a JOURNAL OF THE KENTUCKY ACADEMY OF SCIENCE Official Publication of the Academy pera &. a, Volume 69 Number 2 Fall 2008 The Kentucky Academy of Science Founded 8 May 1914 GOVERNING BOARD 2008 EXECUTIVE COMMITTEE 2008 President: John Mateja, Murray State University/john.mateja@murraystate.edu President Elect: Robin Cooper, University of Kentucky/RLCOOP1@pop.uky.edu Vice President: Nancy Martin, University of Louisville/nancymartin@louisville.edu Past President: Nigel Cooper, University of Louisville Medical Center/nigelcooper@louisville.edu Secretary: Robert Kingsolver, Bellarmine University/kingsolver@bellarmine.edu Treasurer: Kenneth Crawford, Western Kentucky University/kenneth.crawford@wku.edu Executive Director (ex officio): Jeanne Harris, Lexington, Kentucky/executivedirector@kyscience.org Editor: JOURNAL (ex officio): David White, Murray State University/david.white@murraystate.edu Executive Secretary Emeritus: Don Frasier, University of Kentucky/dfrazie@uky.edu DIVISION AND AT-LARGE REPRESENTATIVES Biological Sciences (2009): Sean O’Keefe, Morehead State University/s.okeefe@moreheadstate.edu Biological Sciences (2011): Richard Durtsche, Northern Kentucky University/durtsche@nku.edu Physical Sciences (2010): Eric Jerde, Morehead State University/e.jerde@moreheadstate.edu Physical Sciences (2008): KC Russell, Northern Kentucky University/russellk@nku.edu Social & Behavioral Sciences (2010): Kenneth Tankersley, Northern Kentucky University/cavetank@aol.com Social & Behavioral Sciences (2008): Sean Reilley, Morehead State University/d.olson@moreheadstate.edu At-Large (2009): George Antonious, Kentucky State University/gantonious@kysu.edu At-Large (2010): Cheryl Davis, Western Kentucky University/Cheryl.davis@wku.edu Program Coordinator (ex officio): Robert Creek, Eastern Kentucky University/robertcreek@bellsouth.net Director, Junior Academy of Science: Ruth Beattie, University of Kentucky/rebeat1 @email.uky.edu Editor, NEWSLETTER (ex officio): Susan Templeton, Kentucky State University/susan.templeton@kysu.edu Editor, KAS Webpage (ex officio): Claire Rinehart, Western Kentucky University/Claire.rinehart@wku.edu_ AAAS/NAAS Representative (ex officio): Ruth Beattie (tentative), University of Kentucky/rebeat1@email.uky.edu COMMITTEE ON PUBLICATIONS Editor and Chair: David S. White, Murray State University/david.white@murraystate.edu Abstract Editor: | Robert J. Barney, Kentucky State University/robert.barney@kysu.edu Index Editor: Ralph Thompson, Berea College/ralph_thompson@berea.edu Editorial Board: | Susan Templeton, Kentucky State University/susan.templeton@kysu.edu Claire Rinehart, Western Kentucky University/claire.rinehart@wku.edu All manuscripts and correspondence concerning manuscripts should be addressed to the Editor (david.white@ murraystate.edu). The JOURNAL is indexed in BioOne, Cambridge Scientific Abstracts, Selected Water Resource Abstracts, State Acad- emies of Science Abstracts, and Zoological Record. Membership in the Academy is open to anyone with an interest in science. Interested parties can join online at www. kyscience.org where membership options and benefits are listed. Please contact the executive director at executivedirector@ kyscience.org for additional information. The JOURNAL is made available as a PDF to all active members of the Academy (go to www.kentuckyscience.org). Hard copy subscriptions are available. Subscription rates for nonmembers are $50.00/year domestic, $60.00/year for- eign. Back issues are $30.00 per volume. The JOURNAL is issued in spring and fall. Two issues comprise a volume. All other correspondence concerning memberships or subscriptions may be addressed to the Executive Director, Ken- tucky Academy of Science, P.O. Box 22579, Lexington, KY 40522-2579 or executivedirector@kyscience.org. This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). INSTITUTIONAL AFFILIATES Enhanced Members Berea College Bellarmine University Centre College Eastern Kentucky University Morehead State University Murray State University Northern Kentucky University Spalding University Transylvania University University of Kentucky University of Louisville Western Kentucky University Kentucky Community and Technical College System Fellows Kentucky Biomedical Research Infrastructure Network Sustaining Members Brescia University Campbellsville University Kentucky State University Members Asbury College Kentucky Wesleyan College Lindsey Wilson College Thomas More College University of the Cumberlands Associate Member Pikeville College INDUSTRIAL AFFILIATES Enhanced Member Kentucky Science and Technology Corporation Patron Lumins Associates Sustaining Member Third Rock Consultants Members CEM Corporation Kentucky American Water Associate Members Hoffman Environmental Research Institute Wood Hudson Cancer Research Laboratory 89 a ass oe, RS < Peat NT JOURNAL OF THE KENTUCKY ACADEMY OF SCIENCE ISSN 1098-7096 Continuation of Transactions of the Kentucky Academy of Science Volume 69 Fall 2008 Number 2 J. Ky. Acad. Sci. 69(2):91-100. 2008 Annotated List of the Leaf Beetles (Coleoptera: Chrysomelidae) of Kentucky: Subfamily Chrysomelinae Robert J. Barney’ Community Research Service, Kentucky State University, Frankfort, Kentucky 40601 Shawn M. Clark Monte L. Bean Life Science Museum, Brigham Young University, Provo, Utah 84602 and Edward G. Riley Department of Entomology, Texas A&M University, College Station, Texas 77843 ABSTRACT An examination of leaf beetle specimens (Coleoptera: Chrysomelidae) in the five largest beetle collections in Kentucky, recent inventory work in state nature preserves, and a review of the literature revealed twenty- two species of the subfamily Chrysomelinae present in Kentucky, eight of which were previously unreported for the state. Distribution maps and label data are presented for the twenty-two Kentucky species of the subfamily Chrysomelinae including spatial (state and Kentucky county records), temporal (years and months of collection in Kentucky), and plant association information. The following species are reported from Kentucky for the first time: Chrysolina cribaria (Rogers), Chrysolina quadrigemina (Suffrian), Calligrapha californica coreopsivora Brown, Calligrapha multipunctata (Say), Calligrapha pnirsa Stal, Calligrapha spiraeae (Say), Gastrophysa cyanea F. E. Melsheimer, and Phratora americana americana (Schaeffer). KEY WORDS: Kentucky, leaf beetles, Chrysomelidae, biodiversity, new state records INTRODUCTION This paper is the third in a series intended to present a synopsis of the historical collection data on leaf beetles (Coleoptera: Chrysomelidae) from the ma- jor Coleoptera collections in Kentucky and augment these data with new information gained from recent monitoring in state preserves and other locations. The first two papers presented information on the subfamilies Cassidinae (Barney et al. 2007) ' Corresponding author e-mail: robert.barney@kysu. edu and Donaciinae and Criocerinae (Barney et al. 2008), and this paper will address the Chrysomelinae. The subfamily Chrysomelinae is a group of leaf beetles with 135 species in 16 genera in America north of Mexico (Daccordi 1994: Riley et al. 2002). Several species in this subfamily are of economic importance, espe- cially the Colorado potato beetle, Leptinotarsa decemlineata (Say), and two species of Chry- solina were introduced into North America as biological control agents for Klamath weed (Hypericum perforatum L.): C. hyperici hyperici (Forster) and C. quadrigemina (Suf- frian) (Wilcox 1972). 92 Journal of the Kentucky Academy of Science 69(2) The purpose of this study is to present the historical and current knowledge of the distribution, abundance, and plant associa- tions of chrysomeline leaf beetles in Kentucky. MATERIALS AND METHODS To establish a historical perspective, leaf beetle specimens from the major insect collections in Kentucky (and from collections located in other states, but known to contain Kentucky specimens) were examined, re- identified, and their label data recorded. The following collections were studied: the Brig- ham Young University Collection (BYUC, Provo, UT); the Charles Dury Collection at the Cincinnati Museum of Natural History (CMNH, Cincinnati, OH); the private collec- tion of Charles Wright, the Kentucky Beetles Project Collection (CWIC, Frankfort, KY), which was established in 1991 in an effort to document Coleoptera within the state; the Kentucky State University Insect Collection (KYSU, Frankfort, KY), which houses the specimens generated by the Kentucky Leaf Beetle Biodiversity Project; the private col- lection of Robert J. Barney (RJBC, Frankfort, KY), which comprises two time periods of collecting in Kentucky, 1976-1984 and 2004— 2007; the University of Kentucky Insect Collection (UKIC, Lexington, KY), which contains the Charles V. Covell, Jr. Collection (emeritus professor of the University of Louisville); and the Western Kentucky Uni- versity Collection (WKUC, Bowling Green, KY). We currently are conducting extensive collecting in many grass-dominated barrens and rock outcrop (glade) communities that are known for possessing uncommon plants and plant associations (Jones 2005). These sites are primarily in state nature preserves that have never been surveyed for plant-feeding beetles. Most specimens were collected by the senior author within five state nature preserves in 2004—2007: Crooked Creek Barrens (Lewis County) and Blue Licks Battlefield (Robert- son County) in northeastern Kentucky, East- view Barrens (Hardin County) and Thompson Creek Glades (LaRue County) in central Kentucky, and Raymond Athey Barrens (Lo- gan County) in western Kentucky. Voucher specimens are housed in the Kentucky State University Insect Collection. For each chrysomeline species documented for Kentucky, the following data are present- ed: state-level distribution in the United States (from Riley et al. 2003), Kentucky county records, abundance by year and month in Kentucky, and specimens per collection. When present on specimen labels, other pertinent information, such as the method of collection and plant association information, is presented in the “Comments” section for each species. This information provides the oppor- tunity to determine, from a historical perspec- tive, abundance, seasonality, and distribution. One should note that plant collection records taken from specimen labels are notoriously inaccurate and may not reflect true host plants (Clark et al. 2004). RESULTS According to the “Catalog of Leaf Beetles of America North of Mexico” (Riley et al. 2003), there were 48 species of Chrysome- linae recorded in at least one of the seven states contiguous to Kentucky, thus establish- ing a “ballpark” estimate for the state. However, in that work only 14 species were listed from Kentucky. An examination of 846 chrysomeline leaf beetle specimens from the major collections in the state and others known to contain Kentucky specimens re- vealed 21 species, including 13 of the 14 recorded in Riley et al. (2003) plus eight new state records (Table 1). A twenty-second species, Gastrophysa polygoni (L.), was re- ported from Kentucky by Riley et al. (2003). The state collection at the University of Kentucky (UKIC) contains a total of 488 chrysomeline leaf beetles representing 13 species, including four of the new state records reported herein. This collection con- tains the oldest in-state specimen records for Kentucky leaf beetles, with collection dates as early as 1889. The CWIC collection has 59 specimens representing ten species, including one of the new state records reported herein. The collection at WKUC has 41 specimens of seven species. Recent collecting in state nature preserves (the KYSU collection) has produced 154 specimens of eleven species and three new state records. The RJBC collection contains 61 specimens of nine species from Kentucky. An examination of the BYUC revealed 37 specimens in eleven Kentucky Leaf Beetles—Barney, Clark, and Riley Table 1. List of Chrysomelinae (Coleoptera: Chrysomelidae) recorded from Kentucky, with number of Kentucky specimens examined, number of Kentucky county records, range of years of collection in Kentucky, and new state records. Chrysolina cribaria (Rogers) Chrysolina quadrigemina (Suffrian) Calligrapha alnicola Brown Calligrapha androwi S. Clark & Cavey Calligrapha bidenticola Brown Calligrapha californica coreopsivora Brown Calligrapha multipunctata (Say) Calligrapha philadelphica (L.) Calligrapha pnirsa Stal Calligrapha spiraeae (Say) Zygogramma suturalis (F.) Labidomera clivicollis (Kirby) Leptinotarsa decemlineata (Say) Leptinotarsa juncta (Germar) Gastrophysa cyanea F. E. Melsheimer Gastrophysa polygoni (L.) Phaedon viridis F. E. Melsheimer Phratora americana americana (Schaeffer) Chrysomela interrupta F. Chrysomela knabi knabi Brown Chrysomela scripta F. Plagiodera versicolora (Laicharting) species. Six specimens in two species of chrysomeline leaf beetles were found in the historical Dury Collection (CMNH), which 4 specimens: 3 counties, 2004-2006 (new state record) 4 specimens: 1 county, 2005 (new state record) 1 specimen: 1 county, 1994 1 specimen: 1 county, 1994 17 specimens: 7 counties, 1894-2006 1 specimen: 1 county, 2006 (new state record) 1 specimen: 1 county, 1946 (new state record) 4 specimens: 2 counties, 1988-1994 1 specimen: 1 county, 1948 (new state record) 2 specimens: 2 counties, 1983-1992 (new state record) 111 specimens: 23 counties, 1889-2007 64 specimens: 26 counties, 1892-2007 88 specimens: 10 counties, 1889-2007 85 specimens: 31 counties, 1892-2006 146 specimens: 13 counties, 1890-2004 (new state record) State-level literature record only 6 specimens: 2 counties, 1891-1895 9 specimens: 1 county, 1974 (new state record) 4 specimens: 2 counties, 1988-1994 271 specimens: 31 counties, 1891-2006 8 specimens: 7 counties, 1901-2006 18 specimens: 8 counties, 1965-2006 comprises approximately 75,000 specimens and Raymond Athey Barrens. primarily collected between 1871 and 1931 in the Cincinnati/northern Kentucky area (Vulinec and Davis 1984). 1B) (new state record) Kentucky Counties: Logan Chrysolina_ cribaria (Rogers) (Figure 1A) Years: 2005 (4) (new state record) Months: June (4) Kentucky Counties: Hardin, LaRue, Logan Abundance: 4 specimens: 4-KYSU Years: 2004 (1), 2005 (1), 2006 (2) Months: June (2), July (1), September (1) Abundance: 4 specimens: 4-KYSU Chrysolina cribaria (Rogers) Figure 1. The known distribution of Chrysomelinae (Coleoptera: Chrysomelidae) illustrated in grey shading for Comments: All four specimens were re- cently collected in one month in one man- agement unit of Raymond Athey Barrens Chrysolina quadrigemina (Suffrian) Kentucky counties and states of the United States. New state records reported herein are shown in cross-hatch. Comments: All four specimens were re- cently collected in state nature preserves: Eastview Barrens, Thompson Creek Glades, Chrysolina quadrigemina (Suftrian) (Figure 94 Journal of the Kentucky Academy of Science 69(2) State Nature Preserve. This species was introduced in eastern Canada and the Pacific Northwest to control Klamath weed and feeds on Hypericum spp. (Clark et al. 2004). Calligrapha alnicola Brown (Figure 2A) Kentucky Counties: Rowan Years: 1994 (1) Months: May (1) Abundance: 1 specimen: 1-BYUC Comments: This species feeds on Alnus incana (L.) Moench. (Clark et al. 2004). Calligrapha androwi S. Clark & Cavey (Figure 2B) Kentucky Counties: Rowan Years: 1994 (1) Months: May (1) Abundance: 1 specimen: 1-California Acad- emy of Sciences Comments: This is a recently described species from specimens collected in Ohio, Kentucky, West Virginia and Alabama. This paratype specimen is housed in the Cali- fornia Academy of Sciences (Clark and Cavey 1995). Calligrapha bidenticola Brown (Figure 2C) Kentucky Counties: Anderson, Bracken, Fayette, Franklin, Lewis, Russell, Warren Years: 1894 (4), 1928 (1), 1938 (1), 1992 (2), 1993 (1), 1995 (1), 1998 (2), 2005 (4), 2006 (1) Months: April (1), May (1), June (1), July (5), August (6), September (3) Abundance: 17 specimens: 2-BYUC, 4- CWIC, 4-KYSU, 6-UKIC, 1-WKUC Comments: This species has been associat- ed with Asteraceae (Clark et al. 2004). Calligrapha californica coreopsivora Brown (Figure 2D) (new state record) Kentucky Counties: Franklin Years: 2006 (1) Months: May (1) Abundance: 1 specimen: 1-KYSU Comments: This single specimen, a new state record, was collected in a recently established local nature preserve in Frankfort, KY. Calligrapha multipunctata (Say) (Figure 2E) (new state record) Kentucky Counties: Fayette Years: 1946 (1) Months: October (1) Abundance: | specimen: 1-UKIC Comments: This species feeds on Salix spp. (Clark et al. 2004). Calligrapha philadelphica (L.) (Figure 2F) Kentucky Counties: McCreary, Menifee Years: 1988 (1), 1990 (1), 1994 (2) Months: May (4) Abundance: 4 specimens: 4-BYUC Comments: Hosts are species of Cornus (Clark et al. 2004). Calligrapha pnirsa Stal (Figure 2G) (new state record) Kentucky Counties: Lincoln Years: 1948 (1) Months: April (1) Abundance: | specimen: 1-UKIC Comments: This species feeds on Tilia americana (L.) (Clark et al. 2004). Calligrapha spiraeae (Say) (Figure 2H) (new state record) Kentucky Counties: Fayette, Franklin Years: 1983 (1), 1992 (1) Months: April (1), June (1) Abundance: 2 specimens: 1-CWIC, 1-RJBC Comments: The host of this species is Physocarpus opulifolius (L.) Maxim. (Clark et al. 2004). Zygogramma suturalis (F.) (Figure 3A) Kentucky Counties: Barren, Bracken, Breathitt, Carter, Fayette, F ranklin, Grayson, Hardin, Jefferson, Kenton, LaRue, Laurel, Lewis, Lincoln, Logan, Muhlenburg, Pendle- ton, Robertson, Rowan, Russell, Warren, Washington, Woodford Years: 1889 (6), 1890 (2), 1891(4), 1892 (2), 1893 (1), 1894 (9), 1895 (3), 1912 (1), 1913 (2). 1917 (1), 1937 (10), 1988 (32194453) 1948 (1), 1962 (1), 1965 (1), 1968 (4), 1970 (1), 1977 (1), 1979 (1), 1983 (2), 1984 (11), 1987 (5), 1989 (1), 1991 (1), 1994 (3), 1995 (3), 1998 (3), 2001 (3), 2004 (3), 2005 (9), 2006 (4), 2007 (6) Months: March (1), April (3), May (11), June (17), July (36), August (33), September (7), October (3) Abundance: 111 specimens: 6-BYUC, 3- CWIC, 19-KYSU, 21-RJBC, 52-UKIC, 10- WKUC Kentucky Leaf Beetles—Barney, Clark, and Riley 95 ; Calligrapha androwi S, Clark & Cavey Calligrapha alnicola Brown ‘ Calligrapha multipunctata (Say) Figure 2. The known distribution of Chrysomelinae (Coleoptera: Chrysomelidae) illustrated in grey shading for Kentucky counties and states of the United States. New state records reported herein are shown in cross-hatch. 96 Journal of the Kentucky Academy of Science 69(2) Zygogramma suturalis (F.) Gastrophysa cyanea F, E. Melsheimer ¢¢ Gastrophysa polygoni (L.) Figure 3. The known distribution of Chrysomelinae (Coleoptera: Chrysomelidae) illustrated in grey shading for Kentucky counties and states of the United States. New state records reported herein are shown in cross-hatch. Comments: Label data on UKIC specimens from the 1890s noted ‘smartweed’, “A. corn- nut’, and ‘English bluegrass’. This species feeds on ragweeds, Ambrosia spp. (Clark et al. 2004). Labidomera clivicollis (Kirby) (Figure 3B) Kentucky Counties: Allen, Ballard, Bour- bon, Bullitt, Carroll, Fayette, Franklin, Grant, Grayson, Greenup, Henry, Hopkins, Jeffer- son, Kenton, Lawrence, Logan, Mercer, Nelson, Nicholas, Oldham, Owen, Owsley, Pike, Scott, Warren, Woodford Years: 1892 (1), 1893 (2), 1894 (2), 1921 (1), 1938 (3), 1939 (1), 1940 (1), 1941 (2), 1944 (1), 1946 (6), 1947 (1), 1948 (3), 1950 (4), 1956 (1), 1959 (1), 1961 (1), 1965 (1), 1966 (1), 1967 (1), 1968 (1), 1969 (1), 1971 (4), 1972 (1), 1973 (2), 1978 (1), 1979 (2), 1983 (1), 1984 (1), 1987 (2), 1991 (1), 1993 (2), 1995 (1), 1999 (1), 2002 (1), 2003 (3), 2004 (1); 9005.3), 2007 (ip Months: April (2), May (8), June (16), July (7), August (7), September (10), October (13), November (1) Abundance: 64 specimens: 1-BYUC, 9- CWIC, 3-KYSU, 4-RJBC, 36-UKIC, 11-WKUC Kentucky Leaf Beetles—Barney, Clark, and Riley 97 Comments: Label data on UKIC specimens from the 1890s noted ‘Asclepias incarnata’, and ‘Eupatorium perfoliatum’. This species feeds on Asclepiadaceae (Clark et al. 2004). Leptinotarsa decemlineata (Say) (Figure 3C) Kentucky Counties: Campbell, Fayette, Franklin, Hardin, Jefferson, Russell, Shelby, Warren, Whitley, Woodford Years: 1889 (3), 1892 (34), 1915 (4), 1922 (1), 1928 (2), 1937 (2), 1961 (1), 1965 (1), 1968 (2), 1971 (3), 1978 (1), 1983 (2), 1984 (1), 1988 (1), 1991 (1), 1998 (5), 2001 (1), 2005 (20), 2007 (3) Months: April (1), May (10), June (12), July (56), August (7), September (1), October (1) Abundance: 88 specimens: 5-CMNH, 1- CWIC, 23-KYSU, 3-RJBC, 52-UKIC, 4- WKUC Comments: The Colorado potato beetle has a well documented association with Solanum tuberosum L. Labels noted some specimens collected by Malaise trap. Leptinotarsa juncta (Germar) (Figure 3D) Kentucky Counties: Anderson, Barren, Bell, Boyle, Breathitt, Butler, Campbell, Casey, Christian, Clinton, Daviess, Fayette, Floyd, Franklin, Hardin, Henderson, Jefferson, Jes- samine, LaRue, Lincoln, Marion, Menifee, Mercer, Montgomery, Muhlenberg, Nicholas, Shelby, Spencer, Warren, Whitley, Wolfe Years: 1892 (1), 1893 (1), 1905 (5), 1911 (1), 1922, (2), 1923 (1), 1925 (1), 1927 (2), 1935 (3), 1937 (9), 1938 (8), 1939 (1), 1940 (2), 1942 (1), 1944 (2), 1945 (2), 1946 (5), 1947 (2), 1948 (3), 1949 (2), 1950 (3), 1951 (3), 1952 (2), 1953 (1), 1955 (1), 1956 (1), 1957 (1), 1963 (1), 1967 (3), 1968 (1), 1983 (1), 1987 (2), 1995 (1), 2001 (1), 2003 (2), 2004 (1), 2005 (3), 2006 (3) Months: March (3), April (8), May (24), June (12), July (5), August (8), September (17), October (7), December (1) Abundance: 85 specimens: 1-BYUC, 1- CWIC, 1-KYSU, 8-RJBC, 70-UKIC, 4- WKUC Comments: This species is often called the false potato beetle and feeds on Solanum carolinense L. (Clark et al. 2004). Jacques (1988) cites one Kentucky record for Campbell County in the Purdue University collection. Gastrophysa cyanea F. E. Melsheimer (Fig- ure 3E) (new state record) Kentucky Counties: Clay, Daviess, Fayette, Fleming, Franklin, Hart, Jackson, Jefferson, Jessamine, Kenton, Marshall, Monroe, War- ren Years: 1890 (4), 1891 (69), 1892 (3), 1895 (3), 1901 (1), 1917 (8), 1923 (1), 1925 (1), 1930 (12), 1937 (4), 1938 (2), 1944 (1), 1949 (1), 1963 (1), 1968 (6), 1970 (4), 1971 (1), 1975 (5), 1976 (2), 1979 (5), 1992 (1), 1993 (1), 2000 (2), 2004 (8) Months: January (2), March (2), April (16), May (35), June (89), August (2) Abundance: 146 specimens: 12-CWIC, 7- RJBC, 120-UKIC, 7-WKUC Comments: Label data (UKIC 1890s) noted collection from smartweed, lawn, wheat, dock, red clover, oats, white clover, Rumex crispus, and by Malaise trap. Normal hosts are species of Rumex (Polygonaceae) (Clark et al. 2004). Gastrophysa polygoni (L.) (Figure 3F) Comments: No specimens from Kentucky were seen during this study, but it was listed from the state by Riley et al. (2003). This species feeds on Polygonaceae (Clark et al. 2004). Phaedon viridis F. E. Melsheimer (Figure 4A) Kentucky Counties: Ballard, Fayette Years: 1891 (1), 1894 (4), 1895 (1) Months: May (1), July (1), November (1), December (3) Abundance: 6 specimens: 6-UKIC Comments: Label data (UKIC 1890s) noted collection from alfalfa, grain, clover, and weeds. The Ballard County (Wickliffe) record is from Balsbaugh’s (1983) revision of Phae- don. Hosts of this species are Brassicaceae (Clark et al. 2004). Phratora americana americana (Schaeffer) (Figure 4B) (new state record) Kentucky Counties: Nelson Years: 1974 (9) Months: August (9) Abundance: 9 specimens: 1-RJBC, 8-UKIC Comments: This species is associated with Salix (Salicaceae) (Clark et al. 2004). Chrysomela interrupta F. (Figure 4C) Kentucky Counties: Jackson, Whitley 98 Journal of the Kentucky Academy of Science 69(2) Phaedon viridis F. E. Melsheimer Chrysomela interrupta F. Chrysomela scripta ¥. Phratora americana americana (Schaeffer) Figure 4. The known distribution of Chrysomelinae (Coleoptera: Chrysomelidae) illustrated in grey shading for Kentucky counties and states of the United States. New state records reported herein are shown in cross-hatch. Years: 1988 (2), 1994 (2) Months: May (4) Abundance: 4 specimens: 4-BYUC Comments: Both adults and larvae are associated with Alnus (Betulaceae) (Clark et al. 2004). Chrysomela knabi knabi Brown (Figure 4D) Kentucky Counties: Bath, Breathitt, Cald- well, Fayette, Fleming, Franklin, Gallatin, Grayson, Hardin, Henry, Jefferson, Kenton, Knott, Laurel, Letcher, Madison, Martin, Meade, Menifee, Metcalfe, Muhlenberg, Owen, Owsley, Pike, Powell, Rowan, Union, Warren, Webster, Whitley, Wolfe Years: 1891 (4), 1895 (1), 1901 (12), 1927 (3), 1937 (1), 1939 (10), 1940 (2), 1941 (3), 1942 (10), 1958 (7), 1965 (1), 1966 (1), 1970 (58), 1971 (6), 1980 (11), 1982 (5), 1983 (1), 1990 (1), 1992 (2), 1993 (1), 1994 (17), 1995 (2), 2001 (1), 2003 (5), 2004 (4), 2005 (87), 2006 (14) Months: April (146), May (88), June (26), July (9), October (1) Kentucky Leaf Beetles—Barney, Clark, and Riley 99 Abundance: 271 specimens: 12-BYUC, 1- CMNH, 25-CWIC, 93-KYSU, 5-RJBC, 131- UKIC, 4-WKUC Comments: There is no county, year or month available for one CMNH specimen. This species is associated with Salix (Salica- ceae) (Clark et al. 2004). Chrysomela scripta F. (Figure 4E) Kentucky Counties: Bracken, Fayette, Hardin, Jefferson, Mason, Trigg, Webster Years: 1901 (1), 1970 (1), 1971 (1), 1998 (3), 2005 (1), 2006 (1) Months: May (2), June (2), July (4) Abundance: 8 specimens: 3-BYUC, 1- CWIC, 1-KYSU, 3-UKIC Comments: The Mason County (Maysville) record is from Brown’s (1956) revision of Chrysomela. This species is associated with Salix and Populus (Salicaceae) (Clark et al. 2004). Plagiodera_ versicolora (Laicharting) (Fig- ure 4F) Kentucky Counties: Bracken, Fayette, Grayson, Henry, Jefferson, Knott, Martin, Nicholas Years: 1965 (1), 1981 (7), 1983 (4), 1998 (2), 2003 (2), 2006 (1 Months: May (8), June (6), July (2), August (1) Abundance: 18 specimens: 2-BYUC, 2- CWIC, 1-KYSU, 11-RJBC, 2-UKIC Comments: There is no county, year or month available for one UKIC specimen. This species is associated with Salicaceae (Clark et al. 2004). DISCUSSION We believe the data presented here are the most complete representation of the chry- someline leaf beetles known from Kentucky. The large number of new state records documented here (8 of 22, or 35%) reflects a historical lack of leaf beetle collecting in Kentucky. Three of the eight new state records are based on a single specimen, and eleven of the 22 species have been document- ed from only one or two counties. Six species accounted for 765 of the 846 specimens reviewed in this study, leaving only 81 beetles for the other 16 species. These data, plus the fact that the ballpark estimate for Kentucky Chrysomelinae is 48 species and that we found only 22 species, indicate that many more species may still out there to be found. The host plant switch of the Colorado potato beetle, from the native buffalo bur (Solanum rostratum Dunal) to cultivated potato (S. tuberosum L.), enabled this large showy beetle, first found in the area of the Rocky Mountains, to exploit a changing landscape and become a widespread econom- ic pest (Jacques 1988). Examples such as this are a reminder why it is so important to document the presence and distribution of our native and exotic flora and fauna. ACKNOWLEDGMENTS Thanks are extended to Michael Sharkey and Martha Potts (UKIC), Keith Philips (WKUC), Greg Dahlem (CMNH), and Charles Wright (CWIC) for access to their collections, and Joyce Bender, Lane Linnen- kohl and Zeb Weese of the Kentucky State Nature Preserves Commission for access to preserves. We also thank Joyce Owens and T’Nazja Scott (KYSU) for sorting, organizing, and transcribing, and Sarah Hall (KYSU) for creation of the distribution maps. This re- search was supported by USDA-CSREES Project KYX-10-05-39P. LITERATURE CITED Balsbaugh, E. U. 1983. A taxonomic revision of the genus Phaedon north of Mexico (Coleoptera: Chrysomelidae). North Dakota Insects, Schafer-Post Series No. 15:1—73. Barney, R. J., S. M. Clark, and E. G. Riley. 