or pool cross-section pins were also surveyed. Benchmarks for each segment were chosen by using established NCDOT surveying points or by placing a nail in a nearby tree (benchmarks were established throughout the Tulula floodplain by NCDOT during channel construction). Overall slope of the water surface was calculated by dividing the difference in water surface elevation from the origin to the end of the segment (both points representing the top of a riffle) by the total stream distance. The planview was evaluated before water release and after one year of water flow. The methods used to deterinine the planview (as described above) are destructive of floodplain vegetarion and annual evaluations are not warranted. The other geomorphic characteristics were evaluated before water release and after one and two years of water flow. The goal was to evaluate the geomorphology of the channel annually after the date of water release. Results and Discussion 12 The restored channel was constructed as five separate sections (Fig. 1). Eight random channel segments were chosen in the five sections (Fig. 2) to evaluate stream geomorphology over time. Water release began in Section 1 of th� restored channel in September 2001. We placed two segments for channel evaluation in Section 1, one each in Sections 2 and 3, two in Section 4, and one each in Sections 5 and Sa. The initial bankfull width and changes in the cross- sectional areas of riffles and pools of the channel segments are listed in Table 1. There was essentially no change in the bankfull widths after two years of water flow and therefore, only the initial bankfull widths are reported in Table 1. Restored Tulula Creek . li �-« . � r---�'�.r{ Sedion V Section Va � Sedion I - September, 2001 Sedion I I and I II. - Odober, 2001 Date of water release: ��on IV - November, 2001 Sedion V - May, 2002 Sedion Va - July 2002 Fig. 1. Restored channel sections of Tulula Creek. Fig. 2. Approximate locations of stream segments used for channel evaluations. 13 As anticipated, riffles typically had lower cross-sectional areas and shorter bankfull widths compared to pools (Table 2). Although bankfull widths did not change after two years of water flow, changes in cross-sectional areas were noted for both riffles and pools. The cross- sectional areas of riffles increased after two years of water flow. Nine of 16. riffles had > 10 % increase in cross-sectional area after two years of water flow (Table 3). Ten of 16 pools increased in cross-sectional area but six other pools decreased in cross-sectional area, typically at locations where point bars were forming. The cross section of a stream changes much more rapidly and frequently in meander bends and, therefore, there is more variability in pool cross sections than in riffle cross sections (FISRWG, 1998). A visual representation of riffle and pool cross sections is shown in Appendix A. Changes in cross-sectional area are often used as an indicator of stream channel stability. Increases in cross-sectional area represent areas of stream degradation (sediment erosion) while increases indicate aggradation (sediment deposition) of a stream channel. Changes at Tulula probably represent adjustments of a constructed channel to various flow regimes over the past two years. . Table 2. Bankfull width (ft) and cross-sectional area (ft2) and of riffles and pools in eight stream segments. Segment I Riffle 1 Pool 1 Riffle 2 Poo12 Segment IA Riffle 1 Pool 1 Riffle 2 Pool 2 Segment II Riffle 1 Pool 1 Riffle 2 Poo12 Segment III Riffle 1 Pool 1 Riffle 2 Poo12 Bank Full ----------Cross-Sectional Area------------- Width Initial One Year Two Years 13.58 15.42 11.81 15.42 10.50 10.27 12:96 12.57 16.34 16.01 12.80 14.31 13.29 18.87 16.90 17.88 20.10 33.27 14.59 26.71 13.84 19.07 19.50 18.94 19.67 30.26 13.69 20.29 18.55 31.27 23.89 26.88 18.80 27.93 13.99 27.57 14.42 18.96 19.86 17.97 20.34 25.03 14.81 2335 18.25 30.82 25.44 21.28 21.93 24.21 15.69 28.92 1636 19.76 22.12 18.40 21.93 27.80 16.36 24.7 20.06 32.99 24.70 22.49 Segment N . Riffle 1 Pooll , Riffle 2 Pool 2 Segment Na Riffle 1 Pool 1 Riffle 2 Pool 1 Segment V Riffle 1 Pool 1 Riffle 2 Poo12 Segment Va Riffle 1 Pool 1 Riffle 2 Poo12 Average Riffle 1 Pool 1 Riffle 2 Pool 2 14 12.53 14.08 12.73 14.57 12.40 13.58 15.13 13.52 14.76 16.24 13.78 16.33 16.14 17.15 21.35 24.70 18.91 23.34 26.38 27.33 12.22 14.66 22.29 19.50 19.22 21.89 19.71 19.17 17.50 23.33 22.57 27.59 15.39 21.15 21.50 21.74 17.13 20.51 19.58 24.08 27.03 24.72 15.45 16.70 16.66 28.32 32.97 33.33 9.68 15.24 ---- 11.65 18.14 ---- 15.26 18.57 ---- 10.04 16.68 ---- 12.89 16.61 17.73 14.53 24.97 24.85 13.91 17.98 19.43 1431 22.99 24.23 16.98 19.60 19.43 18.12 18.72 24.20 19.88 24.41 Table 3. Percent change in cross-sectional area of riffles and pools after two years of water flow. Numbers in brackets represent a decrease in cross-sectional area. Segment Riffle 1 Pool 1 Riffle 2 Poo12 I . 9.1 (27.2) 7.5 8.3 Ia 18.2 3.6 13.5 (2.8) II 11.5 (8.1) 19.5 22.0 III 8.1 � 5.5 3.4 (17.1) IV 8.3 9.3 19.3 4.6 Na 25.9 (5.1) 11.9 10.3 V 14.3 2.7 7.8 17.7 Va 11.4 8.1., ., 4.6 8.6 Average 13.4 (1.4) 10.9 6.4 15 The average sinuosity of the restored channel was 1.32 (Table 4), compared to the design sinuosity of 1.62. The slope of the water surface varies for the stream segments and has decreased over two years in four of seven stream segments (Table 4). Table 4. Sinuosity and slope of the water surface over time. Segment I Ia II � N Na V Va Average Sinuosity Initial slope At 1 year 1.23 1.22 1:26 1.43� 1.29 1.22 1.32 1:58 1.32 0.0030 0.0024 0.0022 0.0028 0.0044 0.0022 0.0024 0.0036 0.0010 0.0019 0.0026 0.0047 0.0025 0.0014 At 2 years 0.0006 0.0018 0.0016 0.0059 beaver 0.0018 0.0028 0.0025 0.0020 The width/depth (W/D) ratio of riffles was slightly higher than for pools and decreased after two years of water flow (Table 5). The decrease in W/D was a result of slightly higher average and maximum depths of the channel with no increase in bankfull width. A W/D ratio of 12 is a high end value for "E" stream types (Rosgen 1996). The W/D ratio is used to understand the distribution of energy within a channel. If the W/D ratio increases, the hydraulic stress against the banks also increases and bank erosion is accelerated (Rosgen 1996). Table 5. Width/depth (W/D) ratio and maximum depth (ft) of riffles and pools (represents the average of seven stream segments). Time Riffle 1 Pool 1 Riffle 2 Poo12 , Initial W/D Two Years W/D Initial max depth Two years 11.4 9.2 11.3 _ 10.0 10.0 9.6 10.2 9.5 2.06 2.97 2.21 2.72 2.74 3.07 2.88 3.24 �- 16 Other physical characteristics of the stream segments suggest that the restored channel was not as sinuous as designed. This was reflected in the higher meander wavelengths and radius of curvature and lower belt widths of channel segments (Table 6) as compared with design criteria (Table 1). However, channel configuration has not changed after two years of water flow, suggesting that the geometry of the restored channel was suitable for the various�flow conditions that occur in Tulula Creek. � Table 6. Other physical characteristics of selected meanders in each stream segment. Section I Ia II � IV Na V Average Meander Wavelength (ft) 65.6 68.9 95.1 98.4 137.8 75.5 75.5 .. Arc Belt Radius of Length (ft) Width (ft) Curvature (ft) 45.3 24.3 55.8 66.3 61.4 42.7 59.1 50.5 42.7 43.6 55.8 57.4 77.1 22.9 57.1 50.9 � 19.4 10.2 23.3 21.0 21.3 24.3 22.3 20.3 The cumulative pebble counts of the eight stream segments are shown in Fig. 3. With the exception of stream segment Va, 40 to 70 % of the cumulative pebble counts were found in the silt/clay.fraction. Segment Va is the closest representation of the relic.channel of Tulula, with minor adjustments made to small portions of the stream bank during stream re-construction. Roughly 10 % of the pebble count was sildclay in this segment. With the addition of sands, 80 to 98 % of the pebble counts were accounted for in the eight stream segments. The additional ti stream bed materials consisted of graveL i� � � L Q� .i-r rs. > .� � � � U ... c a� � a� a 100 :� .� 40 � � :.� ..._:�.._.__. � � — I ��— = — �� _�CY'� I _ _-- <Y ...�.��.� _ -- : .....��-_"i�� ........� I T � / _ � � � � — �% // I I �i� I �Y� � � � � 20 I � /.. c,''�ti ' � � 0 1- o.�, Sands �— —�— --F— --�-- ........ �........ —�—��— _..�.—.._ Segment I Segment Ia Segment II Segment III Segment IV Segment V Segment Va Gravels o.ao3 o.00s a.o�o 0.020 o.oao o.oeo o.�eo 0.24m.310 o.a�oo.eso o.eao�.zeo �.sooz.soo Particle Size (in) Fig. 3. Cumulative pebble counts of seven stream segments. Bank inclinations of riffles and pools created for the restored channel were commonly between 20 and 30 degrees (data not shown). Although significant erosion was noted at the bottom of the banks (toe of the bank slope) of riffles and pools (Table 7), overall bank inclinations did not change appreciably after two years of water flow because of the lack of erosion in the middle and upper portions of stream banks. The erosion noted at the bottom of channel banks through erosion control pins can be used to evaluate the lateral stability of a channel. Several points along the re-constructed Tulula channel are at risk of instability based on lateral erosion, most notably the riffle/pool sequence of Section Ia, and to a lesser extent Riffle 2 of Section III and Pool 1 of Section N. The meander width ratio (meander belt width divided by bankfull channel width) is another indicator of lateral stability. Given the lack of changes in meander belt or bankfull width after two years of water flow, the ratio has not changed, suggesting that the re-constructed channel is fairly stable. The overall channel configuration has not changed substantially after two years of water flow. However, changes in channel depth have altered the cross-sectional areas of riffles and pools and changed the W/D ratio. Desirable features have formed in the channel, most notably point bars on inside banks of many meanders. Changes in cross section and bank erosion at certain locations suggest that the channel is still adjusting to the flow regimes of Tulula Creek. Minor adjustments can be made for areas that appear to have unstable banks or stream bed conditions. Table 7. Erosion of channel banks after two years of water flow, based on erosion control pins. Segment I I Ia Ia Ia II II II � � N IV N IV 2. Hydrology Feature Pool 1 Riffle 2 Riffle 1 Pool 1 Riffle 2 Pool 1 Pool 1 Riffle 2 Riffle 1 Riffle 2 Riffle 1 Pool 1 Pool 1 Riffle 2 Location Toe Toe Toe Toe Toe Toe Middle Toe Toe Toe Toe Toe Middle Toe Erosion (inches) 2.02 5.26 11.26 11.98 3.85 1.62 2.02 0.16 0.40 5.66 0.40 4.86 0.81 0.40 18 Concurrent with construction of the new channel, drainage ditches were blocked and filled. The expectation was that re-constructing a meandering channel would decrease water velocity, which, when coupled with blocked drainage ditches, would raise the level of the water table across the floodplain and allow for more frequent overbank flooding. One of our objectives was to determine if site restoration improved the overall site hydrology. Electronic water table wells were installed in July 2000 along transects that were perpendicular to the new channel (Fig 4). In addition, site hydrology has been monitored for over ten years with a series of manual water table wells and piezometers (Fig. 5). Many of the manual wells and all of the piezometers are located in a 4-ha floodplain/fen complex that serves as a reference area for several LTNCA research projects. We have documented seasonal patterns of water-table elevation and vertical hydraulic gradient in this area and'determined the influence of hillslopes and drought on fen hydrology (Moorhead 2001, Moorhead 2003). Fig 4. Transects and individual electronic wells used to assess site hydrology of the restored stream channel. See Appendix A for daily water-table levels of wells. Methods 19 Both electronic and manual water-table wells were used to determine if the floodplain water table was higher because of the new channel and blocked drainage ditches. Methods of installation are described in Moorhead et al. (2001 a). The manual wells were read two to four times a month. The electronic wells were programmed to record the water-table depth on a daily basis. The data for both types of wells were converted to monthly averages to compare the pre- and post-restoration conditions. The monthly data were then used to construct hydrographs over a one-year period that coincided with the release of water in the various stream sections. For example, the months of September through the following August were used for developing hydrographs for electronic or manual wells in stream section I(water release in September, 2001). Differences between the average monthly pre- and post-restoration water-table levels were analyzed with a Student's t-test in Microsoft Excel. Results and Discussion The success of hydrology restoration at Tulula, like many wetland sites, will be determined primarily by changes in water-table depth. The assumption was that after the channel was restored and the drainage ditches were plugged, the overall water table of the site would rise. The electronic wells were also used by NCDOT to determine the success of wetland hydrology as determined by the Section 404 permitting system of the U.S. Army Corps of Engineers (at least 12 consecutive days of inundation or saturation during the growing season; North Carolina Department of Transportation, 2003): 20 Fig. 5. Location of manual wells at Tulula. Wells A160, A120, B200, B160, B120, and D20 were destroyed during site restoration and were not replaced. The electronic wells were installed in July 2000 and one or two years of pre-restoration data were compared to two years of post-restoration data, depending on the date of water release into the various stream sections. The data from individual wells are organized by stream section. As an example, the daily water-table graphs of electronic well X1 (in stream section 1� and the monthly averages .are shown in Fig.6. A comparison of the pre- and post-restoration monthly averages provides an easier visual interpretation of changes in water table depth due to restoration. The remaining monthly averages of water-table graphs of electronic wells.are found in Appendix B. 21 A rise in the water table was viewed as an improvement in site hydrology. For example, the restoration of the stream channel improved the hydrology at Xl. In stream section I, the water table increased in the following electronic wells: H3, G1, G2, and X1 (Appendix B1). It was not as consistent at H2, and although there appeared to be an overall raise at Il, the restoration of Tulula Creek and hydrology did not improve site hydrology at I1 to meet the requirements of wetland hydrology for the permitting process. Water-table graphs from electronic wells in stream section II and III showed a consistent raise in the water table after restoration for electronic wells E1, E2, E3, and F2 (Appendix B2). There was no consistent water table rise for D-transect electronic wells associated with stream section N(Appendix B3). In section V and Va, the water rose after restoration for electronic wells C1, C2, B1, B3, B4, B5, and A3. However, several of these wells were influenced by the flooding of the lower end of the site by beaver dams. In particular, C1, C2, B4, B5, and A3 are located near or in areas of flooded conditions from beaver dams. Data from some of the manual wells have been collected since 1994 (locations of wells shown in Fig. 3). Monthly averages of water-table depth were calculated for seven years of pre- restoration data and two years of post-restoration data. The figures illustrating the pre- and post- restoration water-table data from individual manual wells are found in Appendix C. There are seven years of pre-restoration data including three years of drought conditions (July 1998 through fa112001 (Moorhead 2003). The �iata from manual wells provide a more comprehensive view of site hydrology, given the varied conditions.of annual precipitation before restoration, given the three drought years and the higher than average annual precipitation during June 1994 through 1997. At the eastern side of the site, the depth of the water table of Tulula fen (wells 3C, 3F, 6I, 9I, 7F, and 8C; Appendix C 1 a) and the floodplain adjacent to it (wells II1 and 2, III1 and 2, N 1 and 2; Appendix Clb) showed few statistical differences before and after.restoration of site hydrology. The statistical differences were noted more often in summer months, during periods of plant transpiration. Based on manual wells, the water table of Tulula was improved (higher) for wells located near the stream channel (F1, T13, T14), with little or no improvement documented for wells located farther from the channel (T1, T2, T3, T4, T5, T6, T7, T8, T9, T10, T 11, T 12; Appendix C 1 c). Collecting water-table data over the next few years and comparing pre- and post-restoration water-table levels will provide a more comprehensive view of how site restoration has changed the hydrology of Tulula wetlands. '12 � 0 � -12 -24 _36 -48 � � � � � � � � � � � � a� a � a ? a a� a a � � a � � � X 2000/2001 pre 12 � O � -12 -24 _3g -48 � � � � � � � � � � � � r � � � � x soo, i2oo2 Post X 2002/2003 post '12 O � -12 -24 _36 -48 � � ' � � � � � � � � � � a� a � A � A a� a a m � � � � � 12 •� 0 `' -12 � -24 �� -36 -48 morrthly avg pre and post X S O N D J F M A M J J A month -♦-1 YR PRE —� 2 YR POST 22 Fig. 6. The daily water table and monthly averages tor electromc we11 x t. �tansticai airrerences (P < 0.05) were noted for the monthly averages of all months except June. Depth of "0" represents the surface of the soil. . 