2007. Annotated list of the leaf beetles (Coleoptera: Chry- somelidae) of Kentucky: subfamily Cassidinae. Journal of the Kentucky Academy of Science 68:132-144. Barney, R. J., S. M. Clark, and E. G. Riley. 2008. Annotated list of the leaf beetles (Coleoptera: Chry- somelidae) of Kentucky: subfamilies Donaciinae and Criocerinae. Journal of the Kentucky Academy of Science 69:29—36. Brown, W. J. 1956. The New World species of Chrysomela L. (Coleoptera: Chrysomelidae). The Canadian Entomologist 88(supplement No. 3): 1-54. Clark, S. M., and J. F. Cavey. 1995. A new species of Calligrapha (Coleoptera: Chrysomelidae) from eastern North America. Insecta Mundi 9:329-333. Clark, S. M., D. G. LeDoux, T. N. Seeno, E. G. Riley, A. J. Gilbert, and J. M. Sullivan. 2004. Host plants of leaf beetle species occurring in the United States and Canada. Coleopterists Society, Special Publication No. 2. 476 pp. Daccordi, M. 1994. Notes for phylogenetic study of Chrysomelinae, with descriptions of new taxa and a list 100 of all known genera (Coleoptera: Chrysomelidae, Chrysomelinae). Pages 60-84 in D. G. Furth (ed), Proceedings of the Third International Symposium on the Chrysomelidae, Beijing, 1992. Backhuys. Jacques, R. L. 1988. The potato beetles. The genus Leptinotarsa in North America (Coleoptera: Chrysome- lidae). Flora and Fauna Handbook No. 3. J. E. Brill, Leiden. 144 pp. Jones, R. L. 2005. Plant Life of Kentucky. University Press of Kentucky. 834 pp. Riley, E. G., S. M. Clark, R. W. Flowers, and A. J. Gilbert. 2002. Chrysomelidae Latreille 1802. Pages 617-691 in Journal of the Kentucky Academy of Science 69(2) R. H. Arnett and M. C. Thomas (eds). American beetles. CRC press. Riley, E. G., S. M. Clark, and T. N. Seeno. 2003. Catalog of the leaf beetles of America north of Mexico. Coleopterists Society, Special Publication No. 1. 290 pp. Vulinec, K., and R. A. Davis. 1984. Coleoptera types in the Charles Dury Collection of the Cincinnati Museum of Natural History. Coleopterists Bulletin 38:232-239. Wilcox, J. A. 1972. A review of the North American Chrysomelinae leaf beetles (Coleoptera: Chrysomeli- dae). New York State Museum and Science Service Bulletin 421:1-37. J. Ky. Acad. Sci. 69(2):101-116. 2008. Diversity, Substrata Divisions and Biogeographical Affinities of Land Snails at Bad Branch State Nature Preserve, Letcher County, Kentucky Daniel C. Dourson! Belize Foundation for Research and Environmental Education, P.O. Box 129, Punta Gorda, Belize, Central America and Joel Beverly Apogee Environmental Consultants, P.O. Box 338, Ermine, Kentucky 41815 ABSTRACT The results of this study established that Bad Branch State Nature Preserve has at least 64 native land snail species occurring within its boundaries, representing 14 families and 35 genera. Twenty-five of the snail taxa documented at Bad Branch were new records for Letcher County; 19 species or 29% of the new records accounted for snails under 5 mm. In total, micro snail species (<5 mm) represented 45% or 28 species of the documented fauna. Two species, Paravitrea lamellidens Pilsbry and Paravitrea dentilla Hubricht were previously unknown from Kentucky. Stress in the non-metric multidimensional scaling analysis reached an acceptable level (0.118) with a three-dimensional solution. The ordination of the 19 sites indicated strong differences in the species composition of the land snail fauna between limestone and sandstone substrates. Approximately half of the species were restricted to just one substrate, with 14 restricted to sandstone and 19 restricted to limestone. Abundance (individuals/site), species richness (species/site), and percent species in the total sample were two times to almost four times greater on limestone substrate than sandstone (P = 0.001). This suggested that more sandstone habitat is required to sustain the same number of snail species as is limestone habitat and that land snail diversity in the sandstone regions of Bad Branch are working at a much larger ecological scale than in limestone regions. A number of terrestrial gastropods found in Kentucky’s Pine Mountain region, in particular Bad Branch, are biogeographically significant, as they represent primary associations with the Great Smoky Mountains-Blue Ridge sections of the Cumberland Province or the Blue Ridge region of West Virginia and Virginia. These interesting snail associations are further supported by the discovery of P. lamellidens and P. dentilla at Bad Branch State Nature Preserve. KEY WORDS: Bad Branch, land snails, Kentucky, Paravitrea lamellidens, Paravitrea dentilla INTRODUCTION Bad Branch State Nature Preserve is an outstanding illustration of an undomesticated environment, occurring along the crest of Pine Mountain, in Letcher County, Kentucky (Figure 1). The preserve safeguards one of the state’s finest upper elevation mountain stream ecosystems, and, because of its rela- tively unspoiled landscape and pristine waters, Bad Branch was designated a Kentucky Wild River System in 1989 (Figure 2). The altitude, range of soil types and the preserve’s physio- graphic location has created a wealth of biodiversity. With few exceptions, the flora and fauna remains intact. Several taxa assem- blages found in these primordial mountains, an example being the vegetation, reach ' Corresponding author email: bfreemgr@xplornet.com exceptional diversity at Bad Branch. The preserve was predicted to also have an analogous variety of land snails as well. To better understand the richness, distri- bution and biogeographical affinities of land snails of the southeastern mountains of Kentucky and in particular Bad Branch Nature Preserve, an inventory was initiated in 2006. While the number of snail species occurring in the state has been well docu- mented, approximately 170 taxa (Hubricht 1968, 1985; Branson 1973; Branson and Batch 1968, 1970, 1988), the distribution patterns and ecology of land snails remains incomplete, especially in the southeastern mountains. Recent studies in Kentucky have shown, for example, that where there are comprehensive snail inventories with an emphasis on dimin- utive snail taxa (<5 mm), numerous county records are added; typically more than 20 101 102 Figure 1. Bad Branc h State Nature Preserve is an outstanding illustration of an undomesticated environment, Journal of the Kentucky Academy of Science 69(2) occurring along the great dividing crest of Pine Mountain, in Letcher County, Kentucky. The lower elevation Cumberland Plateau region of eastern Kentucky can be seen in the background. species (Dourson and Feeman 2006; Dourson 2007). Small snails (<5 mm) can represent more than 40% of the snail fauna (Hubricht 1968; Hotopp 1999; Dourson 2007). The lack of species accounts in Kentucky, particularly the micro snail taxa, is also well-illustrated in Hubricht (1985) distribution records for the eastern land snails of North America that show numerous county gaps. If an inventory method fails to collect the leaf litter samples where the majority of snails (<5 mm) fre- quently reside, micro snail species may go unreported. Omitting micro snails from past collections has likely resulted in the vast majority of county gaps. This has resulted in many states listing a number of their terres- trial gastropods as rare or uncommon occur- rences. Site specific land snail richness and shell abundance have long been associated with a variety of geological and ecological factors. For example, terrestrial gastropods living around carbonate cliffs can exhibit large and diverse populations (Nekola 1999) but show significant declines in as little as 50 m from a calcareous source (Kalisz and Powell 2003) or limestone cliffline (Dourson 2007). Additional factors such as gradient, elevation, vegetation, litter moisture and soil cations (particularly calcium) can significantly affect land snail abundance, both in terms of richness of species and numbers of shells (Boycott 1934; Burch 1955, 1956; Atkins 1966; Agocsy 1968; Valovirta 1968; Wareborn 1970; Getz 1974; Pollard 1975; Petranka 1982; Hotopp 2002). Less well-known is how large geophysical landscape edges might serve in bridging distinctive regions and their allied terrestrial gastropod communities. Neighboring land masses may in effect be the driving force behind the distribution and mixing of some snail faunas, acting as travel corridors for dispersal. The end result can be remarkably high land snail diversity in comparatively small Land Snail Diversity at Bad Branch—Dourson and Beverly 103 Figure 2. Bad Branch is one of the finest mountain stream ecosystems to be found in Kentucky and was designated a state wild river system in 1989. Bad Branch falls is in the background. places. The Central Knobstone Escarpment in Powell County, Kentucky, is an example. This area forms a large geophysical landscape edge on Furnace Mountain where the Cumberland Plateau, the Knobs, and the Outer Bluegrass regions of Kentucky converge. The merging of these distinct regions was shown to harbor an exceptional number of land snail taxa, a reported 61 species co-existing within a 2 ha mesic hillside (Dourson 2007). A number of the snails documented at Furnace Mountain were established beyond their reported east- ern and western limits in the state (Branson and Batch 1968; Branson 1973; Hubricht 1985). The snail inventory at Bad Branch also was expected to further the understanding as to the extent of possibly unique biogeography assemblages of terrestrial gastropods in the southeastern mountains of Kentucky. Branson and Batch (1968, 1988) reported that the principal land snail affinity of this region of Kentucky was with the Great Smoky Moun- tains-Blue Ridge regions of the Cumberland Province, with some segments of the fauna showing relationships with more easterly areas, i.e., through Virginia and West Virginia. This view was further supported by Hubricht et al. (1983), whose land snail collections showed further evidence of the region’s faunal affinities to the Ridge and Valley and Blue Ridge Physiographic Provinces. STUDY AREA Woods et al. (2002) placed the study area (Figure 3) within the Cumberland Mountain Thrust Block Ecoregion. The area is charac- terized by steep ridges, hills, coves, narrow valleys, and the Pine Mountain Overthrust Fault. Forest composition is highly variable and is determined by aspect, slope position, historic usage, and degree of topographical shading. Many streams in this ecoregion are cool and high gradient; with a substrate commonly consisting of cobble and boulder (riffles are common). The underlying geology 104 ea Bad Branch State Nature Preserve NGS USA Topographic Maps Kilometers Journal of the Kentucky Academy of Science 69(2) Letcher County, Kentucky a ai e0 See, Ye Ao ie eee 2 oe Figure 3. Map showing land snail sampling locations at Bad Branch State Nature Preserve in eastern Kentucky. consists of Pennsylvanian shale, siltstone, sandstone, conglomerate and coal. The pres- ence of coal mining and acidic mine drainage, as well as logging, has led to the degradation of many streams. Nutrient levels in streams are low, a result of low population density in the area, limited farming, and non-carbonate rocks. In particular, the study area is located on Pine Mountain, which in Kentucky follows a northeast to southwest path, — stretching 177 km from Breaks Interstate Park to the Kentucky-Tennessee border. Much of the mountain is the geographic border between Kentucky and Virginia. The northwest facing Monclinal Mountain consists of a steep over- thrust fault, exposing chunks of limestone bedrock and scattered limestone screes. In contrast, the southeast face of Pine Mountain is gentler, exposing only the Pennsylvanian sandstones, shale and siltstones. The northern slope was significant for our current survey because it provided the necessary carbonate soils to anchor calciphile land snail species not found on the acidic southeast slopes of Bad Branch. Pine Mountain forests are variable, depend- ing on soils and slope exposure (Braun 1935). Both of these variables are based upon geological structure, as the mountain is a monoclinal ridge of strongly dipping strata. Braun (1935) considered four main commu- nities on Pine Mountain; the southeast slope, ravines of the southeast slope, the summit, and the northwest slope. As with land snails, plant communities vary greatly as one pro- ceeds across these communities. Mature soils on the mountain support mixed mesophytic forests while immature soils support a com- munity more dominated by pine, oak, and pine-oak groups. In the mixed mesophytic forest sections, sugar maple Acer saccharum Marshall, basswood Tilia americana L., and yellow buckeye Aescylus flava Solander are considered to be indicator species (Braun 1950). In addition, American beech Fagus Land Snail Diversity at Bad Branch—Dourson and Beverly grandifolia Ehrhart, cucumber magnolia Mag- nolia acuminate L., northern red oak Quercus rubra L., tuliptree Liriodendron tulipifera L., and white ash Fraxinus americana L. also are characteristic (Jones 2005). Historically the upper slopes and ridgetops of this region were dominated by oak Quercus sp. and American chestnut Castanea dentata (Marshall). However, the American chestnut component of the forest was decimated by blight during the early to mid 19th century. Pine species may also be prevalent on some ridges. Especially ubiquitous is Virginia pine Pinus virginiana Miller and shortleaf pine P. echinata Miller. Ravines in this area are typically dominated by eastern hemlock Tsuga canadensis (L.), with an understory of rhodo- dendron Rhododendron species. The forests in this region are considered to be the most biologically diverse of any in the United States (Jones 2005). METHODS The land snail compilation of Bad Branch included collections on both the northwest and southeast slopes of Pine Mountain, and we attempted to survey in as many common to uncommon habitat types as we could find in Bad Branch State Nature Preserve. To best accomplish this task, nineteen sites (each around 0.25 ha) were established, including 5 limestone sites located on the steep north slope of Pine Mountain and 14 sandstone sites located on the more gentle south slopes of the last-named mountain. Fewer samples were taken from the limestone substrata, because it represented only about 41 ha or 3.8% of the totall069 ha that comprises Bad Branch State Nature Preserve. A stratified random sampling method was used for site selection, a technique that focused largely on finding as many variations in both common habitat (e.g., leaf litter) to uncommon habitats or microhabitats (e.g., seeps, tree crotches, fallen and rotting hard- wood trees in advanced stages of decay, depressions of moist leaf litter and or screes) occurring across the preserve. Shells of larger snail species (>5 mm) were hand-picked from under leaf litter, rocks, and logs. When outcrops of rock were encountered, snails were gleaned from the surfaces or from under small overhanging ledges where shells often 105 accumulate. Slugs were collected from under the exfoliating bark of standing dead trees and rotting hardwood logs in advanced stages of decay. To best represent the micro-snails (<5 mm) of Bad Branch, 19 cm? stratified random soil/leaf samples were also collected, each of these occurring within one of the 19 sites. The samples were thoroughly dried and then sifted through a series of sieves ranging from 4.76-0.50 mm. Snails were removed and sorted. All recovered and identifiable shells were assigned to species using Burch (1962), Pilsbry (1940, 1946, 1948), the primary author’s reference collections, and various recent publications. Taxonomy follows Tur- geon et al. (1998). New county records were determined using Branson (1973), Hubricht (1985) and Branson and Batch (1988). The species composition was analyzed with non-metric multidimensional scaling (NMDS). NMDS is an ordination method that uses an iterative approach for finding the best position of n objects in a k-dimensional space. An n X n distance matrix is calculated from a site (p) X species (s) matrix, so the ordination can be based on either position of sites or species. Because it uses rank-distances, it avoids the assumption of linear relationships among variables (Clarke 1993; McCune and Grace 2002). Site X species matrices are often very sparse (i.e., a large number of cells contain zeroes), and, as a result, assumptions of linearity are inappropriate. Therefore, NMDS is often considered the best technique for unconstrained ordination of species or sites (McCune and Grace 2002). The distance matrix was based on the Jaccard index (J) of dissimilarity (1 — J), and the ordination was conducted on the distances between sites. The starting configuration was based on principal components analysis to reduce the possibility of the solution converging on a local rather than global minimum. The maximum number of iterations for convergence was set at 200. Plots of the distance in multidimen- sional space against similarity (Shephard’s diagrams) were used to evaluate stress in the ordination. The number of dimensions in the final solution was determined by a plot of stress against the number of dimensions (6 = k = 2) and a stress threshold <0.15 (Clarke 1993). 106 Differences in diversity and abundance of the snail fauna between limestone and sand- stone substrates were evaluated with one-way ANOVA. Separate ANOVA’s were conducted for each of three dependent variables; the number of individuals per site, number of species per site, and percentage of the total snail fauna collected per site. The dependent variables were standardized by site because of the highly uneven number of samples _be- tween the two substrates. Normal probability plots, histograms, and plots of studentized residuals vs. estimated values were used to evaluate whether assumptions of linearity, homoscedasticity, and normality were met. The assumptions were met in all three analyses. One outlier (5.14 standard deviation units) was detected in the analysis of individ- uals therefore two analyses were conducted for individuals per site; one with and one without the outlier. The results of the analyses were similar, so we retained the record for the final analysis. RESULTS We collected 825 land snail specimens (454 shells from limestone and 371 shells from sandstone) representing 14 families, 35 gen- era, and 64 species (Table 1). Of these, 25 taxa had not been reported from Letcher County by Branson (1973), Hubricht (1985) or Branson and Batch (1988). Twenty-eight snail species or 45% of the fauna were species under 5 mm. The highest species diversity for a single site came from the north side of Pine Mountain in limestone substrata that har- bored 28 land snail species, while the lowest yielding site occurred in sandstone with only 6 species represented. No snails listed as endangered, threatened, or special concern by the Kentucky State Nature Preserves Com- mission (2005) were revealed during the survey. Stress in the non-metric multidimensional scaling analysis reached an acceptable level (0.118) with a three-dimensional solution. The ordination of the 19 sites indicated strong differences in the species composition of the land snail fauna between limestone and sandstone substrates (Figure 4). Approximate- ly half of the species were found in one substrate, with 15 occurring in sandstone and 21 occurring in limestone. Seven of the Journal of the Kentucky Academy of Science 69(2) terrestrial gastropods found in the limestone, Glyphyalinia cumberlandiana (Clapp), Hen- dersonia occulta (Say), Pomatiopis lapidaria (Say), Gastrocopta contracta (Say), Gastro- copta pentodon (Say), Gastrocopta corticaria (Say) and Strobilops labyrinthica (Say) are reported to be calciphile species (Hubricht 1985). Twenty-eight species were less dis- criminatory however, occurring on both sides of Pine Mountain usually in similar situations, under leaf litter, rocks, logs or the loose bark of dead trees in advance stages of decompo- sition. The ordination of the 14 sandstone sites also indicated strong differences in the species composition of the land snail fauna found within the sandstone substrata. Abundance (individuals/site), species rich- ness (species/site), and percent species in the total sample were 2 times to almost 4 times greater on limestone substrate than sandstone (Figure 5; P = 0.001). This suggested that more sandstone habitat was required to sustain the same number of snail species as limestone and that land snail diversity in the sandstone regions of Bad Branch were work- ing at a much larger ecological scale than in the limestone. Species of Special Interest Approximately half of the snail fauna at Bad Branch were considered wide ranging species; the remaining terrestrial gastropods found have biogeographical affinities to other phys- iographic provinces. Land snail assemblages at Bad Branch that had associations with the Great Smoky Mountains were Paravitrea lamellidens, P. placentula Shuttleworth, Ven- tridens lawae (W. G. Binney), P. subtilis Hubricht, V. collisella (Pilsbry), V. theloides (Walker and Pilsbry), Anguispira mordax Shuttleworth (Figure 6), Mesodon normalis (Pilsbry), Carychium clappi Hubricht, Cio- nella morseana (Doherty), and Glyphyalinia caroliniensis (Cockerell). Species that were found to be largely associated with the Blue Ridge, specifically the states of Virginia and West Virginia included Philomycus venustus Hubricht, P. flexuolaris Rafinesque, Paravi- trea multidentata (A. Binney), Paravitrea dentilla, Stenotrema hirsutum (Say), Neohelix albolabris (Say), Glyphyalinia cumberlandi- ana, Triodopsis anteridon Pilsbry and Hen- dersonia occulta. Finally, a measure of the Land Snail Diversity at Bad Branch—Dourson and Beverly snails from Bad Branch survey including Vertigo gouldi (A. Binney), Euconulus fulvus (Miiller) a Holarctic species, Mesomphix inornatus (Say), M. cupreus (Rafinesque), Euchemotrema fraternum (Say), Striatura ferrea Morse, Pallifera dorsalis (A. Binney), Allogona profunda (Say), and Appalachina sayana (Pilsbry) had their affinities with more northerly faunas (Branson and Batch 1968, 1988; Hubricht 1985). Paravitrea lamellidens (new state record) was an interesting addition to Bad Branch for several reasons. Hubricht (1985) showed this snail’s range to center more or less around the Great Smoky Mountains and its habitat to include pockets of deep, moist leaf litter on wooded hillsides of lower to higher elevations forests. It also has a strong affinity to rock talus. The discovery of this species at Bad Branch is a range extension of more than a hundred miles to the north (Hubricht 1985). Specimens were secured from two sites, one from rock talus and another from under moist leaf litter, at elevations of 489 m to 712 m respectively. Both sites occurred in the sandstone bedrock associated habitats. Al- though Branson and Batch (1968, 1988), Petranka (1982), Hubricht (1983) and others conducted concentrated site investigations of Kentucky’s southeastern mountains, mostly in the counties of Bell, Harlan and western Letcher, the species was not found. This suggests that the range of P. lamellidens in Kentucky may be more easterly along Pine Mountain, including eastern Letcher County and perhaps Pike County as well. These two counties are among the least sampled in southeastern Kentucky. This was well illus- trated by Hubricht (1985) distribution records that showed many gaps for common snail species in these two counties. Bad Branch alone produced 23 new records for Letcher County, further illustrating this information deficit. The thought that P. lamellidens occurs more easterly in Kentucky is based on the fact that the species was not found in Bell and Harden Counties despite intensive surveys and a record for P. lamellidens exists less than 50 miles to the east in Craig County, Virginia, in the Jefferson National Forest (Dourson, unpublished data). Paravitrea dentilla (new state record) is another unexpected snail to Kentucky’s land 107 snail fauna. Hubricht (1985) showed the species only from Washington County, Virgi- nia, and McDowell County, West Virginia. It is found in leaf litter on river bluffs (Hubricht 1985). At Bad Branch, P. dentilla was discovered at only one site in moist leaf litter around limestone rock talus occurring on the north slope of Pine Mountain. Four speci- mens were retrieved during our survey suggesting its overall rarity at Bad Branch. Given that it has not been reported by other collectors from the southeastern mountains of Kentucky, it likely represents one of the state’s rarer snails. Two other species of snails found during the survey deserving particular mention in- clude Vertigo parvula Sterkii and Striatura ferrea. Vertigo parvula was first reported in Kentucky by Branson and Batch (1988). They collected one specimen from a moist wooded hillside, 7.3 km north of Richmond on SR 60 (Tates Creek Road), Madison County, Ken- tucky. Although their record from Madison County seems an unlikely site for the species, there are in fact other disconnected popula- tions reported. The primary distribution of V. parvula lies along the Appalachian divide in Virginia, with one relatively isolated location reported from Carter County, Tennessee, and one exceptionally isolated population occur- ring in Summit County in northeastern Ohio (Hubricht 1985). In our study, V. parvula was found from two sites (a total of three specimens) located on the north side of Pine Mountain under leaf litter, at the base of limestone outcrops. The species was scarcely detected by leaf litter sampling. It is a cryptic shell of diminutive size (1.5 mm) remaining well concealed among debris. Vertigo parvula had fewer shells collected than any other species, suggesting that it is one of the rarer snails of Bad Branch or that the collecting methods or locations we used were inade- quate at capturing this species. According to Pilsbry (1948), this is a relatively rare snail. Harlan is the only county in Kentucky previously reported to have Striatura ferrea (Pilsbry 1946; Petranka 1982; Hubricht 1985). Even though Branson and Batch (1988) conducted an extensive survey of Kentucky, sampling terrestrial gastropods from 55 coun- ties, documenting 19 families, 45 genera and 138 species, they did not find S. ferrea. The 108 Journal of the Kentucky Academy of Science 69(2) Table 1. Land snail species documented at Bad Branch Nature Preserve, Letcher County, Kentucky. The ranking system under status are by the authors and do not reflect official state ratings. Keys to table: Status: KR= Kentucky Rare, LR=Limited Range in Kentucky, CO=Common; Record: SR=State Record, CR=County Record; Affinities: GSM=Great Smoky Mountains, VWV=West Virginia/Virginia, NOR= Northern, CKY= Central Kentucky, WID=Wi- despread in eastern North America. Species Substrata Habitat Status Record Affinities HELICINIDAE Hendersonia occulata (Say, 1831) Limestone Leaf litter LR VWV POMATIOPSIDAE Pomatiopsis lapidaria (Say, 1817) Both Moist litter CO WID CARYCHITDAE Carychium clappi Hubricht, 1959 Both Moist litter CO GSM Carychium exile H. C. Lea Both Moist litter CO WID Carychium nannodes Clapp, 1905 Limestone Moist litter CO WID COCHLICOPIDAE Cochlicopa morseana (Doherty, 1878) Both Leaf litter CO GSM PUPILLIDAE Columella simplex (Gould, 1821) Limestone Leaf litter CO WID Gastrocopta contracta (Say, 1822) Limestone Leaf litter CO CR WID Gastrocopta corticaria (Say, 1816) Limestone Leaf litter CO CR WID Gastrocopta pentadon (Say, 1821) Both Leaf litter CO CR WID Vertigo gouldi (A. Binney, 1823) Both Leaf litter CO NOR Vertigo parvula Sterkii, 1890 Limestone Leaf litter KR CR VWV STROBILOPSIDAE Strobilops labrythinca (Say, 1817) Limestone Leaf litter CO CR. WID HAPLOTREMATIDAE Haplotrema concavum (Say, 1821) Limestone Leaf litter CO WID PUNCTIDAE Punctum minutissimum (I. Lea, 1841) Both Seeps CO CR WID Punctum vitreum H. B. Baker, 1930 Limestone Moist litter KR CR WID HELICODISCIDAE Helicodiscus notius Hubricht, 1962 Limestone Leaf litter CO CR WID DISCIDAE Anguispira mordax (Shuttleworth, 1852) Limestone Leaf litter CO CR GSM Discus patulus (Deshayes, 1830) Both Rotting wood CO WID PHILOMYCIDAE Pallifera dorsalis (A. Binney, 1842) Limestone Moist talus CO CR NOR Philomycus flexuolaris Rafinesque, 1820 Both Under bark CO VWV Philomycus venustus Hubricht, 1953 Sandstone — Under bark LR VWV HELICARIONIDAE Euconulus dentatus (Sterkii 1893) Limestone Leaf litter CO CR WID Euconulus fulvus Miiller, 1774 Limestone Leaf litter CO NOR Guppya sterkii (Dall, 1888) Both Leaf litter CO WID ZONITIDAE Mesomphix cupreus (Rafinesque, 1831) Both Leaf litter CO NOR Mesomphix inornatus (Say, 1821) Both Leaf litter CO NOR Mesomphix perlaevis (Pilsbry, 1900) Both Leaf litter CO WID Hawaiia miniscula (A. Binney, 1840) Limestone Leaf litter CO CR WID Paravitrea dentilla Hubricht, 1978 Limestone Rock structure KR, LR SR VWV Paravitrea lamellidens (Pilsbry, 1898) Sandstone Rock talus KR, LR SR GSM Paravitrea multidentata (A. Binney, 1840) Both Seeps CO VWV Paravitrea placentula (Shuttleworth, 1852) Both Leaf litter LR GSM Paravitrea subtilis Hubricht 1978 Sandstone Leaf litter KR, LR CR GSM Glyphyalinia indentata (Say, 1823) Limestone Leaf litter CO WID Glyphyalinia cumberlandiana (Clapp, 1919) Both Rock talus CO VWV Glyphyalinia wheatleyi (Bland, 1883) Sandstone Leaf litter CO CR WID Glyphyalinia caroliniensis (Cockerell, 1890) Sandstone Leaf litter CO CR GSM Gastrodonta interna (Say, 1822) Limestone Rotting wood CO WID Striatura meridionalis (Pilsbry & Feriss, 1906) Both Moist litter CO CR WID Striatura ferrea Morse, 1864 Sandstone _— Seeps KR, LR CR NOR Ventridens demissus (A. Binney, 1843) Sandstone Leaf litter CO WID Ventridens collisella (Pilsbry, 1896) Sandstone Leaf litter LR CR GSM Land Snail Diversity at Bad Branch—Dourson and Beverly 109 Table 1. Continued. Species Substrata Habitat Status Record Affinities Ventridens lawae (W. G. Binney, 1892) Both Leaf litter LR CR GSM Ventridens ligera (Say, 1821) Sandstone Leaf litter CO CR WID Zonitoides arboreus (Say, 1816) Sandstone — Under bark CO WID POLYGYRIDAE Allogona profunda (Say, 1821) Limestone Leaf litter CO NOR Euchemotrema fraternum (Say, 1824) Sandstone Leaf litter CO CR WID Inflectarius rugeli (Shuttleworth, 1852) Both Leaf litter CO WID Inflectarius inflectus (Say.1821) Sandstone Leaf litter CO WID Appalachina sayana (Pilsbry, 1906) Both Leaf litter CO NOR Mesodon normalis (Pilsbry, 1900) Sandstone Leaf litter LR CR GSM Mesodon thyroidus (Say,1816) Sandstone Leaf litter CO WID Mesodon zaletus (A. Binney, 1837) Limestone Leaf litter CO CR WID Neohelix albolabris (Say, 1816) Both Leaf litter CO WID Patera appressa Say, 1821 Both Rock structure CO WID Stenotrema angellum Hubricht, 1958 Both Leaf litter CO CKY Stenotrema edvardsi (Bland, 1856) Both Leaf litter CO WID Stenotrema hirsutum (Say, 1817) Both Leaf litter CO VWV Stenotrema stenotrema (Pfieffer, 1819) Sandstone Leaf litter CO WID Triodopsis anteridon Pilsbry, 1940 Both Leaf litter CO VWV Triodopsis tennesseensis (Walker & Pilsbry, 1902) Both Leaf litter CO WID Xolotrema denotatum (Ferussac, 1821) Limestone — Rotting wood CO CR WID species has its main affinities with the northern states as far north as Maine. In the _ sites, found in the sandstone substrata among southern Appalachians, it is a species of higher moist leaf litter. At one site, two shells of S. elevation hardwood forests, usually found ferrea along with nine other species of snails under moist leaf litter (Hubricht 1985). At were collected from moist leaf litter found in Bad Branch, the species was taken from three 1.4 Site12 1.3 Site19 Site15 Sandstone sites 0.3 Site/8 0.2 Sited2 Axis 2 i?) -0.5 : 0.5 Siteos Site08 Limestone sites -1.6 -1.4 -1.2 -1 -0.8 -0.6 -0.4 -0.2 (6) 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 Axis 1 Figure 4. Non-metric multidimensional scaling ordination of nineteen sites where the land snail fauna was sampled at Bad Branch State Nature Preserve in Letcher County, Kentucky. 110 100 ~ (o>) oO (o) (=) (o) Individuals Per Site NO io) Sandstone Limestone Substrate Ol io) aN (eo) —e— ie) io) Percent Species Per Site Sandstone Limestone Substrate Journal of the Kentucky Academy of Science 69(2) 30 Species Per Site on Sandstone Limestone Substrate Figure 5. Abundance (individuals/site), species richness (species/site), and percent species/site of land snails on two substrates at Bad Branch State Nature Preserve in Letcher County, Kentucky. the crotch of a forked poplar tree about a meter off the ground. Although this seems an uncharacteristic habitat for land snails, many single tree crotches in various eastern states have in fact harbored multiple species, as many as 15 taxa in rare cases. A study in the Great Smoky Mountain National Park showed that a number of infrequent land snails, (including an undescribed Carychium sp.) are occasionally found in tree crotches, sometimes positioned as high as 4 m off the ground (Dourson and Dourson 2006). Most of the snails found in these situations are micro species (<3 mm) including but not always limited to the genera Carychium, Gastro- copta, Vertigo, Punctum, Striatura and Para- vitrea. Other snail species that are intermit- tently arboreal such as Anguispira jessica have been documented as high as 21 m in tulip trees found in the Great Smoky Mountains (Keller 2002). A species occurring rather frequently in the Cumberland Plateau and on Pine Mountain in eastern Kentucky is Appalachina sayana. An Land Snail Diversity at Bad Branch—Dourson and Beverly Figure 6. Anguispira mordax, unique for its prominent, widely spaced ribs and color features. It is one of twelve land snail species found at Bad Branch Nature Preserve that has biogeographical affinities with the Great Smoky Mountains. Two standard views are displayed; shells are 18 mm in diameter. interesting anomaly found in the species has been inadequately discussed in past literature. Although considered a common snail in Kentucky, there are in fact two relatively distinct forms found in the state. Their shell morphologies are dissimilar enough to allow easy separation. One form of A. sayana, the larger and more common of the two is found on the Cumberland Plateau (here referred to as A. sayana CP for Cumberland Plateau, Figure 7a). The smaller form (here referred to as A. sayana PM for Pine Mountain, Fig- ure 7b) is found on Pine Mountain. A. sayana CP averages 3-4 mm larger than A. sayana PM and usually displays both the basal and parietal tooth. A. sayana PM usually displays only the basal tooth and has a thin wire-like lip, remaining somewhat concave in shape its entire length. The umbilicus of A. sayana CP (Figure 8a) is umbilicate where as the umbi- licus of A. sayana PM (Figure 8b) is more or less rimate. Although these two forms were 111 considered the same species by Pilsbry (1940), Hubricht (1985), and others, A. sayana PM represents at the very least an interesting ecological form to the southeastern mountains of Kentucky. Seven snail species that are of interest due to their infrequency or limited range in Kentucky include Punctum vitrem H. B. Baker, Ventridens theloides: form, nodus (Walker and Pilsbry), V. collisella (Pilsbry), V. lawae, Philomycus venustus (Figure 9), Paravitrea subtilis and Hendersonia occulta. Two species found near the Bad Branch parking lot, Mesodon thyroidus (Say) and Inflectarius inflectus (Say), although native to Kentucky, can be indicators that the natural vegetation has been disrupted, which is certainly the case for that site. Neither of these species was found beyond disturbance locales. The discovery of Stenotrema angellum at Bad Branch is an interesting anomaly given that the species primary range occurs in central Kentucky (Hubricht 1985). Specimens of S. angellum found at Bad Branch were smaller, more globose in profile, and generally more hirsute than S. angellum from central Kentucky. DISCUSSION In the early 1900s, Pilsbry considered the broad Valley and Ridge Provinces of eastern Tennessee to be a partial barrier to the intermingling of snail faunas occurring in the Cumberland Plateau (which at that time was more or less grouped with Cumberland and Pine Mountains) with snails from the area of Roan Mountain to the Great Smoky Moun- tains. Collections since by Branson and Batch (1967, 1968, 1988), Petranka (1982), Hubricht (1983), and others have ameliorated the barrier concept, revealing some interesting biogeographical associations. The principal land snail affinities of the southeastern moun- tains of Kentucky are in fact with the Great Smoky Mountain-Blue Ridge physiographic sections of the Cumberland Province, with some divisions of the snail fauna showing relationships with Virginia and West Virginia. Other examples of these broader biogeo- graphical associations at Bad Branch include two small mammals, Sorex dispar, long-tailed shrew and Microsorex thompsoni, Thompson’s pigmy shrew (Caldwell 1980). A small portion 112 Basal tooth Journal of the Kentucky Academy of Science 69(2) b. Parietal tooth Lip Figure 7a—b. Standard side view comparisons between two forms of A. sayana found in eastern Kentucky. The larger and more common of the two forms is found on the Cumberland Plateau (7a, shell diameter 26 mm), the smaller form (7b, shell diameter 16 mm) is found on Pine Mountain and at Bad Branch. Key features for separating the two forms include the parietal tooth (or the absence of it) and their relative size to each other. of the Pine Mountain snails have their affinities with more northerly faunas. Of the 64 species found at Bad Branch, approximate- ly eleven including Paravitrea lamellidens are representative of southern mountain ranges and eight including Paravitrea dentilla are representative with West Virginia and Virginia with around nine having their affinities with more northerly faunas (Table 1). Most of the remaining snails found at Bad Branch are considered to be relatively wide-ranging Figure 8a-b. species across eastern North America, show- ing no close affiliation for a specific region. Besides the interesting biogeographical associations found at Bad Branch, the pre- serve also harbors a rich snail fauna, a result of several geophysical and ecological factors. Land snail abundance, both in terms of diversity of species and numbers of shells, is often highly correlated with increasing soil or litter pH (Burch 1955; Valovirta 1968; Ware- born 1970; Hotopp 2002), soil moisture b. Umbilicus Standard bottom view comparisons between the two forms of A. sayana found in eastern Kentucky, the larger and more common of the two is found on the Cumberland Plateau (8a , shell diameter 26 mm), the smaller form (8b, shell diameter 16 mm) is found on Pine Mountain and at Bad Branch. Bottom views show several key features for separating shells including the umbilicus, lip and their relative size to each other. Land Snail Diversity at Bad Branch—Dourson and Beverly ssc SE SRE SEE SES Ss Soe see ese see Figure 9. Philomycus venustus is one of several land snail species found at Bad Branch Nature Preserve that have biogeographical affinities with the Blue Ridge, specifically the states of Virginia and West Virginia. Length of animal while crawling is 70 mm. (Boycott 1934; Getz 1974; Pollard 1975) and soil cations, particularly calcium (Burch 1956; Agocsy 1968; Atkins 1968; Petranka 1982). Calcium carbonate is required by land snails for regulation of bodily processes, reproduc- tion, but most importantly shell-building (Burch 1962; Fournie and Chetail 1984; Hickman et al. 2003). Litter Ca and salt types can directly increase growth rates and fecun- dity in some species (Wareborn 1969, 1970, 1979). As would be expected, land snail scarcity is associated with low soil pH Burch (1955), declining soil cations, specifically Ca (Petranka 1982), increasing coniferous pres- ence (Jacot 1935; Karlin 1961), and increasing elevation (Petranka 1982). The influence of pH on land snails is thought to be indirect, its main effect being a lowering of the amounts of soil cations, principally Ca (Karlin 1961; Cameron 1970). Calcium occurs in abundance on the north slope of Pine Mountain in the form of limestone outcrops, screes, and finally, the soils. The highest number of shells (454) were found here and are largely correlated to the calcium rich soils and to a lesser degree slope gradient, but in terms of species diversity, the limestone sites (49 species) were nearly equal to the sandstone sites (43 species). These results were not surprising given the large difference in samples between sandstone and limestone (almost three to one). As a result, we felt that conventional diversity indices were not appropriate for the analyses. Never- theless, the analyses we did use clearly showed greater diversity and different species compo- sition between the two substrates when standardized by site. There is a relationship 113 between shell abundance and species rich- ness, sO in one sense, it is not surprising that richness was so much greater on limestone than sandstone (because there were so many more individuals on limestone). But this is much more than a sampling artifact, as there is also a relationship between area and species richness. Given that sandstone had nearly 3 times the number of samples as limestone (hence almost 3 times the area sampled), it would have been expected a greater total number of individuals and species on sand- stone if it had more or less the same levels of diversity as limestone. This was not the case, and it speaks even more to how much greater diversity is on limestone than sandstone. From a management standpoint, protected areas on sandstone substrata would need to be larger in order to preserve similar snail diversity found in limestone. Because many of the calciphile species found on limestone will not likely occur on sandstone and vice-versa, protecting substrata diversity, at least in terms of sustaining snail multiplicity is equally impor- tant. Another factor that has likely increased, to some extent, the numbers of snails on the steep north slopes of Pine Mountain was the slope gradient. Coney et al. (1982) found more species of land snails on steep slopes than on more moderate ones. Petranka (1982) found that 15 of the 56 land snail species found on Black Mountain showed some preference for slope, with 9 species showing an affinity for increasing slope. The impor- tance of leaf litter moisture (thought to be a factor of slope) to land snails was emphasized by Boycott (1934), Getz (1974), Pollard (1975), and others. Although aspect was reported to markedly affect microclimate (Braun 1940, 1942; Geiger 1965), Petranka’s (1982) study found no environmental variable to be significantly correlated with aspect. With respect to elevation, Petranka (1982) reported that pH, potassium, calcium, and magnesium levels would decrease (ppm) with increasing elevation and that the number of snail species and the number of individuals found per site would also decrease with elevation. The results of our work showed no _ particular trend as a function of elevation. Petranka’s (1982) study plots (from 335 m to 1261 m), however, had more than double the elevation 114 ranges from which to sample than was found at Bad Branch (from 457 m to 914 m). Clearly, there have been a number of studies on snail abundance or their scarcity as it relates to soil cations, gradient, elevation, vegetation, and litter moisture. Less well- known is the effect that large geophysical landscape edges have on driving biodiversity, particularly in land snails. The Central Knob- stone Escarpment in Powell County, Ken- tucky, forms a large geophysical landscape edge on Furnace Mountain where the Cum- berland Plateau, the Knobs, and the Outer Bluegrass regions of Kentucky converge. The merging of these distinct regions were shown to have an exceptional number of land snails, a reported 61 species found co-existing within a 2-ha mesic hillside (Dourson 2007). Analo- gous to Furnace Mountain, Bad Branch State Nature Preserve also forms a great “geophys- ical landscape edge” located along the Pine Mountain massif of southeastern Kentucky, bordered to the south by the Valley and Ridge Provinces and to the north by the Cumber- land Plateau. All three regions contain their own snail affiliations. Moreover, the amalgam- ation of these snail rich eco-regions has provided a number of terrestrial gastropods an opportunity to coexist. This large landscape edge, together with the wide range of soil types, elevation and slope gradient has brought together an interesting and_ rare assemblage of species not frequently occur- ring in the state of Kentucky. While the majority of land snails found at Furnace Mountain study were by and large common and wide ranging species, the snails at Bad Branch showed tighter affiliations with a particular region of eastern North America and were generally rarer land snail species. When compared with other collections made in southeastern Kentucky, Bad Branch exceeded their reported numbers. The most limited collection was by Pilsbry (1940) who secured 22 land snail species (unknown numbers) from a single locality in Harlan County, Kentucky. Branson and Batch (1968) reported 47 species (528 individuals) in Bell (3 sites) and Harlan (1 site) counties. A number of their snail accounts however, may have been erroneous. In Kentucky for example, Triodopsis fosteri (F. C. Baker) and Webbhe- lix multilineata (Say), two species reported in Journal of the Kentucky Academy of Science 69(2) their collections, are restricted to counties along the Ohio River (Hubricht 1985). Other questionable species that Branson and Batch (1968) reported from Black Mountain includ- ed Glyphyalinia virginica (Morrison) (an endemic to northern Virginia) and Triodopsis fraudulenta (Pilsbry) (reported from eastern West Virginia, northern Virginia, and a few counties in Pennsylvania). Branson and Batch (1968) also reported one exotic slug Deroceras reticulatum, a native to Europe. The actual number of indigenous snails from their study is probably closer to 40 species. Petranka (1982) reported 56 land snails (12,464 indi- viduals) at 36 sites from Big Black Mountain in Harlan County, Kentucky and Wise Coun- ty, Virginia, adding 21 new county records for Harlan. Petranka’s study covered the largest and most extensively sampled area to date in the southeastern mountains of Kentucky. Hubricht et al. (1983) reported 43 species from scattered locations across Black and Pine Mountains in Harlan County. Snail collections made by Branson and Batch (1968), Petranka (1982) and Hubricht et al. (1983) covered larger sample areas than our study. Further investigations at Bad Branch State Nature Preserve no doubt will add a few more taxa. Additions would most likely include Vertigo bollesiana (Morse), Mesomphix rugeli (W. G. Binney), Zonitoides elliotti (Redfield), Vitrinizonites latissimus (Lewis), Neohelix dentifera (A. Binney), Triodopsis tridentata (Say), Triodopsis vulgata Pilsbry, Triodopsis juxtidens (Pilsbry), Strobilops aenea Pilsbry, Philomycus togatus (Gould), Pallifera secreta (Cockerell) and Discus nigrimontanus (Pilsbry). These species are reported from Harlan County on Pine and Black Mountains (Pilsbry 1940, 1946, 1948: Hubricht 1983, 1985). A number of these land snails also have their affinities with the Great Smoky Mountain-Blue Ridge sections of the Cumberland Province, and Bad Branch has certainly demonstrated these same affinities. The real number of land snail species at Bad Branch State Nature Preserve is probably closer to 70. ACKNOWLEDGEMENTS We would like to express our gratitude to the following persons and_ organizations: Kentucky Nature Preserves Commission for granting permission to study the land snail Land Snail Diversity at Bad Branch—Dourson and Beverly fauna of Bad Branch Nature Preserve; Dr. Ronald S. Caldwell, Lincoln Memorial Uni- versity, and Dr. Steven Brewer, University of North Carolina at Wilmington, for serving as peer reviewers; Maggie Schmitt for field assistance in collecting snails; Dr. Robert Klinger, USGS for his assistance in statistical data analysis; and finally, Judy Dourson for her word processing skills. REFERENCES Agocsy, P. 1968. Data to quantitative conditions in the mollusk faunas of two different substrates in Central Hungary. Acta Zoologica Academiae Scientiarum 14: 16. Atkins, C. G. 1966. Factors affecting the structure and distribution of terrestrial Pulmonata. Proceedings of Iowa Academy of Science 73:408—-416. Boycott, A. E. 1934. The habitats of land mollusca in Britain. Journal of Ecology 22:1-38. Branson, B. A. 1973. Kentucky land mollusca: checklist, distribution and keys for identification. Kentucky Department of Fish and Wildlife Resources, Frankfort, KY. Branson, B. A., and D. L. Batch. 