23 The main concern of NCDOT will be whether the wetlands of the Tulula floodplain have the appropriate hydrology to meet permit conditions. The data required for this determination are collected with the electronic wells and analyzed on a yearly basis (see North Carolina Department of Transportation, 2003 for examples). A more interesting ecological question is how the overall hydrology has changed at Tulula with site restoration. The manual wells will provide more information for this question since they were installed in 1994. B.. Vegetation responses to restoration One of our objectives for restoring wetlands in the Tulula floodplain has been to monitor the response of native wetland plant communities. We have been monitoring the community composition of an intact fen since 1994, and during this funding cycle, we were able to examine the community post-restoration. We also sought to gain a better understanding of the relationship between wetland plants and environmental factors such as hydrology. We used Juncus effusus L. (soft rush), which is an easily recognizable and widespread species in the Tulula floodplain (and elsewhere), as an indicator species to evaluate the effects of hydrology and restoration on plant growth and reproduction. In a previous seed bank study at Tulula, Rossell and Wells (1999) reported that Juncus spp. dominated the wetland seed bank, especially in an early successional area of the fen. Our objectives were to determine whether the growth and reproduction of Juncus ef)`'usus were enhanced by wetland restoration, and• how overall species richness responded to restoration. 1. Plant growth responses to restoration and hydrologic regime Methods We used data from groundwater wells to select four sites at Tulula: an undisturbed wet area, a nearby undisturbed drier area, a restored wet area, and a nearby restored drier area. At each site, we delineated a SOm x l Om study area in relatively uniform plant communities. Within each of the four study areas, we established 20, 0.25-m2 quadrats at randomly selected points (using a table of random numbers). Our only criterion was that all quadrats contained Juncus effusus. If a randomly selected quadrat did not contain J. effusus, it was rejected, and another random quadrat was selected. In July 2003, we surveyed the plant associates of Juncus effusus in each study area. All plants occurring in a1180 quadrats were identified to species, and coverage within the quadrat was visually estimated. We obtained the Region 2(southeastern United States) wetland indicator status for each species by consulting U.S. Fish and Wildlife Service (1996) and U.S. Department of Agriculture (2001). Wetland indicator status categories describe wetland affinities as follows: obligate wetland plants (OBL) occur in wetlands >99% of the time, facultative wetland plants (FACV� occur in wetlands 67-99% of the time, facultative plants (FAC) occur in wetlands 34- 66% of the time, facultative upland plants (FACU) occur in wetlands 1-33% of the time, and upland plants (LJPL) occur in wetlands <1% of the time. 24 In early,September 2003, we used shears to harvest all aboveground plant material within each quadrat. Plant material was placed on tarps, then sorted into four categaries: vegetative stems of Juncus effusus, reproductive stems of Juncus effusus, non-Juncus effusus herbaceous� plants, and woody plants. All plant material was placed into paper bags, air-dried to constant weight in a warm dry building, and weighed. The numbers of Juncus effusus vegetative and reproductive stems were counted. All Juncus effusus inflorescences were clipped off of reproductive stems, and weighed separately. We performed an analysis of variance (ANOVA) to determine the effects of restoration status (restored vs. unrestored) and hydrology (wet vs. dry) on the following variables: number and biomass of vegetative Juncus effusus stems, number and biomass of reproductive Juncus effusus stems, biomass of Juncus effusus inflorescences, biomass of non-Juncus effusus vegetation, and biomass of woody vegetation. Statistical Analysis Systems was used for all analyses (SAS 2001). � Results and Discussion � Juncus effusus (a FACW species), although present in all quadrats, never occupied more � than 25% of the area of any one quadrat. In half to three-fourths of all quadrats, Juncus effusus • occupied <5% of the area of the quadrat. Clearly, although a consistent presence in all of our . quadrats, Juncus effusus was not a dominant species overall. It had many associate species that were reflected in our calculations of taxonomic richness. � � Taxonomic richness was greatest in the restored dry area (48 taxa), and lowest in the • restored wet area (17 taxa). Richness was intermediate in the unrestored dry (33 taxa) and • unrestored wet (37 taxa) areas (Table 8). OBL and FACW species made up the greatest percentage of the flora in the restored wet area (93.3%), and the smallest percentage in the � unrestored dry area (55.5%) (Table 9). The percentage of OBL and FACW species in the � unrestored wet (67.7%) and in the restored dry areas (60.5%) were similar. � � � � � � � � � � � � � � � � Table 8. Taxa and wetland indicator status of plants occurring in 0.25-m2 quadrats in four study areas at Tizlula. Taxon Area Unrest. Unrest. � Wet Aca/ypha rhomboidea Raf. Acer rubrum L. Agalinis purpurea (L.) Pennell Agrimonia parviflora Ait. Agrostis sp. Alnus serrulata (Ait.) Willd. Ambrosia artemisiifolia L. x Ambrosia trifida L. Andropogon virginicus L. X Apios americana Medicus X Aster novae-angliae L. X Aster pilosus Willd. X Bidens frondosa L. Boehmeria cylindrica (L.) Sw. Campanula aparinoides Pursh. X Carex annectens (Bickn.) Bickn. Carex debilis Michx X Carex festucacea Willd. Carex lurida Wahl. X Carex scoparia Schkuhr ex. Wilid. X Carex sp. 1 X Carex sp. 2 Cassia fasciculata Michx. Clematis virginiana L. X Cuscute campestris Yuncker Cyperus strigosus L. Desmodium cuspidatum (Willd.) X Loudon Dicanthelium clandestinum (L.) Gould x Dicanthelium ensifolium X Eleocharis obtusa (Willd.) Schultes Eleocharis tenuis (Willd.) Schultes Epilobium ciliatum Raf. Erigeron ani►uus (L.) Pers. Erigeron philadelphicus L. Eupatorium fistulosum Barratt Eupatorium perfoliatum L. Galium tinctorium L. Grass sp.1 . Grass sp.3 Grass sp.4 Holcus lanatus L. X X X X X X X X X X X X X X X X X X X X , Wetland Restored Restored indicator � Wet status x FAC x FAC x OBL FAC x FACW X FACW x FACU FAC FAC x FACW x. NA x NA x FACW FACW OBL FACW FACW FAC W x X OBL x FACW NA NA x NA x FAC X NA x FACW NA x FACW x NA � OBL FACW X NA x FACU x FAC x FAC x FACW x X FACW x NA x NA x NA FACU 25 Hypericum mutilum L. X Impatiens capensis Meerb. x Juncus acuminatus Michx. Juncus brevrcaudatus (Engelm.)Fem. Juncus effusus L. X Juncus tenuis Wilid. X Lespedeza cuneata (Dumont) G.Don Liriodendron tulipifera L. Lobelia puberula Michx. Ludwigia altemifolia L. Mimulus ringens L. Onoc/ea sensibilis L. Osmunda cinnamomea L. Oxalis sp. X Oxalis stricta L. Panicum virgatum L. Persicaria hydropiper L. X Persicaria sagittatum L. X Persicaria spp. X Potentilla simplex Michx. x Prunella vu/garis L. X Pycnanthemum verticillatum (Michx.) X Pers. Rhynchospora glomerata (L.) Vahl. Rosa palustris Marsh. X Rubus argutus Link X Rubus hispidus L. X Sagittaria latifolia Willd. Sambucus canadensis L. X Scirpus expansus Fem. Scirpus polyphyllus Vahl. Solidago gigantea Aiton X Solidago n.rgosa Miller X Sparganium americanum Nutt. Trifolium campestre Schreb. Trifolium repens L. Vemonia noveboracensis (L.) Michx. X viola sp. X X X X X X X X X X X X X x x x x x x x x x x X X X X X � x x X x X x X X X X x X X x x x X x X x x X X X X FACW FACW OBL OBL FACW FAC NA FAC FACW �OBL OBL FACW FACW UPL UPL FAC OBL OBL NA FACU FAC UPL OBL OBL FACU FACW OBL FACW OBL OBL FACW FAC OBL NA FACIJ FAC NA 26 27 Table 9. Contribution of each wetland indicator status (as a percent of all vegetation in 0.25-m? quadrats) in four study areas at Tulula (for plants with a known indicator status). Unrestored area Restored area Wetland Indicator Status Drv Wet Dry Wet OBL FACW FAC FACU UPL 18.5 26.5 23.7 60.0 37.0 41.2 36.8 33.3 22.2 17.6 23.7 0 14.8 8.8 13.2 6.7 7.4 5.9 2.6 0 The results of our ANOVA showed that vegetative Juncus stems were more numerous (P<0.0001), as well as heavier (P=0.004) in unrestored areas (Table 10). Similarly, non-Juncus herbs were heavier in unrestored areas (P=0.002) (Table 11). Neither the number nor the biomass of reproductive Juncus stems were influenced by restoration (P>0.05). When water table was considered, vegetative Juncus stems were more numerous (P=0.007) as well as heavier (P=0.002) in wet areas (Table 12). Reproductive Juncus stems were more numerous (P=0.015), but not heavier (P=0.064), in wet areas. The biomass of non-Juncus herbs was lower in wet areas than in dry areas (P<0.0001)(Table 13). Table 10. Effects of restoration on vegetative growth and reproduction of Juncus effusus. Within columns, means followed by the same letter do not differ significantly (P>0.05). Vegetative Juncus stems Treatment No. Biomass (g) Unrestored 190.Oa 23.4a Reproductive Juncus Juncus stems � inflorescences No. Biomass (g� Biomass (g) 15.Sa 8.4a 1.6a Restored 69.Ob 13.6b 11.Sa 6.2a 1.2a Table 11. Effects of restoration on biomass of plants occurring with Juncus effusus in 0.25-m2 quadrats. Within columns, means followed by the same letter do not differ significantly (P>0.05). Treatment Unrestored Restored Herbaceous plants (non-Juncus ) (g) 50.3b E:� Woody plants 78.Sa 39.7a Table 12. Effects of hydrology on vegetative growth and reproduction of Juncus effusus. Within columns, means followed by the same letter do not differ significantly (P>0.05). Vegetative Juncus stems Treatment No. Wet 156.2a Dry 102.8b Biomass (g) 23.8a 13.1b Reproductive Juncus Juncus stems inflorescences No. Biomass (g) Biomass (g) 19.7a 9.Sa 1.8a 7.4b 4.3a 0.8a zs In summary, the disturbance that is inherently part of restoration activities clearly benefited the growth of non-Juncus herbaceous plants, perhaps by opening up the canopy and minimizing competition for light. In contrast, Juncus effusus was more numerous and heavier in uridisturbed areas, perhaps because it is less competitive than the associated flora. A high water table benefited Juncus effusus (a FACW species) more than the associated flora, however, and stimulated the production of reproductive stems, ensuring the continued presence of Juncus effusus in the seed bank over the long term. Overall, plant taxonomic richness was greatest in restored dry areas, but lowest in restored wet areas, implying that a high water table inhibited many species and favored the establishment of OBL and FACW plants. � Table 13. Effects of hydrology on biomass of plants occurring with Juncus effusus in 0.25-m2 quadrats. Within columns, means followed by the same letter do not differ significantly (P>0.05). Treatment Wet Dry Herbaceous plants (non-Juncus ) (g) 41.Ob 92.1a Woody plants (g) 98.Sa 36.3a 2. Vegetation dynamics in Tulula Fen and adjacent floodplain To determine the effects of wetland restoration on plant communities in an intact fen at Tulula, we examined the community composition of open and closed canopy areas of fen and adjacent disturbed floodplain. The vegetation in these areas was inventoried twice prior to restoration (1994 and 2001). We repeated the inventory of each area in July 2003, in order to evaluate any changes that might have arisen as a result of the altered hydrology at the site. Methods We inventoried vegetation using the protocol established in 1994, and a grid of 120 ydz plots that was laid out throughout the fen in 1994. Within this grid, 20 plots were randomly selected in an area with a closed canopy, and 20 plots in an area with an open canopy. In each 32.8 ft x 32.8 ft plot, we identified all overstory trees with a DBH > 4 in, and measured its DBH. In nested 13.1 ft x 13.1 ft plots, we identified all understory trees and shrubs with a DBH of 0.8 — 4.0 in, and measured their DBH. In nested 3.3 ft x 3.3 ft quadrats, we identified all herbaceous plants and woody seedlings (DBH < 0.8 in), and visually estimated their percent cover. In an adjacent floodplain that was disturbed by the golf course developers for the purpose of creating a golf fairway, 6, 65.6 ft x 98.4 ft plots were established in 1994. Within each of these 6 plots, overstory trees were inventoried in an 59 ft x 59 ft plot, and understory trees were inventoried in a 23 ft x 23 ft plot (these plot sizes were selected so that the total area inventoried in the floodplain was consistent with the total area inventoried in each area of the fen). Within each of the 6 plots, we inventoried herbaceous and woody vegetation in 4, 3.3 ft x 3.3 ft quadrats (N=24). We obtained the Region 2(southeastern United States) wetland indicator status for all woody species by consulting U.S. Fish and Wildlife Service (1996) and U.S. Department of Agriculture (2001). Importance values (N's) were calculated for all overstory and understory woody species, and for six groups of herbs/woody seedlings (ferns, forbs, grasses, rushes, sedges, and woody seedlings). For the overstory and understory species, N's were calculated based on density, basal area, and frequency of occurrence. For the six groups of herbs/woody seedlings, N's were calculated based on percent cover and frequency of occurrence. 