1968. Land snails from Pine and Big Black Mountains, Kentucky. Sterkiana B24=4 1. Branson, B. A., and D. L. Batch. 1970. An ecological study of valley-forest gastropods in a mixed mesophytic situation in northern Kentucky. The Veliger 12:333- 350. Branson, B. A., and D. L. Batch. 1988. Distribution of Kentucky land snails (Mollusca: Gastropoda). Transac- tions of the Kentucky Academy of Science 49:101-116. Braun, E. L. 1935. The Vegetation of Pine Mountain, Kentucky: An Analysis of the Influence of Soils and Slope Exposure as Determined by Geological Struc- ture. American Midland Naturalist 16:517—565. Braun, E. L. 1940. An ecological transect of Black Moun- tain, Kentucky. Ecological Monographs 10:193-241. Braun, E. L. 1942. Forests of the Cumberland Mountains. Ecological Monographs 12:413-447. Braun, E. L. 1950. Deciduous forests of eastern North America. Hapner, New York. Burch, J. B. 1955. Some ecological factors of the soil affecting the distribution and abundance of land snails in eastern Virginia. The Nautilus 69:26-29. Burch, J. B. 1956. Distribution of land snails in plant associations in eastern Virginia. The Nautilus 70:60-64, 102-105. Burch, J. B. 1962. How to know the eastern land snails. Wm. C. Brown Company Publishers, Dubuque, Iowa. Burch, J. B., and T. A. Pearce. 1990. Terrestrial gastropoda. Pages 201-231 in D. L. Dindal (ed). Soil Biology Guide. John Wiley, NY. Cameron, R. A. D. 1970. Differences in the distributions of three species Helicid snails in the limestone district 115 of Derbyshire. Proceedings of the Royal Society of London 176:130-159. Clarke, K. R. 1993. Non-parametric multivariate analyses of changes in community structure. Australian Journal of Ecology 18:117-143. Coney, C. C., W. A. Tarpley, J. C. Warden, and J. W. Nagel. 1982. Ecological studies of land snails in the Hiwassee River basin of Tennessee, U.S.A. Malacolog- ical Review 15:69-106. Dourson, D. 2007. A selected land snail compilation of the Central Knobstone Escarpment on Furnace Mountain in Powell County Kentucky. Journal of the Kentucky Academy of Sciences 68:119-131. Dourson, D., and J. Dourson. 2006. Land Snails of the Great Smoky Mountains (Eastern Region). Appalachian Highlands Science Learning Center and Great Smoky Mountains National Park. Dourson, D., and K. Feeman. 2006. A survey of terrestrial mollusca in selected areas of the Land Between the Lakes National Recreation Area. Journal of the Kentucky Academy of Sciences 67:9-18. Fournie, J., and M. Chetail. 1984. Calcium dynamics in land gastropods. American Zoologist 24:857-870. Geiger, R. 1965. The climate near the ground. Harvard University Press, Cambridge. Getz, L. L. 1974. Species diversity of terrestrial snails in the Great Smoky Mountains. The Nautilus 88:6-9. Hickman, C. P., L. S. Roberts, and A. Larson. 2003. Animal diversity. 3rd ed. McGraw-Hill, NY. Hotopp, K. H. 1999. Land snails from the mouth of Cornwell Cave, Preston County, West Virginia. Report to the Nature Conservancy West Virginia Chapter. Elkins, West Virginia. Hotopp, K. H. 2002. Land snails and soil calcium in central Appalachian Mountain Forest. Southeastern Naturalist 1:27—44. Hubricht, L. 1956. Land snails of Shenandoah National Park. The Nautilus 70:1—15. Hubricht, L. 1964. Land snails from the talus of Kentucky and Tennessee. Sterkiana 16:34. Hubricht, L. 1968. The land snails of Mammoth Cave National Park, Kentucky. The Nautilus 82:24—28. Hubricht, L., R. S. Caldwell, and J. G. Petranka. 1983. Vitrinizonites latissimus (Pulmonata: Zonitidae) and Vertigo clappi (Pupillidae) from eastern Kentucky. The Nautilus 97:20-22. Hubricht, L. 1985. The distribution of the native land mollusks of the eastern United States. Fieldiana Publication 1359. Jacot, A. P. 1935. Molluscan populations of old growth forest in rewooded fields in Asheville basin of North Carolina. Ecology 16:603—605. Jones, R. L. 2005. Plant life of Kentucky: An Illustrated Guide to the Vascular Flora. The University Press of Kentucky, Lexington, KY. Kalisz, P. J., and J. E. Powell. 2003. Effect of calcareous road dust on land snails (Gastropoda: Pulmonata) and millipedes (Diplopoda) in acid forest soils of the Daniel 116 Boone National Forest of Kentucky, U.S.A. Report to US Forest Service Supervisors Office. Winchester, Kentucky. Karlin, E. J. 1961. Ecological relationships between vegetation in the distribution of land snails in Montana, Colorado, and New Mexico. American Midland Natu- ralist 65:60—66. Keller, H. W., and H. K. Snell. 2002. Feeding activities of slugs on Myxomycetes and macrofungi. Mycologia 94:757—760. Kentucky State Nature Preserves Commission. 2005. Rare and extirpated biota of Kentucky. Journal Kentucky Academy of Science 61:115-132. McCune, B., J. B. Grace, and D. L. Urban. 2002. Analysis of Ecological Communities. MjM Software Design, Gleneden Beach, Oregon, USA. Nekola, J. C. 1999. Terrestrial gastropod richness of carbonate cliff and associated habitats in the Great Lakes Region of North America. Malacologia 41:231—252. Petranka, J. G. 1982. The distribution and diversity of land snails on Big Black Mountain, Kentucky. Master of Science thesis. Eastern Kentucky University, Rich- mond, KY. Pilsbry, H. A. 1940. Land Mollusca of North America (north of Mexico). Volume I, Part II. Academy of National Science of Philadelphia, Philadelphia, PA. Pilsbry, H. A. 1946. Land Mollusca of North America (north of Mexico). Volume II, Part I. Academy of National Science of Philadelphia, Philadelphia, PA. Pilsbry, H. A. 1948. Land Mollusca of North America (north of Mexico). Volume II, Part II, Academy of National Science of Philadelphia, Philadelphia, PA. Pollard, E. 1975. Aspects of the ecology of Helix pomatia L. Journal of Animal Ecology 44:305-329. Journal of the Kentucky Academy of Science 69(2) Reid, F. A. 2006. Mammals of North America, 4th edition. Peterson Field Guide, Houghton Mifflin, New York. Taylor, D. D., G. Chalfant, J. MacGregor, J. Walker, R. Bergeron, C. Miller, and V. R. Bishop. 1997. Landtype association narratives. Report to USDA Forest Service Regional Office. Available from Daniel Boone National Forest. Winchester, KY. Turgeon, D. D., J. F. Quinn, Jr, A. E. Bogan, E. V. Coan, F. G. Hochberg, W. G. Lyons, P. M. Mikkelson, R. J. Neves, C. F. E. Roper, G. Rosenberg, B. Roth, A. Scheltema, F. G. Thompson, M. Vecchione, and J. D. Williams. 1998. Common and scientific names of aquatic invertebrates from the United States and Canada: Mollusks, 2nd ed. American Fisheries Society, Special Publication 26. Valovirta, I. 1968. Land mollusks in relation to acidity on hyperite hills in Central Finland. Annales Zoologici Fennici 5:245-253. Wareborn, I. 1969. Land mollusks and their environments in oligotrophic area in southern Sweden. Oikos 20:461-479. Wareborn, I. 1970. Environmental factors influencing the distribution of land mollusks of an oligotrophic area in southern Sweden. Oikos 2:285-291. Wareborn, I. 1979. Reproduction of two species of land snails in relation to calcium salts in the foerna layer. Malacologia 18:177—180. Wareborn, I. 1992. Changes in the land mollusc fauna and soil chemistry in an inland district in southern Sweden. Ecography 15:62-69. Woods, A. J., J. M. Omernick, W. H. Martin, G. J. Pond, W. M. Andrews, S. M. Call, J. A. Comstock, and D. D. Taylor. 2002. Ecoregions of Kentucky. U.S. Geological Survey, Reston, VA. J. Ky. Acad. Sci. 69(2):117-123. 2008. Molecular and Morphological Evidence for the Occurrence of Two New Species of Invasive Slugs in Kentucky, Arion intermedius Normand and Arion hortensis Férussac (Arionidae: Stylommatophora) Rory J. Mc Donnell! Applied Ecology Unit, Centre for Environmental Science, National University of Ireland, Galway, Ireland Timothy D. Paine and Richard Stouthamer Department of Entomology, University of California, Riverside, California 92521 Michael J. Gormally Applied Ecology Unit, Centre for Environmental Science, National University of Ireland, Galway, Ireland and James D. Harwood Department of Entomology, University of Kentucky, Lexington, Kentucky 40546-0091 ABSTRACT Arion intermedius Normand and Arion hortensis Férussac are invasive mollusks in North America, having been previously reported in thirteen and seven U.S. States, respectively. We report here the first records of both species in Kentucky. Because slug species within the genus Arion Férussac show high degrees of intra- specific variation, identifications were confirmed using both morphological and molecular (partial COI gene sequences) methods. These new records are of concern because invasive slugs are major pests in agriculture, horticulture and floriculture, causing considerable damage to wheat, alfalfa, corn, soybean and tobacco that are amongst the most economically important crops in Kentucky. This study also highlights the important need for additional gastropod surveys throughout the U.S.A. KEY WORDS: slugs, Arion hortensis, Arion intermedius, invasive species, agroecology, molecular diagnosis INTRODUCTION With the recent increases in global trade, the introduction of exotic gastropods contin- ues to pose serious problems throughout North America (Robinson and_ Slapcinsky 2005). Robinson (1999) listed 4,900 gastropod interceptions at U.S. ports from nearly 100 countries between 1993 and 1998. These records comprised 369 species from 197 genera in 71 families. Although the U‘S. Department of Agriculture, Animal and Plant Health Inspection Service (USDA-APHIS) screens shipments at U.S. seaports, airports, and border crossings, it is inevitable that some gastropods make it through due to limited human resources (Robinson 1999) and to the cryptic nature of many species, especially the small size of their eggs and hatchlings. There ' Corresponding author e-mail: rjmcdonnell@gmail. com currently are over 80 non-native snail and slug species with established, self-sustaining pop- ulations in the U.S. and Canada (excluding the Hawaiian Islands and Puerto Rico). Many of these species pose a significant threat to horticultural and agricultural industries in addition to the natural environment (Robin- son and Slapcinsky 2005). However, according to Reise et al. (2006), little has been published on introduced terrestrial gastropods within the U.S. over the past two decades. This is of concern because such introductions tend to dominate the malacofauna in anthropogenic habitats and as a result tend to be economically important as pests of agriculture and _horti- culture. Robinson (1999) suggested that sur- veys of urban and suburban areas by malacol- ogists will be crucial in identifying additional invasive gastropods and should enable mitiga- tion measures to be implemented more rapidly, thereby preventing the establishment Lie 118 Figure 1. of pestiferous species and consequently re- ducing eradication costs. In response to the need for faunistic evaluations of invasive gastropods, a survey was conducted in urban, suburban, agricul- tural, horticultural, floricultural and natural areas throughout Kentucky to document the native and invasive slug species and_ to ascertain the potential impact upon agricul- tural production. The surveys were designed to complement the existing knowledge of the slug fauna of Kentucky (Burch 1962; Branson and Batch 1969; Dourson and Feeman 2006) and resulted in the discovery of the first specimens of A. intermedius (Figure 1) and A. hortensis (Figure 2) in the state. METHODS AND MATERIALS Slug collections were made throughout Kentucky by hand collecting and using simple baited traps (Friskies® canned cat food covered with a black plastic refuse bag pinned at each corner and left overnight). Specimens were kept alive in small plastic containers lined with damp paper towels and fed on organic carrots. On return to the laboratory, specimens were identified using diagnostic morphological characteristics (outlined below) and then preserved in 100% ethanol. Because slugs, especially species within the genus Arion intermedius Normand from B erea College Forest, Madison Journal of the Kentucky Academy of Science 69(2) County, Kentucky. Arion Férussac, show high degrees of intra- specific variation (Pinceel et al. 2004), species- level identification was confirmed using mo- lecular methods. A portion of foot tissue was excised from 14 specimens of A. intermedius and the one individual of A. hortensis. DNA was then extracted using a Qiagen® commercial extrac- tion kit (Tissue protocol) (Qiagen Inc., Valen- cia, CA). Cell lysis was achieved overnight in a 56°C constant temperature shaker with pro- teinase K and Qiagen lysis buffer. The QIAamp Tissue Kit then adsorbed the DNA onto a silica membrane in a spin column that was cleaned with two wash buffers. Finally, purified template DNA was eluted in AE storage buffer. Using this DNA, a polymerase chain reactions (PCR) was used to amplify a 655 bp fragment of the mitochondrial cyto- chrome oxidase subunit I gene (Table 1). Amp- lified DNA was cleaned using the Wizard ® PCR Preps DNA Purification System (Pro- mega, Madison, WI). The PCR product was first purified using a purification buffer and resin, and then washed using 80% isopropanol. Clean DNA was eluted in double distilled water and then direct sequenced in both directions at the University of California, Riverside Genomics Institute Core Instrumen- tation Facility using an Applied BioSystems Two New Species of Arion in Kentucky—Mc Donnell et al. 119 i & oe aie} LO Te) = | Os o fe) ¢ Ep 35 & E aan o E cae) = < as % és oe 5O cs So oO fe Bs £ 19 6 =; ‘= Se n + a ee O a=) Ow x 2 St hese = as |8|e S808 yo Fe Faw ae v Ta) ite | => : iss) (cas i io 2 os ES 5a = & ro 22 | |z o 2 og ge | 12 =¢ |,|2- Be lela os lz n SS 1k E = fo) —_F a pies] a bal ~ ¢ S SC Au Oo S g8 je\87%5 2. s= lz| 337 a oe Bs le(/St3 82. a7 Qo FI5A4as d’—0 < Ero ie Ge TON & Oo me .S 0 Ss oe Om ee =. fe O& ahs =| ssa Ap | |\X@o aS ER =” fa S : eu. Figure 2. Arion hortensis Férussac from an urban & S eo a Bers garden in Lexington, Fayette County, Kentucky. S 4 ae) ea < ; é 5 8 © v 3730 DNA Analyser with a Big-Dye ® V3.1 kit so ~ = (Applied Biosystems, Foster City, CA). Raw - = ae MO | sequences were trimmed by removing the Oe | es primers and aligned manually in Bioedit v a Z 219 : 7.0.5.3 (Copyright© Hall 1997-2004). Using So |e : x these aligned sequences, a BLAST search on ge |5 < z GenBank was performed to confirm our 2S |E/E g initial, morphological identifications (Table & = 2 = 2). Sequences are deposited in GenBank aa = = under accession numbers EU382742 for A. es a O hortensis and EU382743-EU382756 for A. ge | |5 2 intermedius. 3-9 o a a — & COLLECTION DETAILS £6 ¢ < na ae a os o A total of 97 individuals of A. intermedius se | | 7 = (Figure 1) were collected at two locations in ES |g ~ the central Bluegrass region of Kentucky. The aca =3 xo 8 first location was Berea Forest (37°60'N, =i Se a = ae . . 1 84°25’W) in Madison County on 13 and 16 = § o& oS Apr 2007 (n = 70) and the second was <= he se 120 Table 2. Journal of the Kentucky Academy of Science 69(2) Molecular confirmation (BLAST) of the morphology based identifications for Arion intermedius Normand and Arion hortensis Férussac using a 655 bp fragment of the cytochrome oxidase subunit I mitochondrial gene. BLAST Results Species Location Specimens E-value? Identification* Ascension Numbers Arion hortensis Urban Garden 1 0.0 100% EU382742 Arion intermedius Berea Forest 1-10! 0.0 100% EU382743-—EU382752 Stonewall Park 14! 0.0 100% EU382753-EU382756 ' There were no nucleotide polymorphisms within the COI sequences for these individuals. > The E-value is a measure of the random background noise that exists for matches between sequences. The lower the E-value the more “significant” the match between the test sequence and the matching Genbank sequence. ’ The higher this percentage the greater the likelihood that the test species is the same as that predicted by the BLAST search. Stonewall Park (n = 4) in Lexington, Fayette County. Berea Forest is a secondary oak- maple forest interspersed with small stands of hickory, pine and mountain laurel. Specimens were predominantly collected under discard- ed carpet close to an abandoned shack (n = 66). A smaller number (n = 4) were found under dead wood; further surveys on 8 May 2007 captured additional specimens (n = 23) under wood of fallen trees on the edge of the forest floor and adjacent to other collection locations. Other species collected with A. intermedius were Deroceras laeve (Miiller) and Deroceras reticulatum (Miiller). Stonewall Park (38°00'117"N, 84°33'248"W) is a suburban park with managed and unman- aged grassland interspersed with stands of mixed deciduous trees. A stream also runs through the area. Specimens of A. intermedius were collected on 16 Apr 2007 under plywood and decaying logs. Deroceras laeve and D. reticulatum also were collected with A. inter- medius in Stonewall Park. A single specimen of A. hortensis (Figure 2) was collected under broken wood on a lawn in an urban garden in Lexington, Fayette Coun- ty, Kentucky, on 26 Mar 2007 (37°59'3311'N, 84°29'2597’"W). The only other species col- lected with this individual was D. reticulatum. MORPHOLOGICAL IDENTIFICATION Arion intermedius is a small slug up to 25 mm long (Barker 1999). The specimens that we collected in Kentucky were all grayish- yellow with distinctly darker tentacles (Fig- ure 1) and yellow body mucus. The species is best separated from other slugs by the presence of small spikes on the body tubercles that give it a prickly appearance when contracted (Kerney and Cameron 1979). Internally, the genital atrium lacks a stimula- tor (Quick 1960). Arion hortensis is a small slug up to 50 mm long. The body color tends to be dark blue- black (Figure 2), but the sole and body mucus are yellow to orange and sticky (Barker 1999). According to Pfleger (1999) the tip of the tail is the same color as the mucus. However, Davies (1977, 1979) confirmed that A. hortensis s.l. is a species complex comprising A. hortensis s. s., Arion distinctus Mabile, and Arion owenii Davies. The three species are difficult to distinguish on external characters alone, and for that reason dissection and/or molecular analysis should be used to confirm identifica- tions. Because A. distinctus has been reported in other U.S. states (Mc Donnell et al. 2009), it is possible that it also occurs in Kentucky. We therefore provide details of how to reliably identify both species. According to Backeljau and Van Beeck (1986), the shape of the epiphallus structure (i.e., the structure associ- ated with the outlet of the epiphallus in the genital atrium) is the most reliable diagnostic character. In A. distinctus, it is conical, protrudes into the atrium, and covers the outlet of the epiphallus. A gutter or fissure runs from the margin of the epiphallus structure to its centre (Backeljau and Van Beeck 1986). In A. hortensis this structure is a relatively incon- spicuous, oblong plate that covers about half of the epiphallus outlet. It never has a gutter/ fissure (Backeljau and Van Beeck 1986). Arion owenii appears to be the rarest member of the complex (Backeljau and Van Beeck 1986), and there are no records from the U.S (Mc Donnell et al. 2009). The epiphallus structure in this species is variable but it is generally, long, slender, tongue-like and it protrudes from the outlet of the epiphallus. Two New Species of Arion in Kentucky—Mc Donnell et al. DISTRIBUTION AND ECOLOGY Although the native range of A. intermedius is central and western Europe (Barker 1999), it is found as an exotic throughout the world: the Azores (Quick 1960), Australia, Europe, New Zealand, North America, South Africa (Barker 1999), Canada, North Africa, and Polynesia (Forsyth 2004). In the U.S. it has been collected in California, Connecticut, Hawaii, Idaho, Maine, Massachusetts, Michi- gan, New Hampshire, New York, North Carolina, Rhode Island, Tennessee, and Wa- shington (Chichester and Getz 1969; Dundee 1974; Pearce and Blanchard 1992; Pearce and Bayne 2003; Mc Donnell et al. 2009). Arion intermedius is known to frequent hay fields (Bruijns et al. 1959), grasslands (Cameron 1978; Lutman 1978), gardens (Forsyth 2004), pasture (Quick 1960) and, to a lesser extent, disturbed agricultural cropping systems (e.g., Glen et al. 1984). In central and western Europe, where it is native (Barker 1999), it also inhabits forests (Bishop 1977; Tattersfield 1990) and feeds on fungi. Chichester and Getz (1969) reported this species from mixed and deciduous forests in the Great Smoky Moun- tains of Tennessee and North Carolina. According to Barker (1999) and Glen et al. (1984), the species may not be pestiferous, but Christian et al. (1999) cited it as a pest of wheat, and Barker (2002) highlighted it as a pest of pasture due to its effects on plant productivity. Arion hortensis is thought to be native to western and southern Europe (Roth and Sadeghian 2006), but it also is present as an exotic in other parts of the world. However, because it belongs to a species complex, pre- 1977 records of this slug need to be treated with caution as authors did not distinguish between A. hortensis s.s., A. distinctus, and A. owenii. Nevertheless, it has been collected throughout Europe, North America, and New Zealand in recent times (Barker 1999). In the U.S. it is known from California, Hawaii, Idaho, Maine, Michigan, Pennsylvania and Washington (Pearce and Blanchard 1992; Pearce and Bayne 2003; Mc Donnell et al. 2009). This species commonly is associated with disturbed habitats including urban areas such as gardens, roadsides, and forest fringes (Barker 2002). It is a known pest of a wide 121 range of vegetable crops (reviewed by South 1992; Barker 1999), horticultural crops (South 1992), pasture (Barker 2002), and sunflowers (Ballanger and Champolivier 1996). DISCUSSION AND CONCLUSIONS Given the dearth of publications on the invasive slug fauna of North America in the last 20 years, it is imperative that discoveries of new species are published (Mc Donnell et al. 2008). First, such reports are important as they provide a baseline for future studies involved in monitoring the subsequent surviv- al and spread of pest populations. Second, the current lack of information on the distribution of such invasive species means that USDA- APHIS lacks the necessary information to determine if an intercepted slug species represents a potential new threat to agricul- ture, horticulture, and the natural environ- ment (Reise et al. 2006). Our new records are of concern primarily because invasive slugs are major pests in agriculture, horticulture, and _floriculture (South 1992), causing considerable damage to important crops such as wheat (Martin and Kelly 1986; Kemp and Newell 1987), alfalfa (Grant et al. 1982; Barratt et al. 1989), corn (Mallet 1973; Barratt et al. 1989), soybean (Hammond 1985; Barratt et al. 1989), and tobacco (Mistic and Morrison 1979). These crops are the five most economically impor- tant for Kentucky, and the confirmation of two new invasive slug species in the state is of obvious concern to the $3.97 billion farm commodity industry (USDA — NASS Ken- tucky Field Office 2006). Furthermore, or- ganic growers rank slugs and birds as the second most important pests after weeds (Peacock and Norton 1990). With the drive towards agricultural diversification and low- input production, the planting of high value and highly vulnerable commodities such as strawberries could lead to yield loss due to known slug feeding habits within such crops (Prystupa et al. 1987; Duval and Banville 1989). While the most effective means of control is often molluscicides (Hammond et al. 1996), alternative approaches such as cultural and biological control need further examination to accurately evaluate their role in management programs in a_ variety of systems. 