30 Results and Discussion During the nine years of this study, the number of red maples in the overstory of the closed canopy fen increased from 104 in 1994, to 123 in 2003 (Table 14). However, the overall importance of red maple (Acer rubrum L.) declined, from IV=93 in 1994, to N=86 in 2003 (Table 15). At the same time, the importance of white pine (Pinus strobus L.) increased, from N=3 in 1994, to N=11 in 2003. Conversely, in the open canopy area of the fen, the importance of red maple increased during this 9-year period. In 1994 there were no overstory trees (DBH > 4.0 in) in the open canopy area of the fen. By 2003, 14 overstory-sized red maples were present in this area. Table 14. Total number of overstory trees of each species in 20, 1Ox10-m? plots. Closed Fen Open Fen Floodplain Species 1994 2001 2003 1994 2001 2003 1994 2001 2003 Acer rubrum L. Amelanchier sp. Ilex opaca Ait. Malus angustifolia (Ait.) Miclix. Pinus strobus L. 104 109 123 1 - - 1 1 1 - 2 ' 1 2 5 11 - 4 14 - Total 108 117 136 - 4 14 - - - Table 15. Importance values for overstory trees in'1Ox10-m2 plots. Wetland Closed Fen Open Fen Species Status 1994 2001 2003 1994 2001 2003 Acer rubrum FAC 93.2 88.6 86.0 - 100 100 Amelanchier sp. - 1.7 - - - - - Ilex opaca FAC 1.7 1.8 1.4 - - - Malus angustifolia - - 2.1 1.4 - - - Pinus strobus FACU 3.4 7.5 11.2 - - - Floodplain 1994 2001 2003 31 In the understory of the closed canopy fen, neither the total number of stems (Table 16) nor the importance of any species changed appreciably between 1994 and 2003 (Table 17). However, in the open canopy fen, the total number of stems of understory-sized trees (DBH 0.8 - 4.0 in) increased dramatically from 121 stems in 1994, to 234 stems in 2003 (Table 16). Most of this increase was due to the number of red maple and tag alder (Alnus serrulata (Ait.) Willd.) stems that entered this size class. The overall importance of red maple declined from N=89 in 1994, to N=71 in 2003, primarily because the taxonomic richness in this area increased from 3 understory-sized species in 1994, to 12 understory-sized species in 2003 (Table 17). - Table 16. Total number of understory trees of each species in 20, 4x4-m2 plots. Closed Fen Open Fen Floodplain Species 1994 2001 2003 1994 2001 2003 1994 2001 2003 Acer rubrum L. 73 Alnus serrulata (Ait.) Willd. 3 Amelanchier laevis Wiegand - Aralia spinosa L. - Aronia arbutifolia (L.) Ell. - Aronia melanocarpa (Michx.)Ell. Ilex opaca Ait. - Ilex verticillata (L.)Gray 3 Liriodendron tulipifera L. - Malus angustifolia (Ait.)Michx 2 Nyssa sylvatica IVtarsh. 5 Oxydendrum arboreum (L.)DC 1 Pinus stro6us L. 4 Prunus serotina Ehrhart - Rhus copallina L. - Rosa palustris Marsh. - Salix sericea Marsh. - Sambucus canadensis L. 1 Vaccinium corymbosum L. - Viburnum cassinoides L. 9 Total 77 72 lll 157 174 4 1 8 23 33 - - - 2 2 - - - 1 1 1 - - 7 3 - - - 5 - 5 4 - - 1 5 6 - 1 - - - - 1 1 - 6 1 - 6 15 1 - - - - - - - 5 6 - 2 4 - - - 2 7 1 6 2 7 1 6 2 4 1 1 2 4 1 1 12 1 101 lt4 105 121 2I1. 234 1 2 15 1 17 In the ground-layer community, the closed canopy fen showed an increasing dominance by ferns (mostly cinnamon fern, Osmunda cinnamomea L.) over the nine years of this study (N=32.5 in 1994, N=48.5 in 2003) (Tables 18 and 19). The open canopy fen showed a decline in the importance of rushes (N=10.2 in 1994, N= 0 in 2003) and an increase in the importance of woody plants (N=19.5 in 1994, N=29.2 in 2003). Since many rushes flourish in open, sunny areas or those with only partial shade (Thunhorst 1993), it is likely that the shading created during natural succession at Tulula will largely eliminate rushes from this area of the fen. Woody plants increased even more in the adjacent disturbed floodplain (N=19.3 in 1994, N=40.2 in 2003). Table 17. Importance values for understory trees in 4x4-m2 plots. Wetland Closed Fen Open Fen Species Status 1994 . 2001 2003 1994 2001 2003 Acer rubrum FAC 69.8 65.3 65.6 89.0 67.9 70.9 Alnus serru/ata FACW 3.4 3.0 1.4 10.6 7.8 9.2 Amelanchier laevis - - - - - 1.2 1.5 Aralia spinosa FAC - - - - 1.0 1.4 Aronia arbutifolia FACW - 1.2 - - 5.5 2.9 Aronia melanocarpa FAC - - - - 2.6 - Ilex opaca FAC - 4.2 4.6 - - 1.4 Ilex verticillata FACW 3.4 4.2 5.3 - 1.0 - Liriodendron tulipifera FAC - - - - I.0 1.4 Malus anguslijolia - 1.9 1.2 - ' - - - . Nyssa sylvatica FAC 5.3 4.2 6.4 - 2.1 4.3 _ Ozydendrum arbreum UPL 1.5 2.4 2.9 - - - Pinus strobus FACU 6.3 8.5 8.0 3.3 33 1.4 Prunus serotina FACU - - - - 1.0 1.4 Rhus copallina FACU - - - - - - Rosa palustris OBL - - - - 1.0 - Salix sericea OBL - - - - 1.2 3.0 Sambucus canadensis FACW 1.5 - - - 33 1.4 Vaccinium corymbosum FACW - 1.2 1.4 - - - Viburnum cassinoides � FACW 6.9 4.6 4.2 - - - Floodplain 1994 2001 2003 - 41.6 13.4 - 34.2 73.3 - - 13.3 - 9.9 - - 143 - Table 18. Mean percent cover of each plant type in 1 x 1=m2 quadrats. Plant type Fern Forb Grass Rush Sedge Woody Closed Fen Open Fen Floodplain 1994 2001 2003 1994 2001 2003 1994 2001 2003 21.8 0.9 0.4 0.2 12.0 19.5 12.2 0.5 0.2 0 6.4 13.6 15.2 8.9 . 11.0 0.2 4.8 ' 4.3 0.1 12.6 4.5 0 4.6 0.1 2.9 32.9 21.0 4.9 15.6 23.3 6.4 0.3 1.8 3.1 2.7 29.7 10.4 14.7 2.8 23.3 3.0 6.5 0 6.2 1.0 0.8 18.8 2.9 20.4 3.1 17.1 14.7 41.8 42.5 K�� The number of dead trees in the closed canopy region of the fen increased somewhat during our study for understory-sized trees. In 1994, we recorded 7 dead stems in this size class, compared with 13 dead stems in 2001 (we counted 11 dead stems in 2003, but some of those could have been standing since 2001). Because the site restoration was not complete by 2001 and the hydrology had not been altered in this part of the floodplain, the most likely causes of death for these stems aze dry conditions at the site during the 1990's, shading, and/or disease. Table 19. Importance values for plant types in lxl-m2 quadrats. Planttype Fern Forb Grass Rush Sedge Woody Closed Fen Open Fen Floodplain 1994 2001 2003 1994 2001 2003 1994 2001 2003 32.5 7.4 3.6 1.5 24.1 31.0 33.8 3.8 3.3 0 23.3 35.8 48.5 3.6 2.6 0 18.9 26.5 11.4 11.3 17.2 10.2 30.4 19.5 14.9 12.1 13.9 1.2 28.0 29.8 13.6 11.5 13.6 0 32.1 29.2 1.9 29.4 25.3 12.9 11.2 19.3 3.4 17.0 11.4 7.4 23.4 37.5 5.6 20.6 14.3 8.3 11.0 40.2 33 In summary, the changes in the fen reflect what might be expected due to natural succession, but not to changes in hydrology due to site restoration. The closed canopy fen continues to be dominated by red maple, although the overstory shows a small increase in white pine. Given the overall lack of disturbance in recent years, the open fen is reverting to a forested canopy, and is dominated by red maple. Heliophytic herbaceous plants like rushes are decreasing throughout the fen, while shade-tolerant herbs such as ferns are increasing. 3. Survival of commercial red maple stock Methods During the winter of 1995, we planted 77 red maple seedlings in each of three of the 65.6 ft x 98.4 ft plots in the disturbed floodplain (N=231). We re-inventoried these red maple saplings during fa112003, so that we could compare their survival to that determined in several . previous (pre-restoration) years. Results and Discussion . Survival of the commercial red maple seedlings appears to have declined somewhat during 2003 (Table 20). The number of surviving saplings had been relatively steady from 1995 unti12002 (some of the discrepancies in the results of each year's survey are likely due to the fact that there are now thousands of naturally-regenerating red maple saplings in this floodplain, and it is sometimes difficult to determine whether a saplings was planted, or has regenerated on its own). Survival during 2003 was 10% less than it has been since 2000. It is premature to pinpoint the factor(s) that are responsible for the decreased survival of these saplings during 2003, but one of the most significant is likely competition from the aggressive growth of blackberries (Rubus argutus Link) and other tall shrubs in some areas of this floodplain. In these azeas, the planted red maple seedlings.have been overtopped by other vegetation. Other factors that have influenced the survival of the planted saplings over the last few years have included browsing by deer, and the _ spraying of herbicides in the vicinity of a lazge powerline that crosses the floodplain (the herbicides were sprayed by the local power company, in an effort to control vegetation under the powerline). 34 Table 20. Survival of 231 commercial red maple seedlings planted in Tulula floodplain during winter 1995. " Year Survival (%) 1995 77 1996 71 2000 76 2001 81 2002 76 2003 66 Despite the reduced survival of planted saplings, naturally-regenerating red maple saplings continue to flourish in the Tulula floodplain. This trend is documented by Warren et al. (2004), who conducted comprehensive surveys of red maple regeneration across this floodplain in 1994 and 2001. They reported that red maple readily colonized wetland habitats, ��•ith a post- disturbance recruitment window lasting at least twice as long as that reported for tcrrestrial habitats. ' C. Effects of Restoration on Decomposition and Soil Microfauna Decomposition is a primary ecosystem function in the recycling of nutrients (Swift et al. 1979, Seastedt 1984), and is influenced by factors such as soil nutrients, temperature, composition of plant material, and composition and activity of soil fauna. Although many studies have examined decomposition in upland hardwood communities in the southern Appalachians (see Reynolds et al. 2003), and some research has focused on decomposition in cypress-gum wetlands (Battle and Golliday 2001) and playa wetlands in the southern Great Plains (Anderson and Smith 2002), little is known about decomposition in wetlands of the southern Appalachians. The vital role of microarthropods in decomposition and nutrient cycling has been long established (Swift et al. 1979), but research in wetland systems appears to be minimal. Braccia and Batzer (2001) examined invertebrates associated with woody debris in a southeastern floodplain wetland, but their study did not include decomposition. Indeed, these authors emphasized that tenestrial wetland fauna have been overlooked, and they found that non-aquatic (including Acari and Collembola) rather than aquatic arthropods, were the most significant component of overall community structure. We conclude, therefore, that the present research, combining decomposition studies with microarthropod data, is not only useful but ground- 'breaking. In this section, we report on decomposition and microarthropod studies conducted in five plant communities at Tulula, and relate these data to soil pH and organic matter. 35 1. Decomposition Methods Six plots, co-located with water table wells, were established in each of five plant community types at Tulula. Plant communities used were the red maple forest (RM), open (OF) and closed fen (CF), floodplain(FP), and the former fairway — a disturbed alluvial bottomland forest (DA). Twelve fiber-glass screen litter bags, 6 x 6" with mesh size of 1/16 ", containing known weights of air=dried Acer rubrum (red maple) leaves were placed in each plot in a 4 x 3 grid. The fresh-fallen leaves were collected in October, 2002, and the litter bags placed in the field in January, 2003. Each litterbag was anchored with a survey flag and lightly covered with surrounding litter. One litterbag was removed from each plot every other month, beginning in March, 2003 and continuing through May of 2004. Bags were transported in zip-loc bags to the lab, and the litter content weighed after microarthropod extraction. Percent mass of the " remaining litter was calculated. Tukey's Studentized Range (HSD) Test, (SAS version 8), was used for statistical analysis. � Results and Discussion After 17 months in the field, the percent mass of litter remaining averaged 50% for the RM site to 54.8% for the FP site (Fig. 7). The percent mass remaining in RM, CF, and OF was not significantly different, nor was the mass remaining in FP, DA, OF, and CF. However, decomposition was significantly greater in RM than in DA and FP. This indicates to us that in the least disturbed site, which has an intact canopy and least disturbed soil, the important process of decomposition occurs most readily. Therefore, one would assume that the ensuing process of nutrient cycling would also occur most readily in the least disturbed, RM site. Since there is no significant difference in percent mass remaining between the FP and the OF, we conclude that the presence of the closed canopy in the RM site is not as important in determining decomposition rate as the intact soil, although the differences in moisture may be a factor in decomposition. 2. Litter Microarthropods Methods Microarthropods were extracted from litterbags using a modified Tullgren funnel apparatus (Mallow and Crossley 1984). Litterbags were left on the funnels for 3 to 4 days; the extracted microarthropods were preserved in 70% ETOH. Microarthropods were sorted under a stereomicroscope into the following categories: oribatid, prostigmatid, and mesostigmatid mites, Collembola, and others. Microarthropod abundances were determined as the mean number of animals/15.43gr litter. Since the abundance values were not normally distributed, the data were analyzed using a Generalized Linear Model (Proc Genmod SAS version 8e, 2000) (Crawley 1993). Standard errors in graphs are provided for comparison purposes, but aren't statistically rigorous because the data do not conform with the assumptions of �ormality. 36 so 50 vi c c 40 .� E a� d 30 � c � 20 a`� a to 0 Percent Litter Remaining after 17 Months RM CF OF FP DA Plant Communities Fig. 7. Percent litter remaining in litterbags after 17 months in the field. Plant communities are RM=red maple, CF=closed feri, OF=open fen, FP=flood plain, and DA=disturbed alluvial bottomland forest. Each bar is the average of 37 to 481itterbags. Bars with the same letter are not significantly different; error bars are ± 1 SE. Results and Discussion As expected, microarthropod numbers varied significantly among the three dates analyzed (Table 21, Fig. 8). Similar seasonal variations have been reported for upland hardwood forests in the southern Appalachians (Reynolds et al. 2003). We also found differences in litter microarthropod numbers by site for total microarthropods and all individual taxa counted except for prostigmatida, which were not abundant enough for statistical analysis. However, due to significant date*site interactions, the interpretation of significant site differences for mesostigmatida and collembola is unclear (Table 21). In all sites, oribatid mites were by far the most common microarthropod (Fig. 8) and they were most abundant in the RM community, followed by CF. Abundances of oribatids (and total microarthropods) appear to be significantly lower in OF, FP, and DA. These findings could be related to the presence of a canopy in RM and CF, protecting litter-dwelling arthropods from extremes in temperature and from dessication when exposed to solar radiation. However, the low numbers of total microarthropods in DA and FP, compared to sites with more canopy (RM, CF, and OF), before leaves are present in March (Fig. 9), indicates that other factors are involved. We posit that soil disturbance, once again, plays a maj�r role in a critical ecosystem factor — the abundance of litter microarthropods. 37 Table 21. Microarthropod responses to date and site. Data analyzed were average numbers of microarthropods per 15.43 grains of red maple litter from �litterbags collected on each of three dates (March, May, and July of 2003). Organism Log-Likelihood Terms Chi-square df P Total 487.56 Date 40.72 2 <0.0001 Site 16.08 . 4 0.0029 Date*Site 11.31 8 0.1850 Oribatida 400.99 Date 33.20 2. <.0001 Site 14J6 4 0.0052 Date*Site 10.57 8 0.2275 Mesostigmatida 157.29 Date 54.54 2 <0.0001 Site 11.09 4 0.0256 Date*Site 17.17 8 0.0284 Collembola 55.26 Date 9.40 2 0.0091 Site 23.38 4 0.0001 Date*Site 19.25 8 0.0136 38 :1 :1 0 ` � 1� Z a � rn � 100 �ri � � � � � c� 0 U .