122 Finally, it is imperative that faunistic surveys are undertaken in other states and provinces throughout North America, partic- ularly those where the slug fauna is essentially undocumented e.g., North and South Dakota. This need also is reflected in the disjunct distribution of both A. intermedius and A. hortensis in the U.S. It is unlikely that both species are confined to the states mentioned above and more likely that a continuous distribution exists, which is likely to be confirmed with further malacological surveys. It is only by carrying out such surveys that the distribution of exotic gastropods within the U.S. will be clarified and the paucity in knowledge of these invasive pests improved. Such knowledge will be essential in mitigating the potentially severe damage to agricultural and horticultural systems caused by these pests, in addition to helping curb their spread to other states from potential hotspots. We also suggest a more widespread use of molecular methods for confirming the identity of invasive gastropods, particularly those in enigmatic genera such as Arion. The provision of such information may help to uncover additional species complexes in addition to providing potentially useful information for the elucidation of invasive pathways, both of which will be essential for successful control. ACKNOWLEDGEMENTS This research was funded, in part, by the European Union under a Marie Curie Out- going International Fellowship (MOIF-CT- 2005-21592) and the New Crop Opportunities Center at the University of Kentucky through a USDA Special Grant. JDH was supported by the University of Kentucky Agricultural Experiment Station State Project KY008043. We are extremely grateful to Michael J. Eskelson, Anna K. Thomas and Ray Fisher for assisting with the collection of arionid slugs from Kentucky, and to Charles W. Fox and William G. Wallin for photographic assistance. Thanks also to Barry Roth for providing relevant literature and to C. D. Williams for review comments. LITERATURE CITED Backeljau, T., and M. van Beeck. 1986. Epiphallus anatomy in the Arion hortensis species aggregate (Mollusca, Pulmonata). Zoologica Scripta 15:61-68. Journal of the Kentucky Academy of Science 69(2) Ballanger, Y., and L. Champolivier. 1996. Slug damage to sunflower crops in the south-west of France. Pages 321-326 in I. F. Henderson (ed). Slug and snail pests in agriculture. British Crop Protection Council, Farnham, UK. Barker, G. M. 1999. Fauna of New Zealand. Number 38. Naturalised terrestrial Stylommatophora. Manaaki Whenua Press, Canterbury, New Zealand. Barker, G. M. 2002. Gastropods as pests in New Zealand pastoral agriculture, with emphasis on Agriolimacidae, Arionidae and Milacidae. Pages 361-423 in G. M. Barker (ed). Molluscs as crop pests. CAB International Publishing, Wallingford, UK. Barratt, B. I. P., R. A. Byers, and D. L. Bierlein. 1989. Conservation tillage crop establishment in relation to density of the field slug (Deroceras reticulatum (Miiller). Pages 93-100 in I. F. Henderson (ed). Slugs and snails in world agriculture. British Crop Protection Council, Thornton Heath, UK. Bishop, M. J. 1977. The habitats of Mollusca in the Central Highlands of Scotland. Journal of Conchology 29:189-197. Branson, B. A., and D. L. Batch. 1969. Notes on exotic mollusks in Kentucky. The Nautilus 82:102-106. Bruijns, M. F. M., R. van Altena, and L. J. M. Butot. 1959. The Netherlands as an environment for land Mollusca. Basteria 23, supplement. Burch, J. B. 1962. How to know the eastern land snails. W. M. C. Brown Company Publishers, Dubuque, Iowa. Cameron, R. A. D. 1978. Terrestrial snail faunas of the Malham area. Journal of Molluscan Studies 4:715-728. Chichester, L. F., and L. L. Getz. 1969. The zoogeogra- phy and ecology of arionid and limacid slugs introduced into northeastern North America. Malacologia 7:313— 346. Christian, D. G., E. T. G. Bacon, D. Brockie, D. Glen, R. J. Gutteridge, and J. F. Jenkyn. 1999. Interactions of straw dispersal methods and direct drilling or cultiva- tions on winter wheat (Triticum aestivum) grown on a clay soil. Journal of Agricultural Engineering Research 73:297-309. Davies, S. M. 1977. The Arion hortensis complex, with notes on Arion intermedius Normand. Journal of Conchology 29:173—187. Davies, S. M. 1979. Segregates of the Arion hortensis complex (Pulmonata: Arionidae), with the description of a new species, Arion owensii. Journal of Conchology 30:123-127. Dourson, D., and K. Feeman. 2006. A survey of terrestrial Mollusca in selected areas of the Land Between the Lakes National Recreation Area. Journal of the Kentucky Academy Science 76:9-18. Dundee, D. S. 1974. Catalogue of introduced mollusks of eastern North America (North of Mexico). Sterkiana ba: L837, Duval, A., and G. Banville. 1989. Ecology of Deroceras reticulatum (Miill.) (Stylommatophora, Limacidae) in Quebec strawberry fields. Pages 147-160 in I. F. Two New Species of Arion in Kentucky—Mc Donnell et al. Henderson (ed). Slugs and snails in world agriculture. British Crop Protection Council, Thornton Heath, UK. Folmer, O., M. Black, W. R. Hoeh, R. Lutz, and R. C. Vrijenhoek. 1994. DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Molecular Marine Biology and Biotechnology 3:294—299. Forsyth, R. G. 2004. Land snails of British Columbia. University of British Columbia Press, Vancouver, Canada. Glen, D. M., C. W. Wiltshire, and N. F. Milsom. 1984. Slugs and straw disposal in winter wheat. Proceedings of the 1984 British Crop Protection Conference — Pests and Diseases 1:139-144. Grant, J. F., K. V. Yeargan, B. C. Pass, and J.Ge Parr. 1982. Invertebrate organisms associated with alfalfa seedling loss in complete-tillage and no-tillage plant- ings. Journal of Economic Entomology 75:822-826. Hammond, R. B. 1985. Slugs as a new pest of soybeans. Journal of the Kansas Entomological Society 58:364— 366. Hammond, R. B., J. A. Smith, and T. Beck. 1996. Timing of molluscicide applications for reliable control in no- tillage field crops. Journal of Economic Entomology 89:1028-1032. Kemp, N. J., and P. F. Newell. 1987. Slug damage to the flag leaves of winter wheat. Journal of Molluscan Studies 53:109-111. Kerney, M. P., and R. A. D. Cameron. 1979. A field guide to the land snails of Britain and north-west Europe. Harper Collins Publishers, London. Lutman, J. 1978. The role of slugs in an Agrostis-Festuca grassland. Pages 332-347 in O. W. Heal and D. F. Perkins (eds). Ecological Studies: Analysis and Synthe- sis, Vol. 27, Production ecology of British moors and montane grasslands. Springer-Verlag, New York. Mc Donnell, R. J., T. D. Paine, and M. J. Gormally. 2009. Slugs: A Guide to the Invasive and Native Fauna of California. University of California Agricultural and Natural Resources Publications. Mc Donnell, R. J., A. Hansen, T. D. Paine, and M. J. Gormally. 2008. A record of the invasive slug Veroni- cella cubensis (Pfeiffer, 1840) in California. The Veliger 50:81-82. Mallet, C. 1973. Les limaces, enemies des jardins mais aussi des grandes cultures. Phytoma No. 250, July- August 1973. Martin, A. W., and J. R. Kelly. 1986. The effect of changing agriculture on slugs as pests of cereals. Proceedings of the 1986 British Crop Protection Conference — Pests and Diseases 1:411-424. 123 Mistic, W. J., and D. W. Morrison. 1979. Control of slugs in burley tobacco fields in the Appalachian Mountains of North Carolina. International Tobacco 181:60-61. Peacock, L., and G. A. Norton. 1990. A critical analysis of organic vegetable crop protection in the U.K. Agricul- ture, Ecosystems and Environment 31:187-198. Pearce, T. A., and E. G. Bayne. 2003. Arion hortensis Férussac, 1819, species complex in Delaware and Pennsylvania, eastern U.S.A. (Gastropoda: Arionidae). The Veliger 46:362-363. Pearce, T. A., and D. Blanchard. 1992. Arion hortensis s.s., an introduced slug in Michigan. Walkerana 6:243- 244. Pfleger, V. 1999. A field guide in colour to molluscs. Blitz Editions, Leicester, UK. Pinceel, J., K. Jordaens, N. Van Houtte, A. J. De Winter, and T. Backeljau. 2004. Molecular and morphological data reveal cryptic taxonomic diversity in the terrestrial slug complex Arion subfuscus/fuscus. Biological Journal of the Linnean Society 83:23-38. Prystupa, B. D., N. J. Holliday, and G. R. B. Webster. 1987. Molluscicide efficacy against the marsh slug Deroceras laeve (Stylommatophora, Limacidae) on strawberries in Manitoba, Canada. Journal of Economic Entomology 80:936—943. Quick, H. E. 1960. British Slugs (Pulmonata: Testacelli- dae, Arionidae, Limacidae). Bulletin British Museum (Natural History) (Zoology) 6:103—226. Reise, H., J. M. C. Hutchinson, and D. G. Robinson. 2006. Two introduced pest slugs: Tandonia budapes- tensis new to the Americas, and Deroceras panormita- num new to the eastern U.S.A. The Veliger 48:110—115. Robinson, D. G. 1999. Alien invasions: the effects of the global economy on non-marine gastropod introductions into the United States. Malacologia 41:413-438. Robinson, D. G., and J. Slapcinsky. 2005. Recent introductions of alien land snails into North America. American Malacological Bulletin 20:89-93. Roth, B., and P. S. Sadeghian. 2006. Checklist of the land snails and slugs of California. Contributions in Science 3. Santa Barbara Museum of Natural History, Califor- nia. South, A. 1992. Terrestrial slugs. Biology, ecology and control. Chapman & Hall, London, UK. Tattersfield, P. 1990. Terrestrial mollusc faunas from some south Pennine woodlands. Journal of Conchology 33:355-—374. USDA — NASS Kentucky Field Office. 2006. Kentucky agricultural statistics and annual report, 2005-2006 Edition. United States Department of Agriculture National Agricultural Statistics Service and the Ken- tucky Department of Agriculture, Frankfort, Kentucky. J. Ky. Acad. Sci. 69(2):124-133. 2008. Inventory and Analysis of the Pteridophytes of Carter Caves State Resort Park, Carter County, Kentucky Channing Richardson and Allen C. Risk' Department of Biological and Environmental Sciences, Morehead State University, Morehead, Kentucky 40351 ABSTRACT Carter Caves State Resort Park (CCSRP) is comprised of 800 ha in Carter County, Kentucky. A field-based inventory of the pteridophytes of the park was conducted from September 2005—October 2007. Additionally, pteridophyte collections at regional herbaria were examined for specimens from the park. These efforts yielded a pteridophyte flora of 46 species and one hybrid for CCSRP; this is 61% of Kentucky’s known pteridophyte flora. Index of similarity values with other published floras ranged from 22.2% to 84.5%. CCSRP had over twice the predicted number (21) of pteridophyte species based on a species-area curve constructed with data from other regional floras. KEY WORDS: Carter Caves, Cliff Section, ferns, pteridophytes, taxonomy INTRODUCTION Carter Caves State Resort Park (CCSRP) lies within the Cliff Section in the northern portion of the Cumberland Plateau region of Kentucky and is located about 8.0 km north of Olive Hill in western Carter County (Fig- ure 1). The park centroid is located at latitude 38.3713°N and longitude 83.1175°W. The park is regionally and nationally known for its cave systems. Numerous state rare vascular plant species are known from CCSRP including Acer spicatum Lam., Paxistima canbyi A.Gray, Thaspium pinnatifidum (Buckley) A.Gray, Taxus canadensis Marsh., and Viola walteri House. Other botanical inventory work in the Cumberland Plateau has typically included entire vascular floras (e.g., Wofford et al. 1979; Sole et al. 1983; Thompson and Fleming 2004). Huffaker (1975) conducted a vascular plant inventory along Tygarts Creek upstream, adjacent, and downstream of CCSRP and reported 25 pteridophyte species. A 1939 Spring Foray of the Southern Appalachian Botanical Society that included many notable botanists (e.g., E. Lucy Braun and Earl Core) took place in CCSRP, but only two pteridophyte species were cited from the park (Gilbert 1939). Despite its reputation for botanical diversity, no botanical inventories are known to have been conducted in CCSRP. ' Corresponding author e-mail: a.risk@moreheadstate. edu Publications pertaining solely to pterido- phytes in Kentucky are Williamson (1878), McCoy (1938), and Cranfill (1980). William- son (1878) listed 40 species of pteridophytes from Kentucky with two species from Carter County. McCoy (1938) presented county distributions for 62 species, lesser taxa, and hybrids in Kentucky, including 32 from Carter County. Cranfill (1980) listed 70 pteridophyte taxa for Kentucky with 46 species and hybrids as occurring in Carter County. The same number of pteridophyte taxa was listed by Campbell and Medley (2006) for Carter County. The purpose of this study was to inventory Carter Caves State Resort Park (CCSRP) for pteridophytes and compare these results with those of other published floras. STUDY AREA CCSRP is located on the Wesleyville (Philley and Chaplin 1976), Tygarts Valley (Sheppard 1964), Grahn (Englund 1976), and Olive Hill (Englund and Windolph, Jr. 1975) 7.5’ quadrangles. The park is divided into two separate tracts of land with the major segment of the park being to the northwest of the smaller section (Figure 2). The larger section contains most of the buildings found within the park, including a lodge, welcome center, cabins, park residenc- es, and maintenance buildings. This section also includes a golf course and an 18.3 ha lake. Horn Hollow projects from the park to the northeast and is relatively isolated from the 124 Ferns of Carter Caves—Richardson and Risk 125 Figure 1. major infrastructure of the park. Two state nature preserves lie within CCSRP: Bat Cave State Nature Preserve comprises the north- western corner of the park and Cascade Caverns State Nature Preserve is located in the Cascade Caves portion of the park. Elevation overall ranges from 219.5- 341.4 m (720-1120 ft), with the maximum relief in any one area being 79.2-91.4 m (260-300 ft). The park contains a very weath- er resistant sandstone layer at its highest elevations, and below this lies a very weath- erable limestone layer. The sandstone layer is comprised of two formations. The upper layer is the Breathitt Formation, which supports dry Map showing location of Carter Caves State Resort Park (CCSRP) within Carter County, Kentucky. upland communities. The lower layer is the Lee Formation, which often outcrops as cliffs (Sheppard 1964; Englund and Windolph, Jr. 1975; Englund 1976; Philley and Chaplin 1976). Cliffs and their overhangs support a number of pteridophyte species, many of which are restricted to these habitats. The dominant woody species of these dry uplands are Quercus coccinea Miinchh., Q. alba L., Acer rubrum L., Carya spp., Nyssa sylvatica Marshall, Vaccinium pallidum Aiton, and Smilax rotundifolia L. Dominant species immediately below the bases of these cliffs are Betula lenta L., Tsuga canadensis (L.) Carr., and Rhododendron maximum L. 126 Bat Cave State Nature Preserve Cascade Caverns State Nature Preserve Journal of the Kentucky Academy of Science 69(2) Kilometers Figure 2. Topographic map of Carter Caves State Resort Park. The park is located on four 7.5’ topographic quadrangles: I-Wesleyville, II-Tygarts Valley, HI-Olive Hill, IV-Grahn. The Pennington and St. Genevieve mem- bers of the Newman Limestone lie beneath the Lee Sandstone (Sheppard 1964; Englund and Windolph, Jr. 1975; Englund 1976; Philley and Chaplin 1976). Dominant species in calcareous uplands are Quercus muhlen- bergii Engelm., Acer saccharum Marshall, and Fraxinus americana L. Species dominating the mid to lower slope communities are Acer saccharum, Quercus alba, Liriodendron tuli- pifera L., Fagus grandifolia Ehrh., and Acer rubrum. The lowermost geologic material exposed within the park beneath the St. Genevieve Limestone is the Cowbell Member of the Borden Formation. This rock is composed of Ferns of Carter Caves—Richardson and Risk shales (Sheppard 1964; Englund and Wind- olph, Jr. 1975; Englund 1976; Philley and Chaplin 1976). The Borden Formation has little exposure within the park but forms the bedrock of the lower reaches of Cave Branch and Smoky Creek and constitutes the bedrock of Tygarts Creek within CCSRP. MATERIALS AND METHODS Thirty nine collecting trips were made to CCSRP from September 2005 through Octo- ber 2007. A concentrated effort was made to thoroughly examine all representative habitats within the park. Sixty-five pteridophyte col- lections were made. All specimens collected were deposited in the Morehead State Uni- versity Herbarium (MDKY). Additionally, the entire pteridophyte collections in MDKY and the herbaria of Eastern Kentucky University (EKY), University of Cincinnati (CINC), Marshall University (MUHW), and the Uni- versity of Kentucky (KY) were examined for specimens from CCSRP. A literature search for pteridophyte collections within CCSRP was conducted which led to the examination of a Carter County, Kentucky, specimen of Adiantum capillus-veneris L. from the Gray Herbarium (GH). Specimens were tentatively identified in the field and then further examined using micro- scopes in the laboratory for definitive determi- nations. Cranfill (1980), Flora of North Amer- ica Editorial Committee (1993), and Jones (2005) were used for identification. Taxonomy and nomenclature follow the Flora of North America Editorial Committee (1993). The relative abundance of each species in CCSRP (Table 1) was estimated based on field observations and categorized as Common - characteristic and dominant in many habi- tats; Frequent - generally encountered; Infre- quent - scattered localities, populations usu- ally small; and Rare - one to three populations known. A species area Curve was constructed using other published floras from Kentucky and nearby regions for pteridophytes (Wade and Thompson 1991). SPSS 15.0 was used to logio transform pteridophyte species numbers and the study area sizes in hectares, and a power function model was used to create a line of best fit (Figure 3). 127 Sgrenson coefficients of similarity, based only on pteridophytes, were calculated with other published or available floras that in- cluded pteridophytes in the Appalachian Plateau as well as with selected floras from other regions of Kentucky, Tennessee, Ohio, North Carolina, Virginia, and West Virginia. Sgrenson’s coefficient was calculated by using the formula outlined in Sneath and Sokal (1963) as 2W/ (A+B), where W is the number of species the two floras have in common, A is the total number of species in the first flora, and B is the total number of species from the second flora. Only species and hybrids were used in these comparisons; infraspecific taxa were excluded from the calculations. RESULTS Forty-six pteridophyte species and one hybrid were documented for CCSRP. Forty- four of these were found in the field, two (Asplenium Xwherryi D. M. Smith et al., Botrychium biternatum (Savigny) Underw.) were in MDKY, and one (Adiantum capillus- veneris) was in GH. The largest families were Dryopteridaceae with twelve representatives and Aspleniaceae with seven representatives. The largest genera found were Asplenium with seven representatives and Cystopteris, Dryopteris, and Osmunda, each with three members (Table 1). Examination of collections in CINC, EKY, KY, MDKY, and MUHW yielded 74 addi- tional CCSRP specimens. Of the herbaria visited, only three (MDKY, KY, MUHW) contained collections from CCSRP. MDKY had 52 specimens representing 26 different species and one hybrid. KY housed 15 specimens totaling 13 species. In MUHW there were seven specimens accounting for four species from CCSRP. None of the species at the latter two herbaria represented taxa not documented by field work during the present study. McCoy (1938) listed an Adian- tum capillus-veneris specimen collected by a G. G. Jr. (full name unknown) in 1899 housed in GH from the Carter Caves region of Carter County. Examination of this specimen verified the identification. Some pteridophytes showed an exclusive relationship with either sandstone or lime- stone. Those species found either directly on, or in soils derived from, sandstone were 128 Journal of the Kentucky Academy of Science 69(2) Table 1. Annotated pteridophyte species list for Carter Caves State Resort Park. An asterisk (*) indicates new report for CCSRP. Specimens from different herbaria are separated by a semicolon. Overall park abundances are indicated, followed by habitat preferences. Aspleniaceae *Asplenium montanum Willd., Mountain Spleenwort — Richardson 108 (MDKY); infrequent; sandstone cliff faces. Asplenium pinnatifidum Nutt., Pinnatifid Spleenwort — Richardson 28, 49, Meade 139, Huffaker 978 (MDKY); Theiss s.n. (KY); infrequent; sandstone cliff faces. Asplenium platyneuron (L.) B.S.P., Ebony Spleenwort — Richardson 39, Meade 138, Huffaker 179, Broomall 48 (MDKY); common; widespread. *Asplenium resiliens Kunze, Black-Stemmed Spleenwort — Richardson 148 (MDKY); rare; limestone boulder. Asplenium rhizophyllum L., Walking Fern — Richardson 55, Huffaker 179, Ison 74, Meade 140, Mason 117, Huffaker 578 (MDKY); Theiss s.n. (KY); common; limestone cliffs and boulders. Asplenium ruta-muraria L., Wall-Rue — Richardson 147, Meade 133 (MDKY); Terrell 1206, Weller s.n. (KY); Gilbert s.n. (MUHW); infrequent; exposed dry limestone cliffs. *Asplenium trichomanes L., Maidenhair Spleenwort — Richardson 109 (MDKY); rare; sandstone cliff bases. Asplenium Xwherryi D. M. Smith et al., Wherry’s Spleenwort — Meade 152 (MDKY); assumed extirpated. Blechnaceae *Woodwardia areolata (L.) T. Moore., Netted Chain Fern — Richardson 342 (MDKY); rare; seep near sandstone cliff base. Dennstaedtiaceae *Dennstaedtia punctilobula (Michx.) T. Moore, Hay Scented Fern - Richardson 343, Risk 14086b (MDKY); rare; sandstone cliff bases. *Pteridium aquilinum (L.) Kuhn var. latiusculum (Desv.) Underw. ex A. Heller, Bracken Fern — Richardson 196.2 (MDKY); rare; soil below sandstone cliff. Dryopteridaceae Athyrium filix-femina (L.) Roth, Southern Lady Fern — Richardson 24, 29, 35, Huffaker 979a (MDKY); frequent; sandy soils. Cystopteris bulbifera (L.) Bernhardi, Bulblet Bladder Fern — Richardson 45, 54, 59, Meade 147, Levy 73 (MDKY); frequent; limestone cliff bases and limestone cobble dominated areas. *Cystopteris protrusa (Weath.) Blasdell, Southern Bladder Fern — Richardson 60 (MDKY); infrequent; limestone cliffs and calcareous soils. Cystopteris tennesseensis Shaver, Tennessee Bladder Fern — Risk 14027, Lykens 290 (MDKY); rare; moist limestone cliffs and boulders. Deparia acrostichoides (Sw.) M.Kato, Silvery Glade Fern — Richardson 44, 47, Meade 151 (MDKY); infrequent; ravines and shaded slopes. Diplazium pycnocarpon (Spreng.) M.Broun, Glade Fern — Richardson 46, Ison 73, Huffaker 495b, Broomall 49 (MDKY); frequent; moist calcareous areas. Dryopteris goldiana (Hook.) A.Gray, Goldie’s Wood Fern — Richardson 80, Meade 148 (MDKY); Weller s.n. (KY); rare; sandstone colluvium. Dryopteris intermedia (Muhl. ex Willd.) A.Gray, Evergreen Wood Fern — Richardson 52, 62, 63, Huffaker 176, Meade 135 (MDKY); Theiss s.n. (KY); common; sandstone cliffs and moist ravines. Dryopteris marginalis (L.) A.Gray, Marginal Wood Fern — Richardson 30, 38, Huffaker 175, Ison 72, Broomall 82 (MDKY); Smith et al. 3449, 3450, Gilbert 871 (MUHW); common; moist ravines and slopes below sandstone cliffs. Onoclea sensibilis L., Sensitive Fern — Risk 14097, Huffaker 168, Weller s.n. (MDKY); frequent; moist field edges. Polystichum acrostichoides (Michx.) Schott., Christmas Fern — Richardson 22, 26, 32, 33, Huffaker 600, King s.n., Meade 150 (MDKY); common; widespread. *Woodsia obtusa (Spreng.) Torr., Blunt-lobed Cliff Fern — Richardson 214.2 (MDKY); Gilbert 967 (MUHW); infrequent; limestone-dominated areas. Equisetaceae Equisetum arvense L., Field Horsetail — Richardson 51 (MDKY); Theiss s.n. (KY); infrequent; riparian zones. Equisetum hyemale L., Scouring Rush — Richardson 57 (MDKY); Meijer s.n., Theiss s.n. (KY); Trail 34, Watson 39 (MUHW); infrequent; riparian zones. Hymenophyllaceae *Trichomanes boschianum Sturm, Appalachian Filmy Fern — Richardson 110 (MDKY); rare; moist backwalls of sandstone rockhouses. Isoetaceae *Isoetes engelmannii A. Braun., Engelmann’s Quillwort — Risk 14116a (MDKY); rare; streamhead seeps. Lycopodiaceae Diphasiastrum digitatum (Dillenius ex A. Braun) Holub, Running Ground Cedar — Richardson 53, Huffaker 171, Ison 76 (MDKY); Theiss s.n., Meijer s.n. (KY); infrequent; acidic soils. *Huperzia lucidula (Michx.) Trevis., Shining Firmoss — Richardson 61 (MDKY); infrequent; sandstone cliffs and acidic soils. Huperzia porophila (F.E.Lloyd & Underw.) Holub., Rock Clubmoss — Richardson 25, Meade 134 (MDKY)); infrequent; sandstone cliffs. Ferns of Carter Caves—Richardson and Risk 129 Table 1. Continued. Lygodiaceae *Tygodium palmatum (Bernh.) Sw., Appalachian Climbing Fern — Richardson 115 (MDKY); rare; sandstone cliff bases. Ophioglossaceae Botrychium biternatum (Savingy) Underw., Sparse-lobed Grapefern — Meade 313 (MDKY); historical, unencountered during present study. Botrychium dissectum Spreng., Dissected Grapefern — Richardson 51, Meade 142 (MDKyY); frequent; acidic soils below sandstone cliffs. Botrychium virginianum (L.) Sw., Rattlesnake Fern — Richardson 31.1, 36, Carr 52 (MDKY); Warden s.n. (KY); frequent; middle to lower slopes. *Ophioglossum vulgatum L., Adder’s Tongue Fern — Risk 13954, 14098 (MDKY); rare; riparian zone and moist slope. Osmundaceae Osmunda cinnamomea L., Cinnamon Fern — Richardson 23, Meade 130 (MDKY); frequent; moist sandstone cliffs and acidic soils. Osmunda claytoniana L., Interrupted Fern — Richardson 40, 43, King 3701 (MDKY); frequent; moist sandstone cliffs and acidic soils. Osmunda regalis L., Royal Fern — Richardson 48, Meade 173 (MDKY); rare; streamhead seeps and wet sandstone cliff faces. Polypodiaceae Polypodium appalachianum Haufler & Windham, Appalachian Polypody — Richardson 107 (MDKY); Theiss s.n. (KY); rare; sandstone cliff faces and boulders. Polypodium virginianum L., Rock Cap Fern — Richardson 31.2, 41, 56, 64, Broomall 47, Levy 72 (MDKY); frequent; sandstone cliffs and boulders. Pteridaceae Adiantum capillus-veneris L., Southern Maidenhair Fern - G.G. Jr. s.n. (GH); historical, presumed extirpated. Adiantum pedatum L., Northern Maidenhair Fern — Richardson 34, King s.n., Huffaker 186, 345, 580, Brown and Brown 10500 (MDKY); common; moist slopes and lowlands. Pellaea atropurpurea (L.) Link, Purple Cliff Brake — Richardson 67, Meade 139 (MDKY); Theiss s.n. (KY); frequent; limestone cliff faces. Selaginellaceae *Selaginella apoda (L.) Spring., Meadow Spikemoss — Richardson 75 (MDKY); infrequent; moist fields and lawns. Thelypteridaceae Phegopteris hexagonoptera (Michx.) Fée, Broad Beech Fern — Richardson 37, 65, King 3710, Huffaker 595, Browne and Browne 10504 (MDKY); MclInteer 2518, 1198 (KY); common; moist soil. Thelypteris noveboracensis (L.) Nieuwl., New York Fern — Richardson 42, Huffaker 593 (MDKY); common; moist soil. Vittariaceae *Vittaria appalachiana Farrar and Mickel, Appalachian Gametophyte Fern — Risk 14575, 14952 (MDKY); rare; ceilings of sandstone rockhouses. Asplenium montanum Willd., A. pinnatifidum DISCUSSION Nutt., A. trichomanes L., Dennstaedtia punc- tilobula (Michx.) Moore, Dryopteris goldiana (Hook.) A. Gray, D. intermedia (Muhl. ex Willd.) A. Gray, Huperzia porophila (F. E. Lloyd and L. Underw.) Holub., Trichomanes boschianum Sturm, and Vittaria appalachiana Farrar and Mickel. Those pteridophytes with an exclusive relationship to limestone or its derived soils were Asplenium resiliens Kunze, A. ruta-muraria L., all three species of Cys- topteris, Diplazium pycnocarpon (Spreng.) M. Broun, and Pellaea atropurpurea (L.) Link. The formula for the species area curve was Forty-six species and one hybrid represent- ing 61% of Kentucky’s total pteridophyte flora were documented from CCSRP. A large portion of the taxa known from Kentucky (Jones 2005) not documented from the park characteristically occur in wetlands (e.g., Azolla caroliniana Willd. and Thelypteris palustris Schott.) or on dry sandstone uplands (e.g., Chelianthes lanosa (Michx.) D. C. Eaton, Diphasiastrum tristachyum (Pursh) Holub, and Lycopodium obscurum L.). These two habitats are poorly represented within S = 0.972A°* (where S = logig number of species and A = logo area in ha; R® = 0.718, P < 0.001) The predicted value of 21 for CCSRP based on this curve was less than half of the actual number, 47, of species and hybrids documented for the park (Figure 3). CCSRP. Species known from Carter or nearby counties and diligently searched for but not found in CCSRP were Asplenium bradleyi D. C. Eaton, Cheilanthes lanosa, Diphasiastrum tristachyum, Dryopteris car- thusiana (Vill.) H. P. Fuchs, Lycopodium 130 Journal of the Kentucky Academy of Science 69(2) 1.404 log species 1.007 log area (ha) Figure 3. Species area curve for pteridophytes based on published floras with the data point for Carter Caves State Resort Park indicated. Studies included were Basinger et al. (1997); Blackwell et al. (1981); Bush (1988); Campbell and Meijer (1989); Carpenter and Chester (1987); Clements and Wofford (1991); Cranfill (1991); Davies (1955); Feeman (2002); Gaddy (1990); Grubbs and Fuller (1991); Hannan and Lassetter (1982); Ludwig (1999); Murrell and Wofford (1987); Palmer (1990); Schmalzer et al. (1985); Sole et al. (1983); Suiter and Evans (1999); Weckman et al. (2003); Wofford et al. (1979); Woods and Fuller (1988). obscurum, and Pleopeltis polypodioides (L.) and Medley (2006), Cystopteris protrusa (Weath.) E. G. Andrews and Windham. Blasdell, Isoetes engelmannii A. Braun, Ophio- Based upon comparison with county record glossum vulgatum L., and Vittaria appalachiana maps presented in Cranfill (1980) and Campbell —_ are new county records for Carter County. Table 2. Floras with location, size, and index of similarity (IS) to Carter Caves State Resort Park. Index of similarity formula was IS= 2W/(A+B). Study Bush (1988) Sole et al. (1983) Weckman et al.(2003) Wofford et al. (1979) Clements and Wofford (1991) Suiter and Evans (1999) Schmalzer et al. (1985) Gaddy (1990) Murrell and Wofford (1987) Basinger et al. (1997) Grubbs and Fuller (1991) Woods and Fuller (1988) Hannan and Lassetter (1982) Blackwell et al. (1981) Campbell and Meijer (1989) Carpenter and Chester (1987) Feeman (2002) Davies (1955) Cranfill (1991) Palmer (1990) Ludwig (1999) Physiographic Province Size (ha) IS Appalachian Plateau 8 0.222 Appalachian Plateau 220 0.407 Appalachian Plateau 262 0.629 Appalachian Plateau 4000 0.716 Appalachian Plateau 1000 0.727 Appalachian Plateau 95,132 0.756 Appalachian Plateau 4000 0.790 Blue Ridge 10 0.444 Blue Ridge 2843 0.657 Coastal Plain 3702 0.413 Coastal Plain 62,464 0.500 Coastal Plain 99 460 0.639 Interior Low Plateau 202 0.281 Interior Low Plateau 10.3 0.296 Interior Low Plateau 400 0.469 Interior Low Plateau 325 0.545 Interior Low Plateau 46,850 0.703 Interior Low Plateau 20,000 0.776 Interior Low Plateau 163,100 0.845 Piedmont 3500 0.615 Valley and Ridge 8.7 0.291 Ferns of Carter Caves—Richardson and Risk Rare pteridophytes within the region col- lected during this study were Appalachian filmy fern (Trichomanes boschianum), Engel- mann’s quillwort (Isoetes engelmannii), and Goldie’s wood fern (Dryopteris goldiana). Within the park there are numerous sandstone rockhouses. These create excellent habitat for the rare Appalachian filmy fern (Trichomanes boschianum). Four rockhouses within the Horn Hollow region of the park contained populations of the Appalachian filmy fern with the largest population covering approximately 2.79 m’? of substrate. Two populations of Engelmann’s quillwort (Isoetes engelmannii) were found in the Cascade Caves region of the park. These two populations were in slightly swampy streamheads that contained abundant cinnamon fern (Osmunda cinnamomec L.), royal fern (Osmunda regalis L.), and New York fern (Thelypteris noveboracensis (L.) Nieuwl). It is hypothesized that Engelmann’s quillwort may be more common in eastern Kentucky than current collections indicate due to its habitat, atypical pteridophyte morphology, and relatively small size. Goldie’s wood fern (Dryopteris goldiana) was found along a heavily used trail just uphill from the entrance to Saltpetre Cave and along the sandstone cliff line southeast of the Saltpetre Cave site. Pteridophytes that were rare within CCSRP were Asplenium resiliens (one site, but with well developed large plants), Asplenium tri- chomanes (several individuals in one location), Dennstaedtia punctilobula (two sites), Dryop- teris goldiana (three sites along one cliff), Isoetes engelmannii (two sites), Lygodium palmatum (Bernh.) Sw. (two small plants at one site), Ophioglossum vulgatum (two sites), Osmunda regalis (three sites), Pteridium aquilinum (L.) Kuhn (one site), Trichomanes boschianum (four sites along one cliff line), Vittaria appalachiana (two sites), and Wood- wardia areolata (L.) T. Moore (one site). Habitat destruction could be related to the rarity of Dryopteris goldiana and Pteridium aquilinum in CCSRP. The Goldie’s Wood Fern population is bisected by a heavily used trail and the only Bracken Fern site was beside another heavily used trail within 100 m of the lodge. Habitat specificity may account for the rarity of some species, e.g., Isoetes engelmannii, Ophioglossum vulgatum, Os- munda_ regalis, Trichomanes boschianum, 131 Vittaria appalachiana, and Woodwardia are- olata. These ferns, with the exception of T. boschianum and V. appalachiana that inhabit deep sandstone overhangs, require very moist to almost swamp-like conditions, a relatively uncommon habitat within CCSRP. Adiantum capillus-veneris was collected from the Carter Caves region in 1899. This specimen represents an important collection because this is very near the northern limit of its geographical range. However, it is not certain whether the species occurred within the current park boundaries because at the time of the collection the park was not publicly owned and did not have the same boundaries as it does today. Also, the location data on the label is typical for collections of this time period and was not very precise. Because the locality data on the label is vague, the exact site for collection of this specimen is not known. The researchers diligently looked in all appropriate habitats for an extant population of this fern, but in the sites examined none were found. One potential location of suitable habitat was not examined because the site was an inaccessible ledge in the midslope of a large cliff. Another possibility is that the A. capillus-veneris may have been in a location since submerged by the reservoir on Smokey Creek. For the purposes of this study, it is assumed that this species did occur, at least historically, within the present boundaries of the park. This species is presumed extirpated from within the park. A specimen of Asplenium Xwherryi D. M. Smith et al. collected at CCSRP is housed at MDKY. This hybrid is presumed to be extirpated from CCSRP. Examination of the locality data on the label, in conjunction with interviewing a retired park employee, allowed the researchers to relatively confidently pin- point the former population’s location. This population was in a relatively high traffic area, and general human disturbances could have been the cause of its extirpation. Also, excessive collecting may have extirpated this hybrid, since there are several duplicates of this hybrid housed at MDKY. The index of similarity (IS) values between the pteridophyte floras of CCSRP and other areas (Table 2) were fairly high. This was expected because most pteridophyte species 132 have large geographical ranges. The highest index of similarity value was between CCSRP and Hardin County, Kentucky (Cranfill 1991), at 0.845. This is notable because this study occurred in a different physiographic region of the state, the Interior Low Plateau, than that within which CCSRP is found. Hardin County does, however, contain both sand- stone and limestone substrate. The area with the lowest IS with CCSRP was a wet meadow in Barbour County, West Virginia (Bush 1988), at 0.222. This study occurred in the same physiographic province as the current study. It was, however, a very small (8 ha) and specialized habitat that likely accounts for this large disparity. CCSRP lies well above the expected value for an area of its size, being further above the line than any other flora in- cluded for comparison (Figure 3). Some reasons for this high species richness are the presence of many sandstone and limestone outcrops within the park, the relatively low amount of disturbance that has occurred to the forests, and the geographic location of CCSRP. ACKNOWLEDGMENTS We are indebted for the assistance and consultation of John Tierney during field work. We thank the Kentucky State Park System and the Kentucky State Nature Preserves Commission for granting collecting permits. Funding was provided by the Ken- tucky Academy of Science Marcia Athey and Botany Funds. The Undergraduate Research Fellowship Program of Morehead State Uni- versity provided invaluable support for the first author over the course of this project. LITERATURE CITED Basinger, M. A., J. S. Huston, R. J. Gates, and P. A. Robertson. 1997. Vascular flora of Horseshoe Lake Conservation Area, Alexander County, Illinois. Casta- nea 62:82-99. Blackwell, W. H., D. M. Brandenburg, M. D. Baeschle, and P. D. Doran. 1981. Checklist of vascular plants of the “Highbanks,” an oak hickory stand in southwestern Ohio. Castanea 46:300-310. Bush, E. M. 1988. A floristic study of a wet meadow in Barbour County, West Virginia. Castanea 53:132-139. Campbell, J. N., and M. Medley. 2006. Illustrated atlas of vascular plants in Kentucky: a first approximation. July Journal of the Kentucky Academy of Science 69(2) 2006 Draft. Unpublished bound manuscript distributed by the authors, Lexington, KY. Campbell, J. N., and W. Meijer. 1989. The flora and vegetation of Jessamine Gorge, Jessamine County, Kentucky: a remarkable concentration of rare species in the Bluegrass Region. Transactions of the Kentucky Academy of Science 50:27-45. Carpenter, J. S., and E. W. Chester. 1987. Vascular flora of the Bear Creek Natural Area, Stewart County, Tennessee. Castanea 52:112-128. Clements, R. K., and B. E. Wofford. 1991. The vascular flora of Wolf Cove, Franklin County, Tennessee. Castanea 56:268—286. Cranfill, R. 1980. Ferns and fern allies of Kentucky. Kentucky State Nature Preserves Commission, Frank- fort, KY. Cranfill, R. 1991. Flora of Hardin County, Kentucky. Castanea 56:228—267. Davies, P. A. 1955. A preliminary list of the vascular plants of Mammoth Cave National Park. Castanea 20:107-127. Englund, K. J. 1976. Geologic Map of the Grahn Quadrangle, Carter County, Kentucky. U.S. Geological Survey (map scale 1:24,000). Reston, VA. Englund, K. J., and J. F. Windolph, Jr. 1975. Geologic Map of the Olive Hill Quadrangle, northeastern Kentucky. U.S. Geological Survey (map scale 1:24,000). Reston, VA. Feeman, K. 2002. The vascular flora of the Tablelands: a natural region in the northeastern section of the Knobs of Kentucky. M.S. Thesis. Morehead State University, Morehead, KY. Flora of North America Editorial Committee (eds). 1993. Flora of North America North of Mexico. Volume 2. New York and Oxford. Gaddy, L. L. 1990. Glade Fern Ravine, a rich fern site in the Blue Ridge Province of South Carolina. Castanea 59:282-285. Gilbert, F. A. 1939. Spring foray in Kentucky. Castanea 4:132-134. Grubbs, J. T., and M. J. Fuller. 1991. Vascular flora of Hickman County, Kentucky. Castanea 56:193-214. Hannan, R. R., and J. S. Lassetter. 1982. The vascular flora of the Brodhead Swamp Forest, Rockcastle County, Kentucky. Transactions of the Kentucky Academy of Science 43:43—49. Huffaker, W. M. 1975. A preliminary survey of the vascular flora of upper Tygarts Creek, Carter County, Kentucky. M.S. Thesis, Morehead State University, Morehead, KY. Jones, R. L. 2005. Plant life of Kentucky: an illustrated guide to the vascular flora. University Press of Kentucky, Lexington, KY. Ludwig, J. C. 1999. The flora of dolomite and lime- stone barrens in southwestern Virginia. Castanea 64: 209-230. McCoy, T. N. 1938. The ferns and fern allies of Kentucky. American Fern Journal 28:41-46. Murrell, Z. E., and B. E. Wofford. 1987. Floristics and phytogeography of Big Frog Mountain, Polk County, Tennessee. Castanea 52:262-290. Ferns of Carter Caves—Richardson and Risk Palmer, M. W. 1990. Vascular flora of the Duke Forest, North Carolina. Castanea 55:229-244. Philley, J. C., and J. R. Chaplin. 1976. Geologic map of the Wesleyville Quadrangle, northeastern Kentucky. U.S. Geological Survey (map scale 1:24,000). Reston, VA. Schmalzer, P. A., T. S. Patrick, and H. R. DeSelm. 1985. Vascular flora of the Obed Wild and Scenic River, Tennessee. Castanea 50:71-88. Sheppard, R. A. 1964. Geology of the Tygarts Valley Quadrangle, Kentucky. U.S. Geological Survey (map scale 1:24,000). Washington DC. Sneath, P. H., and R. R. Sokal. 1963. Principles of numerical taxonomy. Freeman Publishing Co., San Francisco, CA. Sole, J. D., S. Lassetter, and W. H. Martin. 1983. The vascular flora of Lilley Cornett Woods, Letcher County, Kentucky. Castanea 48:174-188. SPSS (Statistical Package for the Social Sciences). version 15.0. 2006. Suiter, D. W., and D. K. Evans. 1999. Vascular flora and rare species of New River Gorge National River, West Virginia. Castanea 64:23—-49. 133 Thompson, R. L., and C. A. Fleming. 2004. Vascular flora and plant communities of the John B. Stephenson Memorial Forest State Nature Preserve (Anglin Falls Ravine), Rockcastle County, Kentucky. Castanea 69:125-138. Wade, G. L., and R. L. Thompson. 1991. The species-area curve and regional floras. Journal of the Kentucky Academy of Science 52:21-26. Weckman, T. J., J. E. Weckman, and N. R. George. 2003. Checklist of the vascular flora of Pilot Knob State Nature Preserve, Powell County, Kentucky. Journal of the Kentucky Academy of Science 64:36- 54. Williamson, J. 1878. Ferns of Kentucky. John P. Morton and Co, Louisville, KY. Wofford, B. E., T. S. Patrick, L. R. Phillippe, and D. H. Webb. 1979. The vascular flora of Savage Gulf, Tennessee. Sida 8:135-151. Woods, M., and M. J. Fuller. 1988. The vascular flora of Calloway County, Kentucky. Castanea 53:89- 101. J. Ky. Acad. Sci. 69(2):134—140. 2008. Proterometra macrostoma (Faust) (Trematoda: Azygiidae): Further Studies on Strains at North Elkhorn Creek, Scott County, Kentucky Ronald Rosen,' Dikshya Bastakoty, Tsering Dolma, Aaron Fidler, Miluka Gunaratna, Robert Twiggs, Brea Viragh, Jonathan Fleming, Bojana Jovanovic, Aishe Sarshad, Emilie Throop, Fady Zaki, and Andy Ammons Department of Biology, Berea College, Berea, Kentucky 40404 ABSTRACT The objectives of the present study were (1) to compare the frequency of Proterometra macrostoma (Faust) strains and the number of single vs. multiple strain infections in snails, Elimia semicarinata, collected from North Elkhorn Creek during June 2000, 2004, and 2007, (2) to measure selected cercarial strain egg loads and swimming distance under red and white light, and (3) to determine the developmental rate (based on egg stages) of strain I and III adult worms in the bluegill, Lepomis macrochirus. The frequencies of strains I (11.1-28.6%) and III (21.3-35.7%) were the highest, while strains II (4.2-8.8%) and VI (0-1.4%) have been relatively uncommon at North Elkhorn Creek over the last seven years. The majority (76.2-94.3%) of infected snails had multiple strain infections. No significant difference was found between the average egg number (strains I, I, HI, IV, V, VII, and VIII) or mean swimming distance (strains I, HI, and VIII) of cercariae under red or white light. Similarly, no significant difference was found between the average number of stage I, II, and III eggs, respectively, found in strain I and III adult worms on days 12 and 20 postinfection in experimentally infected bluegill. KEY WORDS: Proterometra macrostoma, cercariae, strains, Kentucky, North Elkhorn Creek INTRODUCTION Proterometra macrostoma (Faust) is a digenetic trematode found in streams and rivers east of the Mississippi River. The life cycle incorporates a snail intermediate host and a centrarchid fish definitive host. The latter becomes infected upon ingestion of the parasite’s unusually large (3.0-9.0 mm) pro- genetic cercaria whose swimming is marked by vertical bursts (36.0-150.0 mm) in the water column. The adult worm, once ingested, exits from its cercarial tail and attaches to the esophagus and stomach of the fish host. Horsfall (1934), in her study of Cercaria macrostoma (=P. macrostoma) from. snails collected at Homer Park, Illinois, the Des Plaines River, Illinois, and the Oconomowoc River, Wisconsin, noted that the number, size, and position of papillae or mammalations varied considerably in fresh specimens of the worm. Dickerman (1945) reported three distinct cercarial es of P. macrostoma in snails obtained from the Bass Island region of western Lake Erie based on color, the distribution of mammalations with and with- out spines, morphological measurements, and ' Corresponding author e-mail: ron_rosen@berea.edu the presence/absence of eggs. Initially he thought that these “types” represented three separate species, but experimental infections of fish resulted in the recovery of just a single e of adult worm. Dickerman (1945) con- cluded that his cercarial types were in fact varieties that had not evolved to the point of being considered as separate species. Riley and Uglem (1995) attempted to sort out whether Dickerman’s (1945) cercarial varia- tions, “represent phenotypic plasticity, strains or distinct species.” Working primarily at North Elkhorn Creek, Kentucky, Riley and Uglem (1995) separated this species into eight strains based on the following cercarial features: (1) distribution of papillae with and without spines, (2) daily and seasonal emer- gence patterns, (3) initiation and distance/ duration of swimming under red light, and (4) degree of infectivity in different species of centrarchid fish. They also determined the relative frequency of these P. macrostoma strains at North Elkhorn Creek during May 1990 and 1991. The continued presence and frequency of these eight strains in the Elimia semicarinata (Say) snail population at North Elkhorn Creek since Riley and Uglem’s (1995) original study in 1990 and 1991 has not been evaluated. 