� � N � rn c� � a� > ca 40 0 0 � o .•- .- • • •- i ••••. • : • • .IIC - � . •• i - F�ll (F CF F� D4 Rat G�rrruities Fig. 8: Average number of microarthropods/15.43 gr dry red maple litter for all three collection dates, March, May and July, 2003. Plant communities are RM=red maple, CF=closed fen, OF=open fen, FP=flood plain, and DA=disturbed alluvial bottomland forest. Each bar is the average of 15 to 18 litterbags; error bars are�± 1 SE. ' 35 � � 30 Z v � 25 rn v �ri � 20 � � � co � 15 U .� � �� N � f6 � N Q 5 0 . Average number of total microarthropods for March, 2003 RM CF OF FP DA Plant Communities 39 Fig. 9. Average number of total microarthropods/15.43 gr dry red maple litter for March, 2003. Plant communities are RM=red maple, CF=closed fen, OF=open fen, FP=flood plain, and DA=disturbed alluvial bottomland forest. Each bar is the average of 6litterbags; error bars are ± 1 SE. 3. Soil Properties Methods Eight samples were collected from the top 2" of soil from each litterbag plot with a soil probe in July 2003. Those samples were then composited into one sample/plot, giving 6 samples per plant community. Percent organic cazbon (OC) content was deternuned by the Walkley-Black method (Nelson and Sommers 1982); pH was measured on a 1:1 slurry of soil:distilled water using a Fisher Accumet pH meter and standard electrodes. Average values of pH and OC were calculated for each plant community and comparisons among the 5 sites were done using Tukey's Studentized Range (HSD} test, SAS version 8e (2000). Results and Discussion Average organic cazbon varied from 11.79% to 2.8%, and was highest in soils from the open fen (11.79%), with OC decreasing in this order: closed fen (11.11%) > red maple forest (9.52%) > flood plain (8.53%) > disturbed alluvial forest (2.80%) (Fig. 10). The significantly lower OC for DA is probably the result of bulldozing the area for a fairway. Sites with the least dishubance, the fens and red maple forest, have the highest OC in the soil. Since soil organic matter is known to be strongly influenced by soil fauna (Coleman arid Crossley 1996), these results appeaz to be correlated with the distribution of microarthropod abundances, especially for RM and OF (Fig. 8)., . 40 14 12 10 c 0 � c� 8 U �C �1 6 O \° 0 4 2 0 Organic carbon in soil RM CF OF FP DA Plant Communities Fig. 10. Average percent organic carbon for soil from five plant communities: RM=red maple, CF=closed fen, OF=open fen, FP=flood plain,'DA=disturbed alluvial bottomland forest. Each bar is the average of 48 soil samples. Bars with the same letter are not significantly different; error bars are ± 1 SE. Average soil pH values ranged from 4.36 to 3.72, with the DA having the highest pH (4.36), followed by FP (4.2), OF (4.04), CF (3.98), and RM (3.72) (Fig. 11). pH is.significantly lower for the red maple forest. 5.0 4.5 4.0 3.5 3.0 a 2.5 2.0 1.5 1.0 0.5 0.0 Soil pH RM CF OF FP DA Plant Communities �II Fig. 1 l. Average pH for soil from five plant communities: RM=red maple, CF=closed fen, OF=open fen, FP=flood plain, and DA=disturbed alluvial bottomland forest. Each bar is the average of 48 soil samples. Bars with the same letter are not significantly different; error bars are ± 1 SE. Summary of Decomposition and Soil Fauna We found that the least disturbed plant communities, red maple in particular, have the quickest decomposition, the greatest amount of litter microarthropods, the most soil organic carbon, and the lowest soil pH. We conclude that soil characteristics, related to less disturbance, rather than the presence of a closed canopy, are probably the main influences on decomposition and litter microarthropods. Therefore, the most intact ecosystems appear to be functioning at the healthiest levels. � � � � � D. Amphibian. Use of Tulula Introduction 42 Amphibians are increasingly being used as indicator species in restoration projects for small freshwater wetlands (e.g., Pechmann et al. 2001) because they are often community dominants, are sensitive to site hydrology, and can be easily monitored to assess ecosystem function. Amphibians play key ecological roles in wetlands in the southern Appalachian Mountains, and are the dominant vertebrate group in standing water habitats at Tulula. Because a major goal of wetlands restoration is to restore ecosystem integrity (e.g., to create functional ecosystems where all major community elements are sustained at viable levels), the response of amphibians to site restoration is a useful indicator of ecosystem function. Because of their strong reliance on seasonal wetlands for breeding, the reproductive success of many amphibian species is strongly influenced by hydroperiod (seasonal duration of ponds). The hydroperiod affects the likelihood of amphibian larvae reaching a minimum developmental stage to complete metamorphosis. It also influences the distribution and abundance of predators such as fish and aquatic insects that feed on amphibian eggs and larvae. Short hydroperiods during periods of drought can result in catastrophic mortality of larvae due to premature pond drying, but also reduce or eliminate aquatic predators. Long hydroperiods during wet years provide ample time for amphibian larvae to complete metamorphosis, but may result in heavy mortality from predators such as dragonfly larvae that prefer semi-permanent ponds. '� At the initiation of the study in 1994, the site contained aquatic habitats that varied from . highly ephemeral to permanent ponds. Most natural breeding sites were filled during golf course • construction. During a detailed survey of the site during 1994-1995, we located 155 standing- water habitats that included 11 permanent ponds that were constructed as golf course obstacles. � Permanent ponds contained predatory fish (bluegills, largemouth bass) and were not used as � breeding sites by most resident amphibians. The remaining 144 sites were fish-free, seasonal • habitats that were mostly small, shallow depressions. These included mud puddles, water-filled , tire ruts, test wells for pond sites, sluggish ditches, and stream cut-offs associated with the channelization of Tulula Creek. � � Monitoring of seasonal habitats during 1994-1995 indicated that most breeding sites were • of very low quality because of altered site hydrology associated with stream channelization, ditching; and the filling of low-lying areas. All species of vernal pond-breeders suffered high � larval mortality during 1994 and 1995 because most breeding sites dried prematurely before � tadpoles or salamander larvae could complete their larval stages. Despite heavy rains in late • winter and early spring, about 75% of the breeding sites dried prematurely in 1994 and 60-70% • in 1995. These observations indicated a need to construct larger and deeper ponds to replace natural breeding sites that were destroyed during golf course constxuction. � � � � � � � 43 Ten vernal ponds were constructed between October 1995 and January 1996 to replace natural breeding habitats. Depth and contour were manipulated to create seven temporary and three permanent fish-free ponds that provide suitable habitat for all pond-breeding amphibians at Tulula. At seven sites small standing water habitats existed prior to the construction of ponds. We selected 10 of the lazgest existing breeding sites as reference ponds to compare hydrological, physiochemical, and biotic characteristics. One reference pond was destroyed in 2001 in conjunction with reconstruction of the stream channel. Two others did not fill in 2001=2002 due to construction activity, but were functional in 2003 and 2004. Thirteen new breeding sites were also created in the fall of 1999 when golf course ponds were either filled or partially filled to create shallow ponds. Most of these were stream-fed, and now exist as shallow, permanent sites that contain small fish. In others, fish were eliminated and the sites were converted into temporary ponds. Sections of the restored stream channel also were temporarily blocked with check dams to allow channel re-vegetation prior to restoring stream flow. Small pools formed in the deepest sections of these channel segments and were used as breeding sites by resident amphibians in 2001. Additional pools were formed in conjunction with stream and site restoration in 2001-2003. In February 2004 the site had over 60 breeding sites (Fig. 12). Fig. 12. Location of standing water habitats within the study site (spring 2004). 44 Methods The 10 constructed and 10 reference ponds were sampled 3-19 times annually to obtained data on pond pH, temperature, conductivity, and oxygen saturation. Samples were taken during the day (900-1700 hrs) and all constructed and reference ponds were sampied haphazardly during the same day. Three subsamples of water were taken from each pond at approximately equidistant points along the center of the long axis and approximately 10 cm below the water's surface. Subsamples were pooled and readings were taken from the pooled sample. Samples were placed on ice during warm weather and dissolved oxygen was measured in the field < 3 hours after samples were collected using Corning Check-mate meters: Conductivity and pH were measured using Corning Check-mate and Corning 430 bench meters, respectively. We used the yearly mean for all seasonal samples in statistical comparisons of reference and constructed ponds. Results Reference ponds were smaller and shallower than constructed ponds, which could influence physiochemical characteristics. At full capacity, surface areas of reference ponds averaged 888 ftz (range = 145-2367 ft2) versus 5165 ft2 (range = 2421-9931 ftZ) for constructed ponds. Respective values for maximum depths were 13.4 inches (range = 5.1-23.6 inches) and 24.4 inches (range = 15-34 inches). Comparisons of physiochemical characteristics of constructed and reference ponds from 1996-2004 are in Fig. 13. � 70 60 50 40 30 20 10 0 s.2 s 5.8 5.6 5.4 5.2 5 4.8 Conductivity (mS/cm) 1996 1997 1998 1999 2000 2001 2002 2003 2004 � 1996 1997 1998 1999 2000 2001 2002 2003 2004 Temperature (C) 25 20 . . 15 , 10 5 0 too so eo 70 60 50 40 30 20 1996 1997 1998 1999 2000 2001 2003 2004 Oxygen saturation (%) • • . 1896 1997 1998 1989 2000 2001 2002 2003 2004 -�-r�fere�ioe +oonstrt�ded Fig. 13. Physiochemical characteristics of reference and constructed ponds. Symbols are annual means based on 3- 19 seasonal samples per year. Vertical bars are 1 SE. Asterisks indicate means that differed significantly within years. � 45 Respective grand means (+ 1 SE) based on annual averages for reference versus constructed ponds were 5.46 (0.08) versus 5.60 (0.05) for pH, 14.4°C (0.53) versus 17.1°C (0.86) for temperature, 42.1 (2.09) versus 38.3 (1.60) dS/cm for conductivity, and 58.8 (3.3) versus 77.8 (3.6) for percent OZ saturation. T-tests (alpha = 0.05) indicate that means for pH differed only in 2002 and 2004, while conductivity did not differ significantly for any year (conductivity: P> 0.19). However, constructed ponds were significantly warmer in five of seven years and had significantly higher oxygen saturation levels in all but two years. � 2. Use of constructed and reference ponds by amphibians. Methods All constructed ponds filled with water before amphibians began breeding in February 1996. We monitored all constructed and reference ponds annually to determine patterns of use by resident species. We visited ponds every 1 to 3 weeks between January-August and searched for amplexed adults, eggs, or larvae. Larvae were collected when conducting open-bottom sampling to estimate survival (see below) and when ponds were dip-netted periodically during the spring and summer to sample resident amphibians. Results Resident amphibians rapidly colonized constructed ponds that first filled in 1996 (Fig. 14). Eight species of amphibians bred in the constructed ponds within 1 year of construction and 10 species have used the ponds through 2004. These are the wood frog, green frog, bullfrog, gray treefrog, spring peeper, American toad, spotted salamander, red salamander, three-lined salamander, and the red-spotted newt (Appendix F). The only species unique to constructed ponds was the bullfrog, which prefers permanent or semipermanent habitats. Reference ponds were also used by 10 species of amphibians and only one, the two-lined salamander, was unique to reference ponds (breeding, in 1 of 10 reference ponds). Overall, constructed ponds contained a significantly greater number of breeding species (mean + 1 SE = 4.21 + 0.24 species) than reference ponds (2.74 + 0.16 species) during the 8-year period (paired t-test; P= 0.0002). For individual years, the mean number of species per pond was significantly higher in constructed ponds for five of eight years and approached sigriificance (P < 0.10) for two other years (Fig. 14). Regression analysis indicates that the mean number of species using ponds annually did not increase between 1996-2003 (P values for reference and constructed ponds = 0.92 and 0.19, respectively). The latter suggests that constructed ponds quickly reached saturation levels within one year of construction. A more detailed analysis of pond colonization and community turnover is in Petranka (2000a). ., 6 * � * ,* * 5 „ � a� .� 4 a� � � � 0 3. a� � � � 2 � z , 1 U •ti� •• •�� •�• 1�1 �� —a— C1ef21'@� —�- COf1S�'1.1C�d Fig. 14. Mean number of species that bred in reference and constructed ponds. Symbols are means and bars are + 1 SE. Years with asterisks are significantly different. 3. Response of focal species to constructed ponds. Methods We selected the spotted salamander (Ambystoma maculatum) and wood frog (Rana sylvatica) as focal species for monitoring ecosystem function and restoration success. Both species are widely distributed across the site and are largely, restricted to temporary ponds that � predominated prior to golf course construction. These species lay large egg masses that can be accurately counted, and that serve as an index of the size of the female breeding population. To obtain estimates of the overall response of the focal species to restoration efforts, we conducted a complete count of egg masses on the eastern half of the site beginning in 1995. This census included the 10 constructed ponds, the reference ponds, and all other breeding sites in the eastern sector. ' . To estimate relative changes in embryonic and larval survival across years, we estimated the total population size of hatchlings and larvae nearing metamorphosis in each pond using open-bottomed samplers. Populations were sampled using 30 gallon galvanized trashcans with bottoms that were removed with a blowtorch (approximate'area of can bottom = 1.2 ft2). When sampling, the can was pushed into the pond substrate to trap larvae. Repeated sweeps of the can were made with aquarium nets until no larvae were captured for five consecutive sweeps. 