134 Proterometra macrostoma Strains—Rosen et al. Because possible changes in these strains may be reflective of changes in the preferred centrarchid host community at the site (Riley and Uglem 1995), the purpose of the present study was to summarize the frequency of these strains and the number of single vs. multiple strain infections in snails collected from North Elkhorn Creek during June 2000, 2004, and 2007. In addition, an attempt was made to further characterize selected strains of this parasite by comparing immature egg loads in cercariae, cercarial swimming dis- tance under red and white light, and devel- opmental rate in the fish definitive host based on stage of egg maturity. MATERIALS AND METHODS For all experiments, E. semicarinata were collected from North Elkhorn Creek (lat 38°11'00’N, long 84°29'19”"W), Scott County, Kentucky, during June 2000, 2004, and 2007 and screened for mature cercarial infections with P. macrostoma. The latter was accom- plished by placing snails in an environmental chamber at 20°C and a 12/12 hr light/dark cycle for three consecutive days and monitor- ing the emergence of cercariae. Infected snails were then individually placed in isolated chambers within compartmentalized boxes. Individual snails were assigned numbers to ensure proper tracking during the 21-day experiment, and maintained at 20°C in a 12/12 hr light/dark cycle. Cercarial emergence from each snail was monitored daily, each freshly emerged cercaria identified to strain according to Riley and Uglem (1995), and the number of single and multiple strain infec- tions recorded at the termination of the experiment. Egg counts were also made from some of these cercariae. ANOVA was used to assess possible differences between the mean egg loads of cercarial strains with adequate sample sizes. Previous work had shown significant differ- ences in P. macrostoma cercariae swimming distances under red vs. white light (Rosen et al. 2005a). To assess possible differences in cercarial swimming according to strain in these two light regimens, 2-L graduated cylinders were filled with artificial pond water (APW; 0.5 mM NaCl, 0.05 mM KCI, 0.4 mM CaCls and 0.025 mM MgCl). They were then placed under broad spectrum 20-watt fluo- 135 rescent lights wrapped in black fiberglass window screening or red filters (Edmund Scientific) and adjusted to achieve an intensity of 11 foot candles (fe; Extech Light Meter) at the top of each cylinder. This intensity approximated daylight field values (14.5 fe) at North Elkhorn Creek at 0.6 m depth. One cercaria (less than 2 hr post-emergence) was then placed into a cylinder and exposed in 7- min intervals to red and then white light. Each 7-min time span consisted of a 2-min accli- mation period followed by a 5-minute record- ing period. A single swimming burst distance was recorded at the beginning of each minute of the 5-min recording interval for each light treatment, and the average distance (using a conversion of 1.0 ml = 0.2 mm) was calculat- ed from these five observations. At the termination of each swimming experiment, the cercaria was recovered and _ its strain determined using papillae and spine distribu- tion. ANOVA was used to assess possible differences between the mean swimming distances of strains I, HI, and VUHI under both red and white light. Due to the progenetic nature of the P. macrostoma cercaria, it is only possible to measure developmental rate of this worm in the definitive host by the maturity of its egg stages (Rosen et al. 2000, 2005b). To evaluate possible differences in the developmental rate of strain I and II adults in the bluegill, Lepomis macrochirus Rafinesque young (40— 80 mm) fish were obtained from Pfeiffer Fish Hatchery in Frankfort, Kentucky. They were placed in 37.85-L tanks at 22.7 + 4.6°C and maintained on fish pellets acquired from the hatchery. Individual fish were removed from their tanks and placed in smaller, 3.79-L tanks for infection with 1-3 cercariae previously identified to strain I or III with the aid of a compound microscope. Once all introduced cercariae were ingested (generally less than 1- hr), the fish were returned to their respective holding tanks. Fish were anesthetized in MS- 222 (tricaine methane sulfonate) and necrop- sied on days 12 and 20 PI (postinfection). The eggs in the recovered worms were counted and categorized as either stage I (clear cell at opercular end and dark vitelline mass at opposite end), stage II (vitelline mass restrict- ed to periphery; advanced cleavage), or stage II (dark yellow shell; fully-formed miracidi- 136 Journal of the Kentucky Academy of Science 69(2) 40 #2000 ©2004 ©2007 (a) Q rea Fuauencyie oo I lt Hi iV Strains Figure 1. Relative frequency of eight Proterometra macrostoma cercarial strains at North Elkhorn oo Ss co Vv Vi Vit Vill (N=140 cercariae), 2004 (N=364 cercariae), and 2007 (N=310 cercariae). um) according to Rosen et al. (2000, 2005b). A Student’s t-test was used to determine possi- ble differences in the mean number of each of the egg stages present on days 12 and 20 PI, respectively, between these strains. A proba- bility of P = 0.05 was considered as significant for all statistical tests. RESULTS Strains I and III were the most prevalent, while strains Il and VI were relatively uncommon during June 2000, 2004, and 2007 (Figure 1). Only a small fraction of snails had single strain infections; most had multiple strain infections (2-5), with two strains being the most common (Figure 2). In snails possessing single or multiple infec- tions, there was no obvious pattern regarding the release of the different cercarial strains over the 21-day observation period (Figures 3a—3d)) The number of eggs within the various cercarial strains varied considerably (i.e., range 0-40), but no significant difference (F = 1.869: df = 6, 436; P = 0.085) was found in the average egg numbers between cercarial strains (Figure 4). Similarly, no significant difference was found between the mean swimming distances of strain I, HI, and VIII cercariae under red (F = 2.183: df = 2, 83; P = 0.961 ) or white (F = 2.621: df = 2, 83; P = 0.079) light (Figure 5). No significant difference was found be- tween the average number of stage I (t = 1.287; df = 29; P= 0.208)-or stage II (t = 0.369; df = 29; P = 0.714) eggs in strain I and III worms on day 12 PI (Figure 6). Similarly, no significant difference was found between the average number of stage I (t = 1.875; df = 16; P =0.079), stage II (t = 0.943: df = 16; P = 0.360), or stage III (t = 0.599; df = 16; P = 0.558) eggs in strain I and III worms on day 20 PI (Figure 6). DISCUSSION According to Bryant and Flockhart (1986), there is little chance of population divergence/ speciation when gene exchange is high; conversely, when gene exchange is low, there is a good chance for speciation. Somewhere between these extremes the populations become sufficiently different to be considered as variants or strains (Bryant and Flockhart 1986). Such “moderate” gene flow in the P. macrostoma population at North Elkhorn Creek seems to have been relatively un- changed over the last 17 yr as reflected by the strain types present and their compara- tively consistent frequencies. In the present study, the highest frequencies were observed Proterometra macrostoma Strains—Rosen et al. 60 S S & o - ; 137 @2000 [2004 82007 Figure 2. Relative frequency af single \ VS. ee Proteromibina macrestonie strain infections in EUaa semicarinata at North Elkhorn Creek in June 2000 (N=21 snails), 2004 (N=35 snails), and 2007 (N=22 snails). for strains I and III, while strains II and VI have been relatively uncommon between 2000-2007. Similarly, Riley and Uglem (1995) also found that strains I and III were a Strain Vill — wm al ey 7 res ea bd a 123 4 5 6 7 & & 10 14 12 13 14 45 16 17 18 19 20 21 pay St Experian. ali, OStrainV¥ MWStrain iv Bstrain ih Number Of Cercariag ot id od lh on ~ wn ty) nm bal or 123 4 5 & 7 & & 10 11 12 13 14 15 16 17 18 19 20 21 __ Day Of Experiment ; Number Of Cercariae the dominant forms of P. macrostoma at their two Kentucky sites (North Elkhorn Creek and Cane Run Creek) in 1990/1991. However, their evaluation of infected snails from the i BStrain iV @Strain i PRR 12 3 4 5 &6 7 8 9 10 14 12 13 14 15 16 17 18 19 20 21 Day Of Experiment OStrain Vi BStrain Vi BStraintv Strain # 4123 4 8 © 7 & ¥ 10 41 12 13 14 15 16 17 18 18 20 21 Day Of Experiment Figure a Examples of cercarial shedding patterns over 21 days from individual snails, Elimia semicarinata, infected with 1-4 strains of Proterometra macrostoma. (3a=one strain; 3b=two strains; 3c=three strains; 3d=four strains). 138 10 } eer Las] Mean + SE E SUR na aie Ieee gS so ‘ , : 7 ee een * See [ea = 3 Journal of the Kentucky Academy of Science 69(2) Figure 4. Mean + SE eggs / cercarial strain of Proterometra macrostoma. (Number of cercariae evaluated: [=93; I[=26; I1l=97; IV=87; V=60; VII=46; VIII=35). Olentangy River, Ohio, and Carp Lake River, Michigan, showed a notable decrease in strain I when compared with their Kentucky locales. Based on experimental infections of several centrarchid species, Riley and Uglem (1995) concluded that strains of P. macrostoma have preferred centrarchid hosts, and that when the prevalence of a preferred host changes, so does the frequency of its associated strain. For example, Riley and Uglem (2005) linked the high frequency of strain I in Kentucky to the | ere 3s i 8 90 Red Light & 80 Oo * 70 a op 60 € £50 a ie B30 +) € 20 & = 10 0 I Ut high prevalence of its preferred host, Lepomis gulosus (Cuvier), the warmouth bass. By contrast, the decrease of this strain’s frequen- cy in Ohio and Michigan was associated with the absence of this host species. Because the dominance of strain I has remained relatively stable at North Elkhorn during 1990/1991, 2000, 2004, and 2007, it can perhaps be concluded that the frequency of the associated warmouth population at this site has remained unchanged as well. Riley and Uglem (2005) Clwhite Light under red and white light. (Number of cercariae evaluated: [=45; II[=20; VIII=21). Proterometra macrostoma Strains—Rosen et al. Mean + SE Eggs / Worm Day 12-1 Days Postinfection (Strain) @StagelEggs OStagellEggs MGStage Ill Eggs Day 12 - Ill 139 Day 20-1 Day 20 - Ill awe eeen eee eee renee tenet ei ti Nr At PNR HAR NAAR VS VN AARP NAVAN AN A AVS VV HMMM niMnnnnnnanvann wana nnnnnnnnnnnnnnnnannnnannnnnnnnnnd Figure 6. Mean + SE egg stages in Proterometra macrostoma strain I and HI adults on days 12 and 20 postinfection in experimentally infected bluegill at 22.7°C. (Number of worms evaluated: Day 12 (I)=11; Day 12 (IH)=20; Day 20 (I)=11; Day 20 (III)=7). also suggested that strains with consistently low frequencies (e.g., strains II and VI in this study) might be the result of natural crossings of strains. Such crossbreeding between dige- nean strains has been particularly well-docu- mented for the schistosomes. Theron (1984) described two strains of Schistosoma mansoni Sambon based on cercarial shedding patterns. The strain shedding during the day utilized humans as their primary definitive host while the strain that shed during the night utilized a rat definitive host. When Theron and Combes (1988) crossed these two strains, the resulting F, had an intermediate shedding pattern demonstrating the genetic basis for this phenomenon. Due to the nature of asexual polyembryony of P. macrostoma in its snail intermediate host, the cercariae produced are clones of the original egg ingested by the snail. Thus a snail shedding only one cercarial strain is presumed to have only been infected with one or more egg(s) of the same strain. Riley and Uglem (2005) found that 66% of the snails they monitored over two weeks had pure (single) strain infections. They indicated that this high proportion of pure infections was related to the fact that most snails become infected with only one egg. By contrast, only 5.7-23.8% of the snails monitored for three weeks in our study had single strain infections. Though individual snails can contain a number of P. macrostoma rediae asexually producing cer- cariae, only 0.250-0.833 cercariae/snail/day are shed in sample snail populations (Rosen et al. 2005c). The extended time allowed for cercarial development and emergence (i.e., three vs. two weeks) in our study may have been, in part, responsible for an increase in the number of multiple strain infections detected (i.e., our additional week of obser- vation may have provided more time for cercariae from a different egg and possibly strain to develop and be shed). Although further attempts were made to “biologically” characterize several selected strains of P. macrostoma, no new differences were noted. No significant difference was found in the average number of eggs in the seven cercarial strains we evaluated. By contrast, Dickerman (1945) indicated that his type I cercaria had 30-50 eggs, but types II and III never contained eggs. Riley and Uglem (1995) indicated that Dickerman’s (1945) three types correlated with their strains I, IV and VIII, respectively. While strains IV and VIII occasionally lacked eggs in our study, the majority had a number of developing eggs. We also found no significant difference in the vertical swimming distances of strains I, HI, 140 and VIII under red or white light, and all cercarial swimming bursts were initiated at the bottom of the cylinder. By contrast, Riley and Uglem (1995) found that while the swimming burst distance was similar between strains III and VIII under red light, the bursts were initiated at significantly different heights in the water column. We have no explanation for this discrepancy, though, in our experi- ments, it appeared that tactile stimulation of the cercarial tail by the bottom of the cylinder served to initiate the swimming burst. Finally, Riley and Uglem (1995) found that strain I cercariae established much better in bluegill than strain III worms based on their experi- mental infections. However, the subsequent developmental rate (based on the average number of stage I, II, and III eggs) of strain III in bluegill in the present study was not significantly different from strain I at 12 and 20 days PI. This result demonstrated a lack of connection between the initial infectivity of P. macrostoma and its subsequent developmen- tal rate as a strain characteristic in the definitive host. Future work will focus on experimental infections of snails and fish with known strains of this worm to determine whether or not these proposed strains are “true-breeding.” Such studies will also make it possible to conduct breeding experiments between strains provided that cross-fertilization be- tween adult worms takes place in the fish definitive host. In addition, we will continue our work to characterize genetic patterns in selected strains of P. macrostoma using RAPD (random amplification of polymorphic DNA). ACKNOWLEDGMENTS This study was supported by a grant from the Undergraduate Research and Creative Projects Program (URCPP) at Berea College to Ronald Rosen. We acknowledge Steve Marple and Nick Scudlarek for providing us Journal of the Kentucky Academy of Science 69(2) with young bluegill from the Pfeiffer Fish Hatchery at Frankfort, KY. LITERATURE CITED Bryant, C., and H. A. Flockhart. 1986. Biochemical strain variation in parasitic helminthes. Advances in Parasi- tology 25:275-319. Horsfall, M. W. 1934. Studies on the life history and morphology of the cystocercous cercariae. Transactions of the American Microscopical Society 53:311-347. Dickerman, E. E. 1945. Studies on the trematode family Azygiidae. II. Parthenitae and cercariae of Proterometra macrostoma (Faust). Transactions of the American Microscopical Society 64:138-144. Riley, M. W., and G. L. Uglem. 1995. Proterometra macrostoma (Digenea: Azygiidae): variations in cercar- ial morphology and physiology. Parasitology 110:429- 436. Rosen, R., K. Adams, E. Boiadgieva, and J. Schuster. 2000. Proterometra macrostoma (Digenea: Azygiidae): Distome emergence from the cercarial tail and subsequent development in the definitive host. Journal of the Kentucky Academy of Science 61:99-104. Rosen, R., A. Ammons, A. Boswell, A. Roberts, A. Schell, M. Watkins, J. Fleming, B. Jovanovic, A. Sarshad, E. Throop, and F. Zaki. 2005a. Effect of light wavelength and osmolality on the swimming of cercariae of Proterometra macrostoma. Journal of the Kentucky Academy of Science 66:94—-100. Rosen, R., E. Anderson-Hoagland, C. Barton, B. Berry, J. Hardy, and T. Wangmo. 2005b. Natural and experi- mental infections of centrarchid fish by the digenetic trematode, Proterometra macrostoma: Detection of new infections and host histopathology. Journal of the Kentucky Academy of Science 66:101—106. Rosen, R., J. Fleming, B. Jovanovic, A. Sarshad, E. Throop, F. Zaki, and A. Ammons. 2005c. Location of the redia of Proterometra macrostoma (Trematoda: Azygiidae) in the snail, Elimia semicarinata (Gastro- poda: Pluerocercidae), and daily emergence of its cercaria. Journal of the Kentucky Academy of Science 66:89-93. Theron, A. 1984. Early and late shedding patterns of Schistosoma mansoni cercariae: ecological significance in transmission to human and murine hosts. Journal of Parasitology 70:652-655. Theron, A., and C. Combes. 1988. Genetic analysis of cercarial emergence rhythms of Schistosoma mansoni. Behaviour Genetics 18:201—209. J. Ky. Acad. Sci. 69(2):141-151. 2008. Diatom Species Composition and Environmental Conditions at Four Perennial Springs in Western Kentucky and Tennessee Courtney C. Hunt' and Susan P. Hendricks Hancock Biological Station, Murray State University, Murray, Kentucky 42071 ABSTRACT Diatom assemblage composition and water chemistry at four springs of different geological origin in western Kentucky and Tennessee (Land-Between-the-Lakes National Recreation Area, LBL) were examined using unglazed quarry tiles as artificial substrates. Significant differences in diatom species composition and water chemistry were found among the four springs. A total of 25 taxa were identified with 9 taxa being common to all springs. Calciphilous Planothidium lanceolata var. lanceolata (Bréb.) Grunow and Cocconeis placentula var. lineata (Ehr.) V. H. were the dominant species in the carbonate streams of limestone geology (Mint and Panther springs). Acidophilous Eunotia intermedia (Krasske ex. Hust), Achnanthidium minutissimum var. saprophila Kobaysi and Mayama, and Fragilariforma virescens (Ralfs) Williams and Round were most abundant in the streams with siliceous and argillaceous geology (Barnett and Brown Springs). Statistical analyses indicated that diatom species composition was most highly correlated with conductivity and suggested that the underlying geology of the springs played a role in periphyton community composition. KEY WORDS: Diatoms, water chemistry, springs, geology, Land-Between-the-Lakes, Kentucky, Tennessee INTRODUCTION Diatoms are important in aquatic ecosys- tems because they influence all higher trophic levels and are known for their sensitivity to chemical conditions (Lowe 1974): therefore, they are a useful supplement to chemical analyses in assessments of water quality and ecosystem health. Leira and Sabater (2005) found that freshwater diatoms have affinities towards certain ions. A number of taxa have been characterized as preferring calcium-rich or calcium-poor waters. For example, some Cymbella species are calciphilous, some Dia- toma are halophilous, while some Eunotia are acidophilous (Potapova and Charles 2003). Further, the optimal conductivity for many species of Achnanthes, Cocconeis, and Gom- phonema is relatively high compared with species of Eunotia, Diatoma, and Navicula. Other studies have shown that some of these same taxa having high optimal conductivity requirements also have optimal tolerances to high pH levels. Many taxa with low optimal conductivity requirements also have low optimal pH levels (Pan et al. 1996). Require- ments for nutrients also differ among diatom taxa at both the generic and specific levels (Leira and Sabater 2005). ' Corresponding author e-mail: CC_Hunt@yahoo.com Because underlying parent geology is a regional influence on the chemical composi- tion of groundwater, geology should be considered when evaluating diatom assem- blages or periphyton of flowing water, partic- ularly springs and headwater streams (Pan et al. 1996; Leland and Porter 2000; Leira and Sabater 2005). Even in fairly uniform geologic regions, groundwater characteristics may dif- fer over small distances, depending on the locations of recharge zones, the lengths of underflow paths, and the ages of the water sources. Even small differences in conductiv- ity, alkalinity, pH, or nutrient concentrations may be reflected in periphyton colonization rates and species assemblages (Leland and Porter 2000; Rimet et al. 2004). Only limited information exists on periph- yton distributions for western Kentucky streams (Camburn 1982; Hendricks et al. 2006), Lake Barkley (Jarrett and King 1989), or Kentucky Lake (Barnese 1984), while virtually no information exists on the water chemistry or periphyton of Land-Between- the-Lakes National Recreation Area streams (LBL). The purposes of this study were to provide a first assessment of the diatom species present in four perennial springs at LBL using artificial substrates and to deter- mine the relationships between physicochem- ical parameters and diatom species composi- tions. 