47 Ponds were sampled by walking a zig-zag transect across the entire area of the pond and taking samples at approximately equidistant points along the transect. The number of samples per pond increased with pond size and varied from 15-80. If ponds were not at full capacity, then pond surface area was estimated at the time of sampling based on 3-5 measurements of length and width using a meter tape. The total population size of hatchlings or larvae nearing metamorphosis was estimated using data on the mean number of larvae per sample, the surface area of the sampler, and the surface area of the pond. We obtained an initial sample of hatchlings within 1-3 weeks after > 95% of the egg masses were estimated to have hatched in a pond. We intensively dip-netted ponds as larvae approached metamorphosis, and obtained a final sample immediately after the first metamorphosing larva was observed in each pond. Criteria used to recognize metamorphosing larvae were the emergence of both front legs for wood frog tadpoles and the partial or complete reabsorption of gills and dorsal fins for spotted salamander larvae. We used this estimate as a relative measure of the number of juveniles that were recruited into the terrestrial population each year. Changes in adult population size are the most meaningful measure of the response of amphibians to site restoration efforts. However, a significant time lag in population responses occurs because of the prolonged juvenile stage. That is, juveniles that metamorphose and leave ponds may not return for 2-4 years as breeding adults. We used total egg mass censuses of the eastern half of the site to measure the effects of pond construction and site restoration on breeding populations. Results The responses of breeding populations of wood frogs and spotted salamanders to pond construction are shown in Fig. 15. These data exclude two constructed ponds (7X; l OX) that occurred on the western end of the site and three small reference ponds that were either destroyed (2C) or were nonfunctional in 2002 (3C; 4C) and 2003 (4C only) due to construction activities. During 1996 (first year after pond construction and filling), 71 % of the resident wood frogs and 59% of spotted salamanders bred in the constructed ponds. A corresponding decline in breeding effort occurred in the remaining small depressions, suggesting that many adults abandoned historical breeding sites in favor of newly constructed ponds. The percentage of adult wood frogs that bred in constructed ponds between 1996 and 1999 increased slightly. However, adults decreased use of constructed ponds after 1999 and shifted to other sites. This reflects a progressive increase in the number of ponds on site in association with stream and final site reconstruction. In contrast, use of constructed ponds by spotted salamanders was similar across years, perhaps because adults favor larger, deeper ponds for breeding. In 2004, approximately 48% of wood frogs and 44% of spotted salamanders bred in the constructed ponds, while reference ponds provided breeding habitat for < 8% of the population. . ioo � � 80 \ -0 60 ro J `� 40 a� � a, 20 o+ w 0 1995 1997 1999 2001 2003 so ro 0 60 \ � J 4O a � 20 rn rn w 0 SpottecJ Salamander 1995 1997 1999 2001 2003 -�-- construded -�- reference f other sites � r,onstructed -�- reference -,� other sites 48 Fig. 15. Response of female wood frog and spotted salamanders to pond construction. Symbols are the number of egg masses laid on the eastern half of the site in constructed ponds, reference ponds, and all remaining breeding sites. Numbers are expressed as a percentage of all masses laid in the eastern half of the site. `Other" includes all sites other than reference and constructed ponds, including sites that were created during stream channel restoration. Data for 1995 `constructed' are masses laid in preexisting sites where ponds were constructed. Fig. 16 shows annual changes in the percentage of ponds that successfully produced juveniles (upper graphs) and total yearly output of juveniles from constructed and reference ponds (lower graphs). The percentage of ponds that successfully produced juveniles has declined annual from 60-100% in 1996 to < 30% in 2003. The estimated output of terrestrial juveniles from constructed ponds was exceptionally high during 1996 (N = 253,696 wood frogs; 30,831 spotted salamanders), but progressively declined in later years (e.g., N= 5,819 and 753 in 2003, respectively). A similar trend occurred in reference ponds. � These trends parallel a general decline in the percentage of ponds that have successfully produced juveniles each year. Nonetheless, a small percentage of ponds on site have successfully produced juveniles annually, and viable populations of both species occur on site (see below). Comparisons of the number of hatchlings and number of larvae surviving to the initiation of inetamorphosis (see Petranka 2003b for details) indicate that the decline in juvenile output was primarily due to increased larval mortality rather than increased embryonic mortality. Embryonic survival varied among years, but there was no evidence of catastrophic mortality for any year. In contrast, overall juvenile production per egg mass declined; markedly during the study period for both species and both sets of ponds. The reduction in juvenile production is attributable to at least three factors: (1) premature pond drying and/or the failure of ponds to fill seasonally, (2) outbreaks of a pathogen that caused larval die-offs, and (3)'the accumulation of predators in constructed ponds after 1996. 49 100 � N 7 � .� � � � � O O_ � C � � a� d 0 300 250 v ij 200 � v ° 150 n m 100 'c � � � � !0 vwod Frog 1996 1997 1998 1999 2000 2001 2002 ZOai -o-- refererne t ca�n�ct� Wood fiog 100 y N � � �1 .� .� 7 � O � d Cy V 'XJ a`� a 0 � � v �j 25 � o � °' 15 m 'c �� 5 � 0 � s�oaea s�,�,der 1996 1997 1998 1999 Z000 2001 2002 2003 � �feience t corstructed Spotted SalarrerrJer 1996 1997 1998 1999 2000 �2001 2002 20IX3 1998 t99'7 1998 1999 2000 200� ' 2002 2003 _ �-refe�anCetto15SR1Cted � -�referer�ce-�-oonsWCted , Fig. 16. Estimates of the percentage of ponds that produced juveniles, and total juvenile recruitment from 10 constructed and 10 reference ponds during 1996-2003. Symbols for upper panels are the percentage of ponds that produced juveniles annually, whereas those in the lower panels are the estimated number of larvae surviving to the initiation of inetamorphosis (in thousands). Fig. 17 shows the percentage of ponds that either did not fill or that filled and dried prematurely between 1996-2003. Constructed ponds filled annually and usually held water sufficiently long to allow metamorphosis of both species. An exception is 2001 when 20% of ponds dried prematurely, causing catastrophic mortality. The more shallow reference ponds tended to progressively deteriorate with respect to hydroperiod between 1996-2002. During 2002, 43% and 100% of the reference ponds either did not fill or dried prematurely for Rana and Ambystoma, respectively. This pattern may in part reflect a regional drought that occurred from the summer 1998 to fa112002. The proportion of reference ponds that dried prematurely decreased after 2001-2002 as the drought ended and rainfall increased to average or above average levels. 80 70 60 �, `.�0 � 40 � 30 a`� � 20 10 0 , -10 e �-- Wood Fmg 110 90 � � O � U U1 � � 10 -10 i Spotted Salarrander i 50 1996 1997 1998 1999 2000 2001 2002 2003 1996 1997 1998 1999 2000 2001 2002 2003 —o--ieference f crostnicted ��r�erence —a— constructed Fig. 17. Annual variation in the percentage of constructed and reference ponds that either did not fill or that dried before larvae could initiate metamorphosis, Disease is a second factor that contributed strongly to the decrease in juvenile output between 1996-2003. Outbreaks of a disease that caused catastrophic larval mortality were first observed in 1997. Moribund specimens were sent to the National Wildlife Health Center in Madison, Wisconsin, and detailed histological and molecul'ar studies revealed-that the pathogen is an iridovirus (Ranavirus). , Larvae of both the wood frog and spotted salamander are susceptible to Ranavirus infections. Infected larvae tend to become lethargic, often float at or near the water surface, and develop characteristic bloody, hemorrhagic patches on the body and fins. Infected larvae are first noticed seasonally during the mid- to latter half of the larval stage. Catastrophic mortality typically occurs within 1-2 weeks after the first infected individuals are detected. Typically, outbreaks result in 100% mortality of larvae in a pond. � The extent to which the disease has impacted local populations in reference and constructed ponds at Tulula is shown in Fig. 18. Diseased animals and die-offs were not observed prior to 1997, at which time two die-offs occurred in two ponds. The disease rapidly spread to other ponds on site and has been a major source of larval mortality since 1998. The smaller percentage of reference ponds with die-offs between 1998-2002 reflects the fact that many reference ponds dried prematurely (e.g., prior to the time when the disease normally develops , 100 90 � � 0 8, 70 v � � d � rn o qp U � N a Zp 10 0 Vtibod Frog 1996 1997 1998 1999 2000 2001 2002 2003 70 60 � �� v � 40 m �� U � N a 10 0 SpotG:d Salamarder 1996 1997 1998 1999 2000 2001 2002 2003 51 --�refe2nce tcorstruciedl --o—reference —�constructed Fig. 18. Changes in the percentage of reference and constructed ponds in which catastrophic die- offs of larvae occurred from Ranavirus infections. Egg and larval predation was the third significant source of premetamorphic mortality that contributed to the decline in juvenile output between 1996-2003. In particular, egg predation by green frog tadpoles on wood frogs (Petranka and Kennedy 1999), and wood frog tadpoles on spotted salamanders (Petranka et al. 1998) were significant sources of mortality in certain ponds. Odonates and other predatory aquatic insects accumulated in constructed ponds after 1996 and presumably contributed to higher larval mortality. Despite impacts from drought, disease, and predators, populations of both species have not suffered severe crashes and remain at viable levels (Fig. 19). The size of the wood frog population declined from 1995-1998, increased dramatically (366%) through 2000; and declined thereafter. The population has remained relatively stable since 2002. Female wood frogs require 3-4 years to reach sexual maturity after metamorphosing (Bervin 1982). Thus, the marked increase in population size in 1999 corresponds to when the large output of juveniles in 1996 first returned to breed as adults. The decline since 2000 presumably reflects the impact of Ranavirus and premature pond drying on the adult population. The population of spotted salamanders has not changed as markedly. The size of the breeding population slowly increased from 1995(N = 1,265 egg masses) to 2004 (N = 1,831 masses). Females of this species may require 3-5 years to reach sexual maturity (Petranka 1998), so the gradual increase in breeding population size may reflect recruitment from the relatively large output of juveniles in 1996 and 1997. The decline in 2002 may reflect the impact of Ranavirus outbreaks that began in 1997-1998. However, in 2004 the population reached the highest level (1,831 masses), indicating that recruitment has been sufficient to gradually increase population size. N N 3000 � 2500 p� 2000 W 1500 � 1000 � 500 � 0 Z 1995 1997 1999 2001 2003 --�- reference --.-- all ponds � construcbed wood frog (east of Mason's) ,� � �� i � � � �� ,� i i �� ' � �. � i �"-y---• n `� ` � � 2000 � � 1500 � W 1000 O � 500 � � 0 Z 1995 '1997 1999 2001 2003 �— construc�d ---o— reference --+-- all ponds spotted salamander ' ,.� ' ' ��.,+---+'� �� _ /�'-- �.. `�. � ��' � 52 Fig. 19. Changes in adult breeding population size on the eastern sector based on annual egg mass counts in all breeding sites. � 4. Altered site hydrology and emerging concerns. The completion of reconstruction activities, above average precipitation in 2003-2004, and invasions of the site by beavers have increased the number of habitats with fish. Damming of Tulula Creek by beavers caused spillover into most of the nearby wetlands that parallel the stream on the west end of the site (Fig. 12). Almost all of these sites now contain fish and provide little habitat for seasonal pond breeders. Although reference ponds are too ephemeral to support fish, fish have invaded many of the constructed ponds since 2002 (Fig. 20). Amphibians that use fish-free habitats have responded by not ovipositing in ponds with fish; however, it is uncertain whether adults that avoid ponds with fish are successfully breeding in other habitats on site. " � N L � � � c 0 Q 0 n� � � c a� U � a 90 80 %� 60 50 40 30 20 io 0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Fig. 20. Yearly changes in the percentage of the ten constnicted ponds that contained fish. 53 Summary Data collected from 1996-2004 indicate that constructed ponds are of higher quality than reference ponds based on physiochemical characteristics, seasonal hydroperiod, and use by resident amphibians. The constructed ponds tended to be warmer and have higher oxygen levels Since larval growth is directly proportional to temperature, and high oxygen levels reduce physiological stress, physiochemical conditions are judged to be superior to those of reference ponds. Amphibians rapidly colonized the constructed ponds, and the number of species that � utilize these as breeding sites averaged about 50% higher than that of reference ponds. Reference ponds progressively deteriorated between 1996 and 2002 with respect to seasonal hydroperiod. In 2002 the majority either did not fill or dried prematurely, resulting in catastrophic mortality of pond populations. In contrast, the hydroperiod of most constructed ponds appears to be adequate for most vernal pond breeders. Seven of 10 ponds normally undergo seasonal drying in late summer or fall when larvae have metamorphosed. However, fish have colonized many since 2002 in association with above normal rainfall, beaver activity, and completion of the final phase of reconstruction. Outbreaks of Ranavirus have dramatically reduced the output of juveniles from both constructed and reference ponds. Similar outbreaks of this disease have been reported in several areas of the United States (Daszak et al. 1999) and have resulted in catastrophic die-offs of larvae. Amphibians often exhibit boom-and-bust recruitment patterns in which juvenile recruitment may be near zero in some years and high in others (e.g., Gill 1978, Semlitsch et al. 1996). Local populations are buffered from these effects since the adults may live many years and metapopulation dynamics allow for some recruitment annually. Thus, years with complete reproductive failure in local ponds may not necessarily translate to long-term declines of local � populations. We have documented high rates of reproductive failure in most ponds in most . years. However, annual recruitment from a small subset of ponds annually appears to be sufficient to maintain viable adult populations of wood frogs and spotted salamanders. Scientists currently know very little about the epidemiology of amphibian Ranavirus. For example, it is unknown how the virus is spread between ponds, whether a subset of larvae are resistant to the virus, or whether the infections subside after several years of outbreaks. Preliminary studies that we have conducted suggest that humans and other vertebrates such as raccoons and birds may play a role in spreading the disease via movement of contaminated mud or water between local ponds. One scenario for the Tulula populations is that the severity of die- offs will decline with time as local populations evolve immunity or as the virus undergoes normal erratic patterns of outbreak. A second is that the virus will consistently produce annual die-offs in most ponds that do not dry prematurely. If the proportion of ponds that suffer die-offs increases significantly in the future,. then the latter could result in resident amphibian species undergoing population bottlenecks or even local extinctions. i • � • • 54 • The invasion of beavers (Castor canadensis) and the completion of stream restoration are • influencing site hydrology and the dynamics of amphibian populations at Tulula. Beaver invaded the site shortly before stream channel construction began and were eliminated through trapping. They have since � reinvaded and have significantly altered the landscape. Fish have become far more abundant on site since • 2002 and have invaded most of the constructed ponds. In general, habitat quality for amphibians that use seasonal wetlands has declined. Monitoring of focal species in fiiture years will document how � amphibians respond to altered hydrology from stream restoration;and beaver activity. It will also help • resolve the extent to which Ranavirus infections ultimately impact breeding populations of amphibians. I� D. Bird Use of Tulula ' �� Birds are used as a common indicator for assessing changes in habitat attributes that are associated • with many types of restoration projects (Morrison 1986). Since 1994, we have conducted breeding bird � surveys and measured habitat characteristics of the Tulula floodplain (Rossell et al. 1999, Moorhead et al. 2001). Restoration of Tulula Creek was completed during the summer of 2002. Here we report results of � breeding bird surveys and habitat analyses conducted during 2004. These results are the first year of data • evaluating the response of bird populations to post-restoration habitat changes at Tulula. � 1. Bird Surveys I Methods � Breeding bird surveys were conducted from 17 May to 29 May 2004, at 65, 25-m radius plots • located across Tulula floodplain (Fig. 21). Thirty-two plots were separated by at least� 100 m. An . additiona133 plots were separated by at least 50 m and surveyed because habitat data have been collected at these plots since 1994 (see Bird-Habitat Relations below). Surveys were conducted from sunrise until � 1000 hrs. After a 1-min quiet time, all birds heazd or seen withiri 25 m of the plot center were recorded for • 3 min. Birds that flushed within 25 m of the plot center during tlie approach also were recorded. Plots • were sampled three times during the survey period. Bird richness was defined as the total number of species, and relative bird abundance was defined as the total number of individuals of a species. Results and Discussion Results of breeding bird surveys are presented in Table 22. In 2004, species richness declined 15% from 20021evels, with 33 species recorded. American Woodcock, Common Grackle, and Eastern Wood-pewee were new species recorded during surveys (See Appendix C for complete list of birds and scientiiic names). Common Grackle and Eastern Wood-pewee are common in the mountains of North Carolina (Hamel 1992)!, and both species were likely breeding on site. American Woodcock aze considered rare iin the southern Appalachians, although they have no designated conservation status (Hamel 1992). The American Woodcock is associated with moist woodland thickets and bottomland forests that have an abundance of dead leaves on the ground (Hamel 1992). American Woodcock have been observed in past years using the Tulula floodplain for singing grounds; this species likely breeds in low numbers throughout the site. 55 C Vegetation, 1998 0 grass shrubs forest � water 200 0 200 400 600 800 1000 1200 Meters Fig. 21. Location of bird survey and habitat plots. S= survey plots, H= habitat plots, and B= survey and habitat plots. � Relative bird abundance in 2004 decreased 52% from 20021evels, with a 166 total observations (Table 22). Song Sparrow and Rufous-sided Towhee continued to be the most abundant species on site, however, their numbers decreased by almost 50% from 20021evels. Red-winged Blackbird also continued to be one of the most abundant species on site, but its numbers held steady relative to 20021evels. Many species of conservation concern declined substantially in 2004 (Hamel 1992). The most notable declines included the Golden-winged Warbler, Hooded Warbler, and Yellow-breasted Chat. Golden-winged Warblers and Yellow- breasted Chats have declined steadily since 1998. Other species that declined in 2004 included Red-eyed Vireo and White-eyed Vireo. Brown-headed Cowbirds, which were breeding at Tulula in 2002, were conspicuously absent in 2004. � The declines in species richness and relative bird abundance are likely associated with the large proportion of the floodplain that was inundated with standing water. Beaver have colonized the western end of Tulula Creek, constructing a series of dams that flooded much of the interior of the'site. The site was so wet during the spring of 2004 that chest waders had to be worn to conduct surveys. Species associated with standing water, such as Red-winged Blackbirds and Wood Ducks, have generally increased in abundance, while species associated with early-successional habitats, including many of the Neotropical migrants of conservation concern, have generally decreased in abundance. - � 56 The Golden-winged Warbler is the species of highest conservation concern breeding at Tulula. This species is federally listed as a species of special concern (LeGrand and Ha112004). Since 1994, the Golden-winged Warbler has decreased 94% (31 to 2 birds) in breeding bird surveys at Tulula. Golden-winged Warblers require a varie'ty of seral stages for breeding, including patches of herbaceous cover, shrub thickets, and a forested edge (Klaus and Buehler 2000, Rosse112001, Rossell et. al. 2002). As a result of stream construction and backfilling the old stream channel during the spring of 2002, most of the h'erb and shrub layers were eliminated from the interior of Tulula. This area encompassed a substantial portion of many Golden-winged Warbler territories (Rossell et al. 2002). In 2004, additional habitat was lost due to the flooding of the site by beaver. i � General observations of Golden-winged Warblers at Tulula indicated that 6-8 territories were established in 2004. The majority of territories were located along the periphery of the floodplain where conditions were drier and where there wa's a large shrub component. Areas with large amounts of standing water were generally not inhabited by Golden-winged Warblers. Interestingly however, all Golden-winged Warbler territories established in 2004 contained some standing water. Table 22. Relative abundance and migratory status of birds recorded during breeding bird surveys in 65, 25-m radius (0.2 ha) plots during 1994, 1998, 2000, 2002, and 2004. 1994 1998 2000 2002 2004 Migratory Species � Number , Status* Acadian Flycatcher American Goldfinch American Robin American Woodcock Belted Kingfisher Blue-gray Gnatcatcher Blue-headed Vireo Brown-headed Cowbird Brown Thrasher Black-and-White Warbler Blue Jay Carolina Chickadee Carolina Wren Common Yellowthroat Chestnut-sided Warbler Cedar Waxwing Common Grackle Downy Woodpecker Eastern Phoebe Eastern Wood-Pewee 2 19 0 0 0 11 0 0 1 0 15 3 7 23 9 0 6 0 0 14 13 1 0 1 13 0 0 0 3 2 4 6 1 2 10 0 1 0 0 3 7 0 0 0 10 0 0 0 0 7 3 0 4 0 2 0 0 1 �5 � 12 I 0 '0 �9 �1 �`2 �4 �0 '0 10 2 �2 ; 14 9 '0 3 '1 '0 i � 5 2 1 1 0 11 0 '0 1 0 8 7 3 0 1 2 0 1 N Y D D. Y N N D D Y Y Y N D Y Y D N 57 Golden-winged Warbler Gray Catbird Hooded Warbler Indigo Bunting Kentucky Warbler Mourning Dove Northern Bobwhite Quail Northern Cardinal Northern Flicker Northern Parula Northern Rough-winged Swallow Ovenbird Pileated Woodpecker Red-eyed Vireo Ruby-throated Hummingbird Rufous-sided Towhee Red-winged Blackbird Scarlet Tanager Song Sparrow Swainson's Warbler Tufted Titmouse � White-breasted Nuthatch White-eyed Vireo Wood Duck Wood Thrush Yellow-breasted Chat Yellow-throated Vireo Yellow-throated Warbler Yellow Warbler Total Species Total Individuals 31 4 11 83 17 0 0 8 1 17 0 2 0 21 6 22 0 0 4 1 3 1 22 0 0 18 4 3 0 21 0 21 55 9 2 0 3 0 24 2 6 2 28 5 24 0 1 11 4 5 0 26 0 1 23 1 4 1 8 0 6 15 9 0 2 4 0 10 0 2 1 28 6 14 0 1 11 0 8 1 29 0 0 12 3 1 0 6 0 12 17 2 1 7 12 1 26 4 5 2 25 7 26 13 0 31 0 11 1 20 1 3 7 3 3 0 2 0 4 13 9 0 1 5 0 11 0 0 1 10 3 15 12 0 16 0 5 1 3 2 1 1 0 0 0 31 36 29 39 33 378 350 215 321 166 N Y N N N Y Y Y Y N N N Y N N Y D N Y N Y Y N D N N N N N *Note: Migratory status from Hamel (1992). N= Neotropical migrant, D= Short-distance migrant, Y= Year-round resident. 58 2. Bird-Habitat Relations Methods Habitat data were collected in 41, 25-m radius (0.2 lia) permanent plots from 7 June to 28 June 2004. Bird-habitat plots were selected in 1994 based on the criterion that they had at least one bird species recorded in two out of three surveys. Within each plot, herbaceous cover, shrub thickness, and canopy cover were estimated at 16 regularly spaced points along two perpendicular transects. Understory (2.5-10 cm dbh) and overstory (> 10 cm dbh) tree densities were also estimated in each plot using the closest individual method (Bonham 1989). Herbaceous cover was estimated for vegetation < 0.5 m in lieight using a 0.25-m2 quadrat. Shrub thickness was estimated for vegetation 0.5-2 m tall using a shrub profile board (Hays et al. 1981), Canopy cover was estimated using a spherical densiometer (Hays et al. 1981). � Bird richness and relative bird abundance were calculated for each plot. Cedar . Waxwings and American Goldfinches were excluded from ;the analysis because their flocking i behavior tended to inflate estimates. Correlation analysis was used to examine associations between the habitat variables and bird richness and relative bird abundance. Analysis of variance � (ANOVA) tests were used to compare differences among years for bird richness, relative bird � abundance, and the habitat variables. If a significant difference was found wit� ANOVA, then • Tukey's Studentized Range test was used to determine between year differences. � r Results and Discussion � � I ' • Means of bird richness, relative bird abundance, and habitat variables for the 41 habitat • plots are summarized in Table 23. Both bird richness and relative bird abundance were significantly lower in 2004 than in 2002 (P < 0.05). In �2004, herbaceous cover was significantly � greater than in 2002 (P < 0.05), while all other habitat variables were similar between the two � years (all P> 0.05). There was a significant negative correlation between relative bird abundance • and overstory tree density (r =-0.14, P= 0.04). A similar relationship was evident between bird • richness and overstory tree density, although the correlation was not statistically significant (r =- 0.12, P= 0.09). All other correlations between bird richness or relative bird abundance and the � habitat variables were extremely low (all Pearson r, between -0.07 and 0.05; all P> 0.05): 59 Table 23. Means (SD) of bird richness, relative bird abundance, and habitat variables for 41, 25-m radius (0.2 ha) plots during 1994, 1998, 2000, 2002, and 2004. Variable Year 1994 1998 2000 2002 2004 Bird Richness 4.6 (2.1)b 4.0 (1.8)b 2.8 (1.9)a 3.7 (2.2)b 1.8 (1.9)a Rel. Bird Abund. 6.6 (3.0)a 5.2 (2.8)a 3.4 (23)ab 4.4 (2.7)a 2.2 (1.6)b Herb. Cov. (%) 60.0 (17.5)a 53.9 (20.6)a 52.4 (17.9)a 28.1 (15.6)b 48.5 (18.7)a Shrub Thick. (%) 35.2 (15.9)ab 28.5 (14.7)b 38.9 (17.7)a 25.9 (16.7)ab 32.6 (12.0)b Canopy Cov. (%) 59.2 (23.8) 45.4 (21.8) 51.7 (25.0) 45.6 (26.5) 47.4 (26.0) Understory dens. (no./0.2 ha) 11.5 (15.3) 6.3 (18.8) 21.7 (27.1) 18.5 (30.2) 22.2 (31.0) Overstory dens. (no./0.2 ha) 7.1 (13.9)a 7.6 (13.8)a 10.8 (20.5)ab 8.9 (16.0)ab 21.8 (40.1)b Note: Values followed by the same or no letters within a row are not significantly different (P > 0.05 ). The negative trends in bird richness and relative bird abundance in the habitat plots support the results of the breeding bird surveys. As discussed in the Results and Discussion of the Bird Survey section of this report, the declines in bird richness and relative bird abundance are related to loss of habitat due the large proportion of the site with standing water. In addition, the negative correlations found between bird richness and relative bird abundance and o��erstory tree density also help to explain the declines in species that require early-successional habitats. These declines in early-successional species are likely to continue as succession proceeds and overstory tree densities increase across the site. In 2002, significant reductions in herbaceous cover and shrub thickness reflected high levels of disturbance of the interior of Tulula that occurred during restoration activities. These habitat changes were accompanied by significant increases in bird richness and relative bird abundance as a result of generalist species colonizing the site. Many of the generalist species � that experienced large increases in 2002, such as the American Robin, Rufous-sided Towhee, and Song Sparrow, declined dramatically in 2004 as a result of the site being flooded by beaver (Table 22). The significant increase in herbaceous cover in 2004 compared to 2002 reflects the large increase in areas with standing water colonized by sedges and rushes. Observations during surveys indicated that few bird species use this rush/sedge dominated habitat, with the exception of a few blackbirds and wood ducks. Bird surveys and habitat analyses are scheduled for 2006 to continue monitoring the responses of bird populations to post-restoration habitat changes. Results reported here indicate that some type of management is needed at Tulula to maintain the productivity of the habitat for birds (especially the habitat of the interior of the site). Management objectives should include taking appropriate actions to eradicate beaver or control the flooding caused by beaver, and maintaining a variety of eazly-successional habitat types. .1 DISCUSSION Tulula continues to change as restoration proceeds and as natural processes respond to changing site conditions. We have developed a fairly comprehensive understanding of annual and seasonal variability in the structural and functional attributes of this restoration project. The overall pattern of the restored stream channel has not changed since water was released in the first restored section in September 2001. We have noticed isolated areas of bank and bed erosion, but the channel is performing remarkably well after two years of water flow. Most of the notable areas of bank erosion are associated with beaver dams at the lower end of the site in the past year. Banks have eroded behind the dams and at the point of water entry back into the restored channel. After two years of water flow, bankfull widths did not change. However, we anticipate an increase in erosion problems during subsequent channel surveys in the middle and the lower end of the site due to beaver activity. Most of the channel has degraded to some extent, as noted by increasing cross-sectional areas of riffles and pools. � � Changes in site hydrology were noted through the network of manual and electronic � wells. The data from manual wells provide a more comprehensive understanding of changes in � hydrology. Differences that suggest an improvement in hydrology (rise in water table depth) • were observed with manual wells located in the floodplain close to the restored channel. Restoration of Tulula Creek did not appear to influence the hydrology of the fen. The assessment � of site hydrology has been made more difficult from overbanking of water from beaver dams at � the lower end of the site. � � Restoration has had a clear influence on the composition of some plant communities. For � example, restored wetland areas contained fewer species than unrestored areas or restored dry • areas, and the species that dominated the restored wet areas were OBL and FACW plants. In • restored areas that did not benefit from hydrologic change, soil disturbance led to a great increase in species richness. The production of both vegetative and reproductive stems of a common � rush was influenced by both restoration and hydrologic change. � � � � � � Natural succession continued to change the composition of wetland plant communities across the site, increasing the number and dominance of woody species. Although red maple continued to dominate the canopy of a forested fen that we� have studied since 1994, the importance of white pine in the canopy increased. Had the water table in the fen increased after restoration, this'increase in white pine may not have occurred. Red maple saplings continued to flourish in open areas of the site, with naturally-regeneratirig saplings outperforming nursery- stock saplings that were planted in 1995. � We found that the least disturbed plant communities (particularly red maple) had the � quickest decomposition, the greatest amount of litter microarthropods, the most soil organic � carbon, and the lowest soil pH. We conclude that soil characteristics related to low disturbance, • rather than the presence of a closed canopy, probably have;the greatest influence on speed of decomposition and numbers of litter microarthropods. The'refore, the most intact ecosystems � appear to be functioning at the healthiest levels. � � � � � � � 61 Researchers have rarely conducted long-term studies of vertebrates to determine changes in community assemblages, or to document population resilience and viability following the creation of breeding habitats at restoration sites. This information is critical for setting realistic time frames and criteria for assessing restoration success. Our project is yielding important information concerning the long-term dynamics of restored wetlands and the response of resident amphibians to environmental perturbations. The data will be useful in designing more meaningful assessment criteria for future restoration projects throughout the eastern U.S. To enhance population resilience, we used a metapopulation design that involved the creation of a large array of breeding sites that differed in size, depth, and hydroperiod. Many ponds were unproductive in certain years because of premature drying associated with drought, the outbreak of a deadly viral disease, and the invasion of ponds by fish due to beaver activity and site reconstruction. Despite catastrophic mortality in most ponds each year, a few ponds have produced juveniles annually, and total juvenile output has been sufficient to maintain viable adult populations of resident species. The extent to which Ranavirus and predatory fish will ultimately impact adult populations of the wood frog and spotted salamander will become more clear as we continue to track changes in breeding population size in 2004 and beyond. However, the data to date indicate the wetland design at Tulula has resulted in populations that are resilient to major site perturbations associated with drought, disease, and predators. The significant declines in bird species richness and relative bird abundance in 2004 were associated with substantial amounts of standirig water in the interior of the floodplain due to flooding by beaver. Habitat productivity for birds in these areas generally decreased as the structure of the habitat became more homogenous, being dominated by a thick cover of low- growing rushes and sedges. Most affected by the flooding were species that require early- successional habitats; most notably many Neotropical migrants of conservation concern including the Golden-winged Warbler, Chestnut-sided Warbler, Hooded Warbler, and Yellow- breasted Chat. Tulula is the first wetlands mitigation bank in the Blue Ridge Province of North Carolina. Most mitigation banks in North Carolina aze located in the Coastal Plain, a�d are considerably different from Tulula in terms of their hydrology and ecology. Our database on hydrology, soils, flora, and fauna continues to provide a framework for documenting the success of restoration at Tulula. These data were important in the development and design of restoration strategies, and have influenced considerations for site management. Tulula has provided research experience to more than 50 undergraduates at UNCA, including numerous senior research projects. 62 RECOMMENDATIONS 1. Beaver dams are exerting a localized but significant impact on bank erosion of the restored channel and are influencing the water table of nearby wetlands. Evaluating the geomorphology and stability of the restored stream channel and site hydrology will require controlling of beaver activity. 2. Eradicating or controlling beaver activity will also maintain or restore the productivity of the habitat for amphibians and birds by controlling or eliminating flooding in key portions of the site. 3. Future site management of the Tulula site should include efforts to retain portions of the site in an early successional stage (for example, using bush-hogging or burning). This would enhance habitat for small but unique plants that are presently being outcompeted, and for uncommon animals such as the Golden-winged Warbler. 4. Monitoring of floral and faunal communities at Tulula should continue, to document how they respond to the hydrologic changes. Ideally, a long-term monitoring program should be developed to gauge the success of the wetlands restoration over a decade or longer. 5. As the site re-establishes as a wetland, it should be monitored for the presence of in��asive plant species such as cattail (Typha latifolia). Cattails are present in small numbers in some of the wettest habitats, but have the potential to dominate these areas, causing a local decline in species diversity. 63 LITERATURE CITED � Anderson, J.T. and L. M. Smith. 2002. The effect of flooding regimes on decomposition of Polygonum pensylvanicum in playa wetlands (Southern Great Plains, USA). Aquatic Botany 74:97-108. Battle, J.M. and S.W. Golladay. 2001. Hydroperiod influence on breakdown of leaf litter in Cypress-gum wetlands. American Midland Naturalist 146:128-145. Bervin, K. A. 1982. The genetic basis of altitudinal variation in the wood frog Rana sylvatica: An experimental analysis of life history traits. Evolution 36:962-983. . Binkley, D., D. Richter, M. B. David, and B. Caldwell. 1992. Soil chemistry in a loblolly/longleaf pine forest with interval burning. Ecological Applications 2:157-164. Bonham C. D. 1989. Measurement of terrestrial vegetation. John Wiley & Sons, Inc., New York, New York. 478 pp. Braccia, A. and D.P. Batzer. 2001. Invertebrates associated with woody debris in a southeastern U.S. forested floodplain wetland. Wetlands 21:18-31. Coleman, D. and D. A. Crossley, Jr. 1996. Fundamentals of soil ecology. San Diego: Academic Press. 205 pp. Crawley, M.J. 1993. GLIM for ecologists. Oxford: Blackwell Science. 379 pp., Daszak, P., L. Berger, A. A. Cunningham, A. D. Hyatt, D. E. Green and R. Speare. 1999. Emerging infectious diseases and amphibian population declines. Emergirig Infectious Diseases 5:1-14. FISRWG (1998). Stream Corridor Restoration: Principles, Processes, and Practices. By the Federal Interagency Stream Restoration Working Group (FISRWG)(15, Federal agencies of the U.S. government). GPO Item No. 0120-A. Franzreb, K. E., and R. A. Phillips. 1996. Neotropical migratory birds of the southern Appalachians. U.S. Forest Service, Gen. Tech. Rep. SE-96, Asheville, North Carolina. Gill, D. .E. 1978. The metapopulation ecology of the red-spotted newt, Notophthalmus viridescens (Rafinesque). Ecological Monographs 48:145-166. Hamel, P. B. 1992. The land manager's guide to the birds of the south. The Nature Conservancy, Southeast Region, Chapel Hill, North Carolina. 437 pp. Harcombe, P. A. 1987. Tree life tables. BioScience 37:557-568. Hays, R. L., C. Summers, and W. Seitz. 1981. Estimating wildlife habitat variables. U.S. Fish and Wildlife Service. FWS/OBS-81-47. 111 pp. Kartesz, J.T. 1994. A synonomized checklist of the vascular flora of the United States, Canada, and GreenIand. Timber Press, Portland, OR. 64 Klaus, N. A. and D. A. Buehler. 2001. Golden-winged Warbler breeding habitat characteristics and nest success in clearcuts in the southern Appalachian mountains. Wilson Bulletin 113:297- 301. LeGand, H.E., Jr. and S.P. Hall. 2004. Natural Heritage Program list of the rare animal species of North Carolina. North Carolina Natural Heritage Program, Raleigh, NC. 91 pp. Mallow, D. and D. A. Crossley, Jr. 1984. Evaluation of five techniques for recovering postlarval stages of chiggers (Acarine; Trombiculidae) from soil habitats. Journal of Economic Entomology 77:281-284. McGraw, J. B. 1987. Seed bank properties of an Appalachian sphagnum bog and a model of the depth distribution of viable seeds. Canadian Journal of Botany 65:2028-2035. Mooney, H. A. 1988. Lessons from Mediterranean-climate regions. In E. O. Wilson (ed.). Biodiversity. National Academy Press, Washington, D. C. Moorhead, K. K. 2001. Seasonal water table dynamics of a southern Appalachian floodplain and associated fen. Journal of the American Water Resources Association 37:105-114. Moorhead, K. K. 2003. Effects of drought on the water-table dynamics of a southern Appalachian mountain floodplain and associated fen. Wetlands 23:792-799. Moorhead, K. K. and I. M. Rossell. 1998. Southern mountain fens. Pp. 379-403 In:'M. G. Messina and W. H. Conner (eds). Southern Forested Wetlands: Ecology and Management. Lewis Publishers, Boca Raton, Florida. Moorhead, K. K., I. M. Rossell, C. Reed Rossell, Jr., and J. W. Petranka. 2001a. Ecological Assessment of a Wetlands Mitigation Bank. Phase I: Ecological Conditions and Initial Restoration Efforts. Final Report to the Center for Transportation and the Environment. North Carolina State University, Raleigh, North Carolina. Moorhead, K. K., I. M. Rossell, J. W. Petranka, and C. R. Rossell, Jr. 2001b. Tulula wetlands mitigation bank. Ecological Restoration 19:74-80. Morrison, M. L. 1986. Bird populations as indicators of environmental change. Pp. 429-451 In: Current Ornithology. Vol. 3, R. F. Johnston (ed). Plenum Press, New York, New York. Nelson, D. W. and L. E. Sommers. 1982. Total carbon, organic carbon, and organic matter. P. 539-579. In A. L. Page, R. H. Miller, and D. R. Keeney (eds). Methods of Soil Analysis, Part 2, 2"a ed. Agronomy Monograph 9, ASA and SSSA, Madison, WI. North Carolina Department of Transportation. 1997. Stream and Wetland Mitigation Plan. Tulula Creek Wetlands Mitigation Bank Graham County; North Carolina. Raleigh, North Cazolina. North Carolina Department of Transportation. 2003. Annual Report for 2003. Tulula Bog Mitigation Site. Office of.Natural Envuonment and Roadside Environmental Unit, Raleigh, NC. 65 North Carolina Natural Heritage Program. 1999. List of the rare plants of North Carolina. Division of Parks and Recreation, Raleigh, NC. 95 pp. Noss, R. F. 1983. A regional landscape approach to maintain diversity. Bioscience 33:700-706. Odum, E. P., J. E. Pinder III, and T. A. Christiansen. 1984. Nutrient losses from sandy soils during old-field succession. American Midland Naturalist 111:148-154. Pechmann, J. H. K., R. A. Estes, D. E. Scott and J.Whitfield Gibbons. 2001. Amphibian colonization and use of ponds created for trial mitigation of wetland loss. Wetlands 21: 93-111. Pendelton, G. W. 1995. Effects of sampling strategy, detection probability, and independence of counts on the use of point counts. Pp. 131-134. In: Monitoring bird populations by point counts. (eds. C. J. Ralph, J. R. Sauer, and S. Droege. U. S. Forest Service, Gen. Tech. Rep. PSW-GTR-149, Albany, California. Petranka, J. W. 1998. Salamanders of the United States and Canada. Smithsonian Institution Press, Washington D. C. 587 pp. Petranka, J. W., A. W. Rushlow, and M. E. Hopey. 1998. Predation by tadpoles of Rana sylvatica on embryos of Ambystoma maculatum: Implications of ecological role reversals by Rana (predator) and Ambystoma (prey). Herpetologica 54:1-13. Petranka, J. W. and C. A. Kennedy. 1999. Pond tadpoles with generalized morphology: is it time to reconsider their functional roles in aquatic communities? Oecologia 120:621-631. Petranka, J. W., C. A. Kennedy, and S. M. Murray. 2003a. Response of amphibians to restoration of a southern Appalachian wetland: a long-term analysis of community dynamics. Wetlands 23:1030-1042. Petranka, J. W., S. M. Murray, and C. A. Kennedy. 2003b. Response of amphibians to restoration of a southern Appalachian wetland: perturbations confound post-restoration assessment. Wetlands 23: 278-290. Radford, A. E., H. E. Ahles, and C. R. Bell. 1968. Manual of the Vascular Flora of the Carolinas. University of North Carolina Press, Chapel Hi11, NC. Reynolds, Barbara C., D. A. Crossley, Jr., and Mark D. Hunter. 2003. Response of soil microarthropods to forest canopy inputs along a productivity gradient. Pedobiologia 47: 127- 139. Rosgen, D. 1996. Applied River Morphology. 2"d Edition. Wildland Hydrology, Pagosa Springs, Colorado. Rossell, C. R.. Jr., I. M. Rossell, J. W. Petranka, and K. K. Moorhead. 1999. Characteristics of a partially disturbed southern Appalachian forest-gap bog complex. Pp. 81-89 In: R. P. Eckerlin (ed). Proceedings of the Appalachian Biogeography Symposium. Virginia Museum of Natural History Special Publication Number 7, Martinsville, Virginia. .: Rossell, C. R.,.Jr. 2001. Song perch characteristics of Golden-winged Warblers in a mountain wetland. Wilson Bulletin 113:246-248. Rossell, C. R., Jr., I. M. Rossell, M. M. Orraca, and J. W. Petranka. 2002. Epizootic disease and high mortality in a population of eastern box turtles. Herpetological Review 33:99-101. Rossell, C.R., Jr., S.C. Patch, and S.P. Wilds. 2003. Attributes of Golden-winged Warbler territories in a mountain wetland. Wildlife Society Bulletin 31:1099-1104. Rossell, I. M. 1996. Noteworthy collections from North Carolina: Lilium canadense ssp. editorum (Liliaceae). Castanea 61:196-197. Rossell, I. M. and C. L. Wells. 1999. The seed banks of a southern Appalachian fen and an adjacent degraded wetland. Wetlands 19:365-371. SAS Institute, Inc. 2001. Version 8.02 for Windows. SAS Institute, Inc. Cary, NC. Seastedt, T.R. 1984. The role of microarthropods in decomposition and mineralization processes. Annual Review of Entomology 29:25-46 , Semlitsch, R. D., Scott, D. E., Pechmann, J. H. K., and J. W. Gibbons. 1996. Structure and dynamics of an amphibian community: Evidence from a 16-year study of a natural pond. In: Long tertn studies of vertebrate communities. M. L. Cody and J. Smallwood (eds.). Academic Press. Short, C. 1988. Mitigation Banking. Biological Report 88(41). U. S. Fish and Wildlife Service, Washington D.C. Swift, M.J., Heal, O.W., and Anderson, J.M. 1979. Decomposition in Terrestrial Ecosystems. University of California, Berkeley, CA 372 pp. Thunhorst, G.A. 1993. Wetland planting guide for the northeastern United States — plants for wetland creation, restoration, and enhancement. Environmental Concern, Inc., St. Michaels, MD. 179 pp. U.S. Department of Agriculture. 2001. The PLANTS database, Version 3.1. National Plant Data Center, Baton Rouge, LA: (http://plants.usda. o�v). U.S. Fish and Wildlife Service. 1996. The 1996 national list of vascular plant species that occur in wetlands. U.S. Fish and Wildlife Service, Washington, D.C. 209 pp. Warren, R.J., II., I. M. Rossell, and K.K. Moorhead. 2004. Colonization and establishment of red maple (Acer rubrum) in a southern Appalachian wetland. Wetlands 24:364-374. Weakley, A. S. 2000. Flora of Carolinas and Virginia (working draft). The Nature. Conservancy, Southern Resource Office, Durham, NC. 898 pp. Weakley, A. S. and M. P. Schafale. 1994. Non-alluvial wetlands of the southern Blue Ridge -- diversity in a threatened ecosystem. Water, Air, and Soil Pollution 77:359-383. ' Wilbur, R. B. and N. L. Christensen. Y983. Effects of fire on nutrient availability in a North Carolina Coastal Plain pocosin. American Midland Naturalist 110:54-61. 67 . . . . r-. . r • rn i i • . . _ _ _ _ _ _ . � r Appendix A2. Cross sections of riffles and pools in stream.segment Ia. 68 - �.��.is....r.xaa� p .......... Rr�urveY-100] --- R��urvey-]00) \ C -I � \ \ \ ' ..............�� ........... .. — 2 \ ... ____ L��_�� V .� Segment II - Ri(tle 2 .� 0 3 1 b � n�:�.�� n«o.. cn.�o�i �n I �a �i — Ieili�ISUrveY � 2001 p .......... R...rvsY � l002 ' --- Re..rvay � lOOJ ♦ \ \ .1 \\ \ � ' \ . / \ ' � \' .Z \'. // \' .. / \_ . J �...... \ ...... ��. ; ' ��_—_� Segm ent 11 - Pool 2 ' -� . o z � a � �. .. Di�tence Acro.e Chennei it�. Annendix A4_ Cr�ss secti�n� �f riffles and nn�ls in ctream seament TTT 69 Annendix AS_ Cr�ss secti�n� nf riffles anci nnnls in stream ceument TV_ — �.;,;.� s...., - �aa� p .......... Ra.urvey . 3003 --- R..orrey - 3001 � \. \. \ .� �` C v / ....... 5 / � � r �. / �. l � � 2 \^.� //. {' V , '�.\\.. _ f i -\ / �J � : / `... � \ / Segmen� IV • Riffle 2 � _� a s � s � w i¢ Di��snce Acroaa Ch�onel (fl) A ««..«,7:.. A L !"�«,..... ..,.,.4;,.«.. ..F«:�fl,... ..«,7 «,...1.. :« ..4«,...«.. ....,'.......«4 T�7.. �o — ,.;,;., s.,.... zoaz — ,.;,;.� se..., . zooz o .......... a...rv.y . zoo� o .......... a......y - zoo� � �.` ___ R.....vy . zoaa � .�.� ��.... �\ ___ x..�...y - ioo� ` ' \ 'v.` ��.�. ♦ � �.< i:' � "i ��'-� �� . /� � `.� ` .� � � :� � � Z = � � i � n ri ��.� :Y .. � :�i ��\ i \ i ' . . a -1 �\ : 1 . � I � I + I Segment V- Rifflt 2 _� Segment V- Pool 2 ��,\_,�� __� -� — a i � a � �o iz o z � s � �a iz �� ic Distance Acro�� Chennel (ft) Di�tance Acro�a Chsnne4(fl) ppen ix . ross ec ions o n es an pools m s eam segmen a. � ., � � �_ ., $egment Va - Riffle 1 I � 6 � 10 I I �Z .J � — �.�aa sn..r -�oox ........ R��v�Y � 200� $egment Ve - Poo12 x � � � Din�nce Acrou Chaonel (f�) �i Appendix B. Water-table data for electronic wells (see Fig. 3 for well location). Months with asterisks are significantly different. B 1. Pre- and post-restoration water-table data from the electronic wells of stream Section I. 12 c 0 -12 L � -24 �° -36 -48 O N D J monthly avg pre and post 11 12 -� 0 � -12 � -24 �° -36 �8 0 -10 c � -20 t � -30 -40 -50 O N D J F M A mo�th t 1yr pre �-2yr post F M A M J month � 1yr pre —a-2yr post morrthly avg pre and post HZ J A S O N D J F M A M J J A S month -♦-1 yr pre � 2yr post monthly avg pre and post H3 M J J A S 12 -� 0 L -12 � -24 �° -36 -48 12 c 0 `� -12 s .+ y -24 � -36 -48 12 c � � -12 � -24 �° -36 �8 monthly avg pre and pc�st O1 S O N D J F M A M J J A month —�--1 yr pre —�— 2-yr pos� monthly avg pre and post G2 S O N D J F M A M J J A month t 1 yr pre --� 2yr post maithly avg pre and pc�st X S O N D J F M A M J J A month —*-1 YR PRE --E— 2 YR POST �z B2. Pre- and post-restoration water-table data from the electronic �wells of Section II and IlI. 12 � 0 s -12 a -24 d '� -36 -48 N D J F M A M J month --�-1yr pre —�— 2yr post monthly avg pre and post E�l 12 � 0 r -12 a -24 m a _� -48 N D J monthly avg pre and post Q J A S O 12 c 0 �- -12 t �. -24 �° -36 -48 N D J F M A M J J A S O month �-1yr pre —�-2yr post morrthly avg pre and post E3 F M A M J J A S O , month —�-1yr pre —a— 2yr post 73 12 0 c _- -12 t .. � -24 � -36 -48 N monthly avg pre and post F1 12 c 0 - -12 r � -24 �° -36 -48 D J F M A M J J A S O month t 1yr pre -i- 2yr post monthly avg pre and post F2 N D J ' F M A M J J A S O month -�-1yr pre -f- 2yr post 74 B3. Pre-. and post-restoration water-table data from the electronic wells of Section N. 12 c 0 L -12 � -24 �° -36 -48 D J monthly avg pre and post D1 12 c � � -12 � -24 �° -36 -48 D J 2 r . « a a� � - F M A M J J A S O N month . t 1yr pre —�— Tyr post morrthly avg pre and post D2 F M A M J J A S O N month —� 1yr pre �-2yr post monthly avg pre and post D4 D J F M A M J J A S O N m onth 75 76 B4. Pre- and post-restoration water-table data from the electronic wells of Section V and Va. 12 � 0 _ -12 a -24 'o -36 -48 12 c 0 s '�2 a -24 � -36 -48 12 � 0 r -12 a -24 d � -36 -48 monthly avg pre and post C1 . J J A S O N D J F M A M m onth . �-1 yr pre —F- 2 yr post morrthly avg pre and post C2 J J A S O N D J F M A M month t 1 12 yr pre —a— 2 yr post monthly avg pre and post 61 A S O N D J F M A M J J month --�— 2yr pre —�- 2 yr post � c r " .. a - d � . 12 c 0 r -12 a -24 � -36 -48 12 c 0 r -12 a -24 � � -36 -48 A S O N morrthiy avg pre and post 62 * * ,� * * * * * A S O N D J F M A M J J month � Series 1 �— Series2 monthly avg pre and post B3 A S O N D J F M A M J J month �- Tyr pre —� 2 yr post moMhly avg pre and post B4 D J F M A M J J month t 2yr pre � 2 yr post �� 12 c 0 r -12 a -24 � -36 -48 12 c � s -12 a -24 m '� -36 -48 morrthly avg pre and post B5 A S O N D J F M A M J J month t 2yr pre —�— 2 yr post morrthly avg pre and post A1 r c. 12 � 0 d -12 � _- -24 a`� -36 � � A S O N D J F M A M J months t 2yr pre —a— 2 yr post A S O N D J F M A M J J m onth . � 2 yr pre —f— 2 yr post monthly pre and post avg A2 J �g 12 � 0 t -12 a -24 d 'a -36 -48 A S O N D J F M A M J J m onths � 2yr pre --a-- 2 yr post morrthly avg pre and post A3 12 c 0 _ -12 a -24 m 'a -36 -48 A S O N D J F M A M J J month --�- 2yr pre —�-- 2 yr post morrthiy avg pre and post A4 � 0 79 Appendix C. Water-table data from manual wells (see Fig. 3 for location of site location of wells). Months with asterisks are significantly different. C 1. Pre- and post-restoration water table from manual wells located on the eastern side of Tulula. C 1 a. Six manual wells in Tulula fen 3C: Monthly water table depth pre and post 12 -- c 0 — t -12 Q -24 �a -36 * * -48 O N D J F M A M J J A S month �- 7 yr pre --F- 2 yr post 91: Monthly water table depth pre and post 12 - c � L -�2 a -24 � -36 * * -48 ' O N D J F m A M J J A S Month -♦- 7-yr pre f 2-YR Fbst 8C: Monthly water table depth pre and post 12 c � t -12 a -24 'a -36 � * * * O N D J F m A M J J A S Month -�- 7yr pre -a- Tyr post 3F: Monthiy water table depth pre and post 12 .- ---, _� 0 I r -12 �. -24 i � -36 -48 , * O N D J F M A M J J A S Month �- 7 yr pre -a- 2 yr post 7F: Monthly water table depth pre and post 12 -- � 0 -- L -12 �. -24 � -36 �8 * O N D J F m A M J J A S month t 7-yr pre --F- 2-yr post 61: Monthly water table depth pre and post 12 c � t -12 g, -24 v _36 * * -48 O N D J F m A M J J A S Month -� 7yr pre --E- 2yr post m Appendix Clb. Six manual wells located on floodplain adjacent to Tulula fen. 111: Monthiy water table depth pre and post 12 _ _ ....__ . ............._ __...._.. �....__.. _ ...._.__, c 0 t -12 � d -24 G � -48 * O N D J F M A M J J A S Month -� 7-yr pre avg �- 2-yr post avg 1111: Monthly water table depth pre and post 12 c 0 t -12 a -24 g' _36 ' � w * O N D J F M A M J J A S Month �- 7yr pre avg -a- Tyr post avg N1: Monthy water table depth pre and post , 12 -•— c � L -12 � -24 � -36 � * w O N D J F M A M J J A S Month -�--- 7 yr pre �- 2 yr post 112: Monthiy water table depth pre and post 12 --- --� 0 � .6 _12 t � -24 v _� � * * * * O N D J F m A M J J A S Nbrdh �- 7-yr pre -� 2-yr post 1112: Monthly water table depth pre and post 12 c � . t -12 m -24 9 -36 � •* * O N D J F m A M J J A S Month -+- 7-yr pre -a- 2-yr post N2: Monthly water table depth pre and post 12 c � r -12 � -24 9 -36 -48 * * O N D J F m A M J J A S Month �- 7-yr pre --s- 2-yr post 81 C 1 c. Transect wells (T-series) located near Tulula Fen. T1: Monthy water table depth pre and post 12 0 — t -12 � -24 � -36 * * * * * � * * * -48 O N D J F m A M J J A S month --�— 7-yr pre —�-- 2-yr post � T3: Monthly water table depth pre and post 12 - ---•- c � r -12 d -24 'o -36 � * * * ,r O N D J F m A M J J A S month --�— 7-yr pre --s— 2-yr post T5: Monthly water table depth pre and post 12 c � t -12 � -24 � -36 � * * O N D J F m A M J J A S Month �— 7-yr pre �- 2-yr post T2: Monthiy water table depth pre and post 12 - -_____.�._...__-----------._..... -- i 0 —a - -12 ' " -24 � . � . _� I -48 '� * O N D J F m A M J J A S month T4: Monthly water table depth pre and post 12 c � t -12 � -24 '� -36 * * -48 O N D J F m A M J J A S Month , —� 7-yr pre —a— 2-yr post T6: Monthly water table depth pre and post 12 --- c 0 � -12 m -24 'o -36 � * * O N D J F m A M J J A S Month -t-7-yr pre —�-2-yr post 82 T7: Monthly water table depth pre and post 12 c � " r -12 a -24 m '° -36 * * -48 O N D J F m A M J J A S m onth —� 7-yr pre ---�— 2-yr post T9: Monthly water table depth pre and post 12 -- c 0 r -12 a -24 � -36 -48 O N D J F m A M J J A S month —� 7-yr pre —a— 2-yr post T11: Monthly water table depth pre and post 12 •--„ __, c � r '�2 � -24 �° -36 -48 O N D J F m A M J J A S Month --�— 7-yr pre —s�— 2-yr post T8: Monthly water table depth pre and post 12 __ ...,..,. _. c � L -�2 a -24 m 'o -36 -48 "` . * * : O N D J F m A M J J A S Month —�- 7-yr pre --�— 2-yr post T70: Monthly water table depth pre and post 12 _.. - � � — r '1Z � -24 '° -36 � * * * O N D J F m A M J J A S Month �-7-yr pre —�-2-yr post T12: Water table depth pre and post 12 - c o v ��•1 � -24 -36 -48 � O N D J F m A M J J.A S Month �- 7-yr pre —a— 2-yr post 83 12 0 c _�2 � -24 -36 �8 84 T14: Water table depth pre and post O N D J F m A M J J A S Nbnth APPENDIX D. Amphibian and Reptile species at Tulula Common Name Family Ambystomatidae spotted salamander Family Plethodontidae four-toed salamander Ocoee salamander black-bellied salamander Blue Ridge two-lined salamander three-lined salamander black-chinned red salamander Blue Ridge spring salamander southern Appalachian salamander southern red-backed salamander Family Salamandridae red-spotted newt Family Bufonidae American toad Family Ranidae bullfrog green frog wood frog Family Hylidae northern spring peeper gray treefrog Family Chelydridae common snapping turtle Family Emydidae bog turtle eastern box turtle Family Iguanidae (Phyrynosomatidae) eastern fence lizard Family Scincidae five-lined skink Family Colubridae northern water snake eastern garter snake eastern ribbon snake northem ringneck snake black rat snake northern black racer Family Viperidae timber rattlesnake northern copperhead Scientific name Ambystoma maculatum Hemidactylium scutatum Desmognathus ocoee D. quadramaculatus Eurycea bislineata wilderae (= E. wilderae) E. guttolineata Pseudotriton ruber schencki Gyrinophilus porphyriticus danielsi Plethodon oconaluftee Plethodon serratus Notophthalmus v. viridescens Bufo a. americanus Rana catesbeiana Rana clamitans melanota Rana sylvatica Pseudacris c. crucifer Hyla chrysoscelis Chelydra s. serpentina Clemmys muhlenbergii Terrepene c. carolina Sceloporus u. undulatus Eumeces fasciatus Nerodia s. sipedon Thamnopliis s. sirtalis Thamnophis s. sauritis Diadophis punctatus edwardsii Elaphe o. obsoleta Coluber c. constrictor Crotalus horridus Agkistrodon contortrix mokasen 85 APPENDIX E. Bird Species at Tulula Wetland (1994-2002). (1) Probably breeding. (4) Foraging, but not breeding. (2) Nest found. (5) Winter resident. (3) Migrant. Common Name Family Ardeidae (herons and bitterns) Great Blue Heron (4) Green Heron (4) Family Anatidae (waterfowl) Wood Duck (4) Family Cathartidae (American vultures) Black Vulture (4) Turkey Vulture (4) Family Accipitridae (hawks) Red-tailed Hawk (4) � Red-shouldered Hawk (4) Broad-winged Hawk (2) Cooper's Hawk (4) Family Pandionidae (ospreys) Osprey (3) Family Strigidae (typical owls) Eastem Screech Owl (4) Baned Owl (4) Great Horned Owl (2) Family Tetraonidae (grouse) Ruffed Grouse (4) Family Phasianidae (quail, pheasants, etc.) Northem Bobwhite (1) Family Meleagrididae (turkeys) Wild Turkey (2) Family Scolopacidae (sandpipers) American Woodcock (1) Common Snipe (4) Solitary Sandpiper (3) Spotted Sandpiper (3) Family Columbidae (pigeons and doves). Mourning Dove (1) Family Cululidae (cuckoos) Yellow-billed Cuckoo (4) Black-billed Cuckoo (3) Family Caprimulgidae (goatsuckers) Whip-poor-will (1) Family Apodidae (swifts) Chimney Swift (4) Family Trochilidae (hummingbirds) Ruby-throated H»mmingbird (2) Family Alcedinidae (l�ngfishers) Belted Kingfisher (4) Scientific Name Ardea herodias Butorides striatus Aix sponsa Coragyps atratus Cathartes aura Buteo jamaicensis Buteo lineatus Buteo platypterus Accipiter cooperii Pandion haliaetus Otus asio Strix varia Bubo virginianus Bonasa umbellus Colinus virginianus Meleagris gallopavo Scolopax minor Capella gallinago Tringa solitaria Actitis macularia Zenaida macroura Coccyzus americanus Coccyaus erythropthalmus Caprimulgus vociferus Chaetura pelagica Archilochus colubris Ceryle alcyon 86 87 Family Picidae (woodpeckers) Northern Flicker (2) Colaptes auratus Pileated Woodpecker (4) Dryocopus pileatus Hairy Woodpecker (4) Picoides villosus Downy Woodpecker (1) Picoides pubescens Family Tyrannidae (flycatchers) Acadian Flycatcher (1) Empidonax virescens Alder Flycatcher (3) Empidonax alnorum Eastern Pewee (1) Contopus virens Eastern Phoebe (1) Sayornis phoebe Eastein Wood-pewee (1) Contopus virens Family Hirundinidae (swallows) Northern Rough-winged Swallow (4) Stelgidopteryx serripennis Tree Swallow (4) Tachycineta bicolor Barn Swallow (4) Hirundo rustica Family Corvidae (jays and crows) Blue Jay (1) Cyanocitta cristata Common Raven (4) Corvus corax American Crow (4) Corvus brachyrhynchos Family Paridae (titmice) Carolina Chickadee (1) Parus carolinensis Tufted Titmouse (1) Parus bicolor Family Sittidae (nuthatches) White-breasted Nuthatch (1) Sitta carolinensis Red-breasted Nuthatch (3) Sitta canadensis Family Certhiidae (creepers) Brown Creeper (4) Certhia americana Family Troglodytidae (wrens) Carolina Wren (1) Thryothorus ludovicianus Winter Wren (3) Troglodytes troglodytes Family Mimidae (mockingbirds, catbirds, thrashers) Gray Catbird (1) Dumetella carolinensis Brown Thrasher (1) Toxostoma rufum Family Turdidae (thrushes) American Robin (1) Turdus migratorius Hermit Thrush (3) Catharus guttatus Wood Thrush (1) Hylocichla mustelina Family Sylviidae (kinglets, etc.) Blue-gray Gnatcatcher (2) Polioptila caerulea Golden-crowned Kinglet (3) Regulus satrapa Ruby-crowned Kinglet (3) Regulus calendula Family Bombycillidae (wagwings) Cedar Waxwing (1) Bombycilla cedrorum Family Vir,ionidae (vireos) White-eyed Vireo (1) Vireo griseus Yellow-throated Vireo (1) Yireo flavifrons Solitary Vireo (1) Vireo solitarius Red-eyed Vireo (1) Vireo olivaceus Family Parulidae (wood warblers) Black-and-white Warbler (1) Swainson's Warbler (1) Worm-eating Warbler (3) Golden-winged Warbler (1) Blue-winged Warbler (3) Northern Parula (2) Pine Warbler (1) Black-throated Blue Warbler (3) Black-throated Green Warbler (3) Yellow-throated Warbler (1) Chestnut-sided Warbler (1) Yellow Warbler (3) Ovenbird (2) Kentucky Warbler (2) Common Yellowthroat (1) Yellow-breasted Chat (1) Canada Warbler (3) Hooded Warbler (2) , American Redstart (3) Prairie Warbler (1) Family Icteridae (blackbirds) Common Grackle (1) Red-winged Blackbird (4) Brown-headed Cowbird (1) Family Traupidae (tanagers) Scarlet Tanager (1) Family Fringillidae (finches, etc.) Northern Cardinal (1) Indigo Bunting (2) Blue Grosbeak (3) American Goldfmch (1) Rufous-sided Towhee (2) Northern Junco (5) White-throated Sparrow (5) Field Sparrow (3) Fox Sparrow (3) Swamp Sparrow (5) Song Spanow (1) Mniotilta varia Limnothlypis swainsonii Helmitheros vermivorus Vermivora chrysoptera Vermivora pinus Parula americana Dendroica pinus Dendroica caerulescens Dendorica virens Dendroica dominica Dendroica pensylvania Dendroica petechia Seiurus aurocapillus Oporornis formosus Geothlypis trichas Icteria virens Wilsonia canadensis Wilsonia citrina Setophaga ruticilla Dendroica discolor Quiscalus quiscula Agelaius phoenicus Molothrus ater Piranga olivacea Cardinalis cardinalis Passerina cyanea Guiraca caerulea Carduelis tristis Pipilo erythrophthalmus Junco hyemalis Zonotrichia albicollis Spizella pusilla Passerella iliaca Melospiza georgiana Melospiza melodia 88