141 142 Legend A Springs C3 LBL Boundary Figure 1. Map of study spring at Land-Between-the- Lakes National Recreation Area, western Kentucky and Tennessee. STUDY AREA The springs were located in the Land- Between-the-Lakes National Recreation Area (LBL) managed by the U.S. Forest Service. Two springs were in northern Lyon Co., Kentucky (Brown Spring; 36°57.363'N, 88°11.351’W, and Barnett Spring; 37°00.797'N, 88°04.529'W) and two in Stewart County, Tennessee (Panther Spring; 36°30.373'N, 87°58.857'W, and Mint Spring; 36°28.360'N, 88°01.501’W) (Figure 1). The four springs are perennial with uniform flow year around and were well known to the moonshiners of the Prohibition Era for their excellent taste quality (Mauer 1974). The springs are in mature oak-hickory forests and are heavily canopied. Panther and Mint Springs are geologically similar overlying alluvium, Warsaw Limestone and the Fort Payne Formation (USGS Geo- logic Quadrangles: Hamlin and Paris Landing GQ, 1966; Tharpe TN GQ, 1967). The alluvium is composed of silt and clay and in places is stratified with lenses of sand and gravel. Thickness ranges from 0-12 m. War- Journal of the Kentucky Academy of Science 69(2) saw Limestone is composed of tightly packed fine to medium fossil fragments, is inter- bedded with microcrystalline shale and chert limestone, and weathers to silty clay and blocky chert composed of silicified fossil grains or rounded dense nodular chert. Thickness ranges from 41-47 m. The Fort Payne Formation is composed of 10-20% chert and silty limestone that is hard and thin- bedded. The rock weathers deeply to crum- pled layers of residual chert, silt, and tripolitic (diatomaceous earth) clay. The thickness usually is >91 m. Barnett Spring directly overlies the alluvi- um (USGS Geologic Quadrangle: Eddyville KY GQ, 1963). Immediately surrounding the alluvium is the Tuscaloosa Formation, a gravel mixture that is sandy, silty, and in some places argillaceous (containing clay). The rock for- mations are thin-bedded with scattered lenses of chert, sand, and tripolitic (diatomaceous earth) silt or clay. Thickness ranges from 0- 53 m. Brown Spring overlies the Tuscaloosa Formation and sand (USGS Geologic Quad- rangle: Birmingham Point KY GQ, 1966). The sand is slightly argillaceous, silty, and uncon- solidated in places. The sand is composed of poorly sorted medium-fine sand grains, of which 85%-95% are quartz, 5%-15% are chert, and =2% are mica or other minerals. METHODS Unglazed ceramic quarry tiles (area = 91 cm?) (Wellnitz et al. 1996) were placed downstream from each spring in October 2006. Tiles were allowed to colonize with periphyton for approximately four weeks and then were collected in triplicate from each spring every four weeks until July 2007. Tiles were placed as closely as possible to the spring outlet. The tiles at Mint and Panther springs were placed ~8-10 m from their sources, while tiles at Barnett and Brown springs were placed ~3-5 m from their sources. Tiles were collected, put in Zip-Loc® bags, and placed in a cooler with ice for transport to the laboratory for analysis. In the laboratory, periphyton was scraped off each tile with a toothbrush into a beaker. The diatoms were then cleared with a mixture of 30% Hs,O2 (hydrogen peroxide) and KyCr2,O07 (potassium dichromate), followed Diatoms and Springs—Hunt and Hendricks by several rinses using deionized water (Carr et al. 1986). Permanent slides were mounted using Naphrax. Identifications were made under a light microscope (Zeiss Axioplan, Germany) using 1000 magnification and oil emersion (Eaton et al. 2005). A minimum of 200 diatom valves were counted on several transects across each slide surface, and species were identified using standard taxonomic keys (Patrick and Reimer 1966; Krammer and Lange-Bertalot 1986, 1988, and 1991). Water temperature, pH, conductivity, dis- solved oxygen (DO), and turbidity were measured in triplicate in the vicinity of the tiles during each diatom sampling period with a YSI multi-probe meter (Yellow Springs Instruments, Yellow Springs, OH). Triplicate water samples from each site were collected and analyzed in the laboratory for NH4*, NO3 +NOz , SiOs, PO, , Cl. and SO,” using a Latchet Autoanalyzer (Milwaukee, WI) and standard reagents (Prokopy and Switala 1994). Alkalinity was determined using single-point titration methods and formula of Eaton et al. (2005). Light at the stream surface at mid-day was measured at each site with a LI-COR model 185B photometer. Discharge was determined at each site using a Marsh McBirney flow meter and cross sectional area measurements of the stream. Analysis of variance (ANOVA) was _ per- formed to determine if there were any significant differences in physicochemical and nutrient variables among the four springs. The Tukey Kramer multiple comparison procedure was used to identify where the differences occurred. Level of significance was set at a = 0.05. Cluster analysis was used to determine the grouping structure of the springs based on similarities in their physicochemical and nutrient variables and taxa (either genus only or genus and species) composition and to give a visual representation of how similar the taxa composition and physicochemical variables were at each spring. A principal component analysis (PCA) was used to reduce physicochemical data dimen- sions from 13 variables to the most relevant subset based on their eigenvalues. Using SYNTAX 2000, a canonical correlation analy- sis (CCA) was then performed using the reduced data from the PCA results to 143 determine the strength of the relationships between the taxa composition and environ- mental variables at each spring. RESULTS Different diatom assemblages (Table 1) and relative abundances (Figure 2) characterized each of the four springs with Mint and Panther Springs being most similar. Dominant at Mint Spring were Cocconeis placentula var. lineata (74.3% of the total valves count- ed), Planothidium lanceolata var. lanceolata (17.5%), Gomphonema minutum (4.6%), and Eunotia intermedia (1.9%). Dominant at Panther Spring were Planothidium lanceolata var. lanceolata (56.4%), Cocconeis placentula var. lineata (22.3%), Gomphonema minutum (18.42%), and Eunotia intermedia (1.0%). The most abundant species at Barnett Spring were Eunotia intermedia (56.6%), Fragilariforma virescens (28.1%), Gomphonema parvulum (5.07%), Pinnularia gibba (3.5%), Planothi- dium lanceolata var. lanceolata (2.8%), and Gomphonema minutum (1.5%), while the most abundant species at Brown Spring were Eunotia intermedia (52.8%), Achnanthidium minutissima var. saprophila (24.6%), Meri- dion circulare var. circulare (8.0%), Fragilar- forma virescens (5.4%), Gomphonema parvu- lum (2.7%), and Planothidium lanceolata var. lanceolata (2.2%). There were no significant differences in light availability among sites although Panther Spring appeared to have less canopy cover at mid-day (P = 0.05) (Table 2). Discharge was higher at Mint and Panther Springs (0.018 and 0.022 m* sec™') compared with the smaller Barnett and Brown Springs (0.001 and 0.002 m* sec"'). Conductivity and dissolved oxygen (DO) were significantly different among all four sites. Mint and Panther Springs had significantly higher average conductivity (205.4 and 250.8 uS cm™', respectively) com- pared with Barnett and Brown Springs (34.2 and 53.5 US cm!, respectively). Water tem- peratures at each spring did not vary greatly throughout the sampling period with the ranges of <2 °C among the springs. Panther and Barnett Springs both had average tem- peratures of 14.2 °C. Mint Spring averaged 14.8 °C and Brown Spring averaged 13.1°C. Average DO at Mint and Panther Springs (7.1 and 7.9 mg L”', respectively) was significantly 144 Journal of the Kentucky Academy of Science 69(2) Table 1. Checklist of diatom taxa found in Mint, Panther, Barnett, and Brown Springs in western Kentucky and Tennessee (Land-Between-the-Lakes National Recreation Area). Taxon Achnanthidium minutissimum var. saprophila Kobaysi & Mayama Amphora perpusilla (Grun.) Grun Caloneis hyaline Hust. Cocconeis placentula var. lineata (Ehr.) V.H. Cymbella species of unknown or uncertain identity Eunotia intermedia (Krasske ex. Hust) Nérpel & Lange-Bertalot Eunotia paludosa Grun Eunotia species of unknown or uncertain identity Fragilariforma virescens (Ralfs) Williams & Round Frustulia species of unknown or uncertain identity Gomphonema acuminatum Ehr. Gomphonema minutum (Agardh) Agardh Gomphonema parvulum (Kiitz) Gomphonema species of unknown or uncertain identity Meridion circulare var. circulare (Grev.) Navicula placentula Ehr. Navicula species of unknown or uncertain identity Nitzschia species of unknown or uncertain identity Pinnularia gibba var. mesogongyla Ehr. Planothidium lanceolatum var. dubia Planothidium lanceolatum var. frequentisima Planothidium lanceolata var. lanceolata (Bréb.) Grunow Stauroneis species of unknown or uncertain identity Staurosira species of unknown or uncertain identity Surirella species of unknown or uncertain identity lower than at Barnett and Brown Springs (9.2 and 8.5 mg L"'!, respectively). Percent satura- tion ranged from 70% in Mint Spring water to 80% in Brown Spring. Alkalinity at Mint and Panther Springs (1.65 and 2.09 mEq L"', respectively) was considerably higher than at Barnett and Brown Springs (0.15 and 0.20 mEg L"', respectively). pH at Mint and Panther Springs was higher (7.2 and _ 7.4, respectively) as compared with 6.4 and 6.2 at Barnett and Brown, respectively. There were no differences among the springs in ammonium or silica concentrations (Table 3). NO3- + NOs concentrations were much higher in Barnett and Brown springs (0.144 and 0.123 mg L"', respectively) com- pared with Mint and Panther springs (0.053 and 0.070 mg L"', respectively) as were SO,?- concentrations in Barnett and Brown springs (5.54 and 10.87 mg L"', respectively) com- pared with Mint and Panther springs (2.11 and 2.94 mg L"', respectively). PO4* concen- trations were significantly higher in Mint and Panther springs (0.092 and 0.042 mg L"', respectively) compared with Barnett and Brown springs (0.004 and 0.006 mg L"', Mint Panther Barnett Brown X X XxX X Xx X X X X X XxX XxX XxX X X X X X X X X X XxX XxX X X X XxX XxX X XxX XxX X X X X X X Xx X X Xx X XxX X XxX X X X X X X XxX X X X X XxX X xX X X X X respectively), and Cl concentrations were found to be significantly higher in Panther and Barnett springs (2.16 and 1.76 mg L", re- spectively) compared with Mint and Brown springs (1.65 and 1.6 mg L"', respectively). Cluster analysis showed that there were very distinct taxa compositions at each spring (Figure 3). Cluster analysis also showed that Mint and Panther springs were similar in physicochemical and nutrient variables (Fig- ure 4). Mint and Panther springs had signif- icantly higher concentrations of PO, (and significantly lower concentrations of NO3” + NOz,") compared with Barnett and Brown springs. Barnett and Brown springs differed from each other more in their physicochem- ical and nutrient variables. Although there was a distinct dominance of certain taxa at each spring, there were taxa present that comprised less than 1% of the total abundance (Table 1). Principle components analysis (PCA) was performed on all 13 environmental variables. Those with eigenvalues greater than 0.7 were selected as variables that explained the most variance in species composition at each spring (Table 4). Six variables selected by the anal- Diatoms and Springs—Hunt and Hendricks Mint Spring MB C.placentula 100 ZiZza P. lanceolata ESE G.minutum Gia E intermedia 80 g ro) = 60 @ xe} ¢ =| a ¢ aoe 3 o a : | 20 ib i , tf 7 it OH Oh : | y : ‘ ‘| i Is T | é ; i; lax la 1 i | IT A Nov Dec Jan Feb March April May June July Month (Nov 2006-July 2007) MMM E. intermedia Barnett Spring F. virescens GESS G. parvulum 400 EEE P. gibba fd P. lanceolata EEE G. minutum 80 < C0) & 60 E | fe 3 a | ER G. parvulum Eig) P. lanceolata GH CC. placentula 80 = ca) e Ss 60 xo) < 3 a x g 40 i ; 5 | o 5 | a | 20 } y h (+ wl Be BE ‘|: f A pole GS: Bes 3 : is a: ; 0 ee Ae Ao as ee Bh oh LE ir Nov Dec Jan Feb March April Month (Nov 2006-July 2007) May June = July Figure 2. The relative abundance of taxa found at each spring in Land-Between-the-Lakes National Recreation Area, western Kentucky and Tennessee. ysis were conductivity, alkalinity, PO,’ , SO, NOs” + NOs, and pH. PCA results showed the component loadings for the most important environmental variables (Table 5). Conductivity, alkalinity and PO,°? explained 68% of the variance in species composition at each spring. The first component axis ex- plained 41.6% of the total variance in species composition between the springs. The second component axis explained 15.6% of the total variance. The component loadings showed that conductivity, alkalinity, PO, , SO,*, Table 2. Comparison of mean physicochemical variables (s.d.) among four springs in western Kentucky and Tennessee (Land-Between-the-Lakes Natioanl Recreation Area). Means marked with the same letters (a—d) are not significantly different from each other (P = 0.05). Variable Mint Conductivity (uS cm™') 205.4? (4.8) DO (mg L~) 1 (0.4) Temperature (°C) 14.8 (0:1) Light (umoles m~ sec"') 501* (408) Alkalinity (mEq L”) 1.65: (0.09) Flow (m° sec™') 0.018 wu 006) pH 7.2 Panther Barnett Brown 250.8 (14.8) 34.2¢ (3.5) 500° (7.6) 7.2? (06) 9.2° (0.8) 8.5¢ (0.8) 14.2° (0.2) 14.2 (0.5) 1S2i(1.2) 288: (54) 621* (110) 581* (107) 2.09" (0.15) 0.15* (0.03) 0.20° (0.04) 0. = : 007) 0.001 (0.000) 0.002 (0.002) 6.4° 6.2? 146 Journal of the Kentucky Academy of Science 69(2) Table 3. Comparison of mean concentrations (s.d.) of nutrients in four perennial springs in western Kentucky and Tennessee (Land-Between-the-Lakes). Means marked with the same letters (a-d) are not significantly different from one another other (P = 0.05). All variables are expressed in mg L”. Variable Mint Panther Barnett Brown PO,g? 0.092: (0.007) 0.042 (0.003) 0.004 (0.003) 0.006 (0.003) NO3 + NOs” 0.053" (0.002) 0.070° (0.002) 0.144° (0.027) 0.1234 NH,4* 0.004" (0.004) 0.004" (0.004) 0.003* (0.003) 0.004 (0.003) SiOg 8.717: (0.168) 8.661: (0.206) 7.207* (3.972) 8.611* (4.035) Cr 1.648* (0.034) 2.156? (0.127) 1.758° (0.138) 1.598 (0.058) SO,° 2.112* (0.408) 2.937" (0.308) 5.539 (0.650) 10.868* (0.710) NO3; + NOs, and pH were all highly correlated with the first axis and that pH and SO,?> were correlated with the second axis. Using the six variables that explained the most variance in species composition, a canonical correlation analysis (CCA) was performed to determine the strength of the relationship among the most abundant species and the physicochemical variables at each spring. Based on the CCA (Figure 5), diatom species compositions at Mint and Panther Springs were most influenced by conductivity, alkalinity, pH, and PO,* concentrations. Species composition at Barnett Spring was most influenced by NO3° + NOg” concentra- tions, while the species composition at Brown Spring was most influenced by SO,?- concen- trations. Of the seven most abundant taxa, each had distinct correlations with one or more water quality parameters (Table 6). Spring water conductivity, alkalinity, pH, PO, NO3 + NO,” and SO, concentrations appeared to have the most influence on diatom species assemblages. Achnanthidium minutissimum was negatively correlated with pH and posi- tively correlated with SO,?- concentrations. Cocconeis placentula was positively correlated with alkalinity and negatively correlated with NO; + NOs, and SO, concentrations. Eunotia intermedia was negatively correlated with conductivity, pH, alkalinity, and PO,” but positively correlated with NO3;” + NOg” Group | Group 2 (Mint Panther Group 3 Spring) Spring); 4) (Bamet Group 4 Spring) (Brown 25 Spring) = 5 E BAS a 13 7 6 4 2 1817 5 8 9 10 11 12 13 16 14 15 19 25 21 22 24 23 20 26 27 35 30 36 28 33 31 34 29 32 Figure 3. Cluster analysis showing the grouping structure of the springs based on their taxa composition. Each spring has a distinct taxa composition: Numbers 1-9 (Group 1) represent the taxa found at Mint Spring during the study period. Numbers 10-18 (Group 2) represent the taxa found at Panther Spring; numbers 19-28 (Group 3) represent the taxa found at Barnett Spring, and numbers 29-36 (Group 4) represent the taxa found at Brown Spring. Diatoms and Springs—Hunt and Hendricks : Group | 3B (Panther A and Mint Springs) 147 Group 2 (Barnett Group 3 Spring) (Brown Spring) 123 7 4 9 5 8 6 10 17 14 18 11 13 15 16 12 19 20 25 24 26 27 21 23 22 28 29 36 30 35 31 32 33 34 Figure 4. Cluster analysis showing the grouping structure of the springs based on their physicochemical and nutrient variables. Numbers 1-9 represent the physicochemical measurements collected for Mint Spring, November through July. Numbers 10-18 represent the physicochemical measurements collected for Panther Spring; numbers 19-28 represent the physicochemical measurements collected for Barnett Spring and numbers 29-36 represent the physicochemical measurements collected for Brown Spring. Mint and Panther spring (Group 1) are physicochemically similar; Barnett Spring is Group 2, and Brown Spring is Group 3. and SO,’ concentrations. Fragilariforma vir- escens showed a negative correlation with conductivity and alkalinity. Gomphonema minutum was positively correlated with alka- linity. Pinnularia gibba was negatively corre- lated with conductivity and alkalinity, and positively correlated to NO3;— + NOs. Pla- nothidium lanceolata was positively correlated with conductivity, pH, and alkalinity, and negatively correlated with SO, concentra- tions. DISCUSSION The diatom flora of the four spring streams was not diverse and primarily contained taxa known to be common in western Kentucky Table 4. Physicochemical variables with eigenvalues >0.7 explaining the highest percentage of total variance in species composition among four perennial springs in western Kentucky and Tennessee (Land-Between-the- Lakes). Component Eigenvalue Percentage of Variance Explained Conductivity (In) 6.241 41.6 Alkalinity 2.337 15.6 PO4? 1.657 11.0 SO,” 1.078 fe NO; + NO.” 1.030 6.9 pH 0.797 5.3 streams (Camburn 1982, Hendricks et al. 2006). Southern LBL’s Mint and Panther Springs had higher average conductivity, alkalinity, and pH indicative of their underlying lime- stone geology as compared to northern LBL’s Barnett and Brown Springs which were underlain by siliceous and argillaceous geolo- gy. Chetelat et al. (1999) found that phospho- rous concentrations were positively correlated with specific conductance, suggesting strong relationships between nutrient concentrations and drainage basin geology. Although phos- phorus is usually the limiting nutrient in streams, nitrogen can be limiting in streams that have geologic sources of phosphorus, such as limestone bedrock (Hendricks and White 2000). Barnett and Brown Springs also had much higher concentrations of SO,’ than Mint and Panther Springs. SO,’ has many sources including weathering of sedimentary rocks and atmospheric deposition (Allen 1995). Mayer et al. (2007) found that high SO, concentrations in stream water are caused by chemical weathering involving the oxidation of sulfides in siliceous rocks. SO,2- and bicar- bonate concentrations tend to be inversely related in stream water, especially in areas where alkalinities are lower (Allen 1995). 148 Table 5. Principle Components Analysis results show- ing the component loadings (correlations of each variable on each component) for the environmental variables explaining the most variance in species composition among four springs in western Kentucky and Tennessee (Land-Between-the-Lakes National Recreation Area). Variable Component 1 Component 2 Conductivity (In) 0.951 0.061 Alkalinity 0.950 O.1t PO, 0.840 0.057 SO,4” —0.836 =(1225 NO; + NOs” —0.791 0.052 pH 0.761 0.349 Since atmospheric deposition of SO,’ has decreased over the past 20 years in Kentucky Lake (Yurista et al. 2004), it is likely that geologic weathering is the main source of SO,?- in LBL streams. Although we do not know if specific sulfide minerals such as pyrite (FeS:) are proximate to any of the springs, LBL has been known historically for it’s low- grade iron mining activities (Harris 2002) Axis 2 mS, -3 Journal of the Kentucky Academy of Science 69(2) whereby such minerals may have been important. The most abundant species in Mint and Panther Springs (Cocconeis placentula, Pla- nothidium lanceolata, and Gomphonema min- utum. C. placentula and P. lanceolata (syn. Achnanthes lanceolata)) are classified as being alkaliphilous occurring at pH 7 with best development over pH 7 (Lowe 1974). C. placentula has an optimal conductivity of 270 uS cm" and P. lanceolata has an optimal conductivity of 286 uS cm™! (Potapova and Charles 2003). Many species of Gomphonema are classified as indifferent with best develop- ment around pH 7 (Lowe 1974). G. minutum has been found to have an optimal conduc- tivity of 324 uS cm"! (Potapova and Charles 2003). Barnett and Brown springs both had abundances of Eunotia intermedia. Many species of Eunotia have been described as acidophilous (occurring at pH 7 with best development below pH 7) to indifferent (Lowe 1974), and many species have an Mint and Panther Axis | Figure 5. Triplot based on the canonical correlation analysis. The circled numbers represent the taxa found at each spring November through July. Numbers 1-18 represent the taxa found at Mint and Panther springs. Numbers 19-27 represent the species composition at Barnett Spring. Numbers 28-36 represent the taxa found at Brown Spring. Each line represents the physicochemical variables which had the most influence on the taxa composition at each spring (conductivity, alkalinity, pH, PO, °, NO; + NO. °, and SO, ”). Diatoms and Springs—Hunt and Hendricks 149 Table 6. Canonical correlation results showing the correlation coefficients between the most abundant species and the physicochemical and nutrient variables at each spring. Genus and species names are found in Table 1. (*) represents correlation coefficients =0.5. Variable A .min C. lin E. int F. vir G .min G. par M. cir P. gib P. lan Conductivity —0.376 0.730% =0;857" ~ =0:647* 0487 =0.361. =—0179 —0,647* O72" pH —0.683* 0.473 —0.631* —0.144 0.449 -0377 —-0.255 —0.154 0.664* Alkalinity —0.482 0.720* -—0.856* —0.565* 0.528" =0:450° .=0,226° °—0.5657 0.765* PO, —0.421 O:927* ~—O721*. 0490 0.210 —0.298 —0.188 —0.479 0.307 NO3 + NOs 0.215 ==) 730" 0.833* 0.443 —0.279 0.199 —0.044 0.551* —0.465 SO4 0.766" -=0.672* 0.682* 0.215 —0.393 0.202 0.400 0.171 —0.554* optimal conductivity range from 48 to 116 wS cm"! (Potapova and Charles 2003). Barnett Spring had a good population of Fragilariforma virescens (syn. Fragilaria vir- escens). F. virescens is classified as indifferent (Lowe 1974) and species of Fragilarioforma have an optimal conductivity of 86 u~S cm”! (Potapova and Charles 2003). Achnanthidium minutissimum (syn. Achnanthes minutissima) which has been classified as indifferent (Lowe 1974) with an optimal conductivity of 229 uS cm™ (Potapova and Charles 2003) was common in Brown Spring. Because the pH and conductivity of Brown Spring do not meet the optimal environmental conditions for A. minutissium, it is possible that the high SO, concentrations had a positive influence on the abundance of this species at this spring. Bonny and Jones (2007) found that A. minutissimum was abundant in springs with high SO,?" concentrations. It is also important to note that the abundance of Meridion circulare was very high in the month of March in Brown spring. Although M. circulare generally achieves maximum abundance dur- ing the spring (Krecji and Lowe 1987) and is considered an alkaliphilous species (Lowe 1974), it remains unclear why it became so dominant. There were many taxa present at each spring that never became dominant. All of these taxa occur at a pH 7, but the conditions might not have been optimal in other ways for their development. For example, many spe- cies of Amphora are classified as alkaliphilous (Lowe 1974), but the optimal conductivity for each are very high, ranging from 515- 634 US cm™' (Potapova and Charles 2003). The extremely high optimal conductivity requirements for Amphora perpusilla could be one reason why it did not become a dominant species in Mint Spring. Many species of Caloneis are also classified as alkaliphilous (Lowe 1974) and have high optimal conductivities (>365 US cm~!, Pota- pova and Charles 2003). We conclude that these environmental conditions were not met in Barnett Spring for Caloneis hyalina. Optimal environmental conditions for many species of Cymbella vary greatly. They have been classified as acidophilous, indifferent, alkaliphilous and alkalibiontic (occurring only in alkaline water) (Lowe 1974). Their optimal conductivity range is 89-347 uS cm"! (Pota- pova and Charles 2003). Cymbella was present in both Brown and Barnett Springs but was never abundant probably because of non-optimal conductivities. In summary, it is apparent that most of the dominant species occurring in the southern LBL springs also were present (but not dominant) in the northern LBL. springs. Likewise, most of the dominant species in the northern springs were present (but not dominant) in the southern springs. ACKNOWLEDGMENTS We thank Dr. Mark Luttenton of Grand Valley State University, Allendale, MI, for his generous assistance with algal identifications. We thank Dr. Kate He for advice on the PCA and CCA programs. We thank personnel at the Hancock Biological Station for assistance with nutrient analyses and for the use of laboratory and field equipment. We also thank numerous graduate students for their assis- tance in the field and laboratory. This paper was submitted as partial requirement for the Masters in Water Science program at Murray State University, Murray, KY. This research was supported, in part, by USGS 104b funding through the Kentucky Water Re- 150 sources Research Institute of the University of Kentucky. The views and conclusions con- tained in this document are those of the authors and should not be interpreted as necessarily representing the official policies, either expressed or implied, of the U.S. Government. We acknowledge the support and collaboration of the Department of the Interior, U.S. Geological Survey, and the Kentucky Research Foundation, under Grant # 06HQG0087. LITERATURE CITED Allen, D. J. 1995. Stream Ecology: Structure and func- tion of running waters. Chapman and Hall, London, UK. Barnese, L. E. 1984. The algal composition and popula- tion dynamics in selected embayments and channel areas of Kentucky reservoir. Masters Thesis, Depart- ment of Biological Science, Murray State University, Murray, KY. Bonny, S. M., and B. Jones. 2007. Diatom mediated barite precipitation in microbial mats calcifying at Stinking Springs, a warm sulphur spring system in Northwestern Utah, USA. Sedimentary Geology 194:223-244. Camburn, K. E. 1982. The diatoms (Bacillariophyceae) of Kentucky: A checklist of previously reported taxa. Transactions of the Kentucky Academy of Science 43:10-20. Carr, J. M., G. L. Hergenrader, and N. H. Troelstrup, Jr. 1986. A simple, inexpensive method for cleaning diatoms. Transactions of the American Microscopical Society 105:152-157. Chetelat, J., F. R. Pick, A. Morin, and P. B. Hamilton. 1999. Periphyton biomass and community composition in rivers of different nutrient status. Canadian Journal of Fisheries and Aquatic Sciences 56:560—-569. Eaton, A. D., L. S. Clesceri, E. W. Rice, and A. E. Greenburg. 2005. Standard methods for the examina- tion of water and wastewater. American Public Health Association. 2—28, 7-53, 10-14, 10-19, 10-37. Harris, S. E. 2002. Summary review of geology of Land-Between-the-Lakes. Pages 19-52 in E. W. Harris and J. S. Fralish (eds). Land Between the Lakes, Kentucky and Tennessee: Four Decades of Tennessee Valley Authority Stewardship. The Center for Field Biology, Austin Peay University, Clarksville, TN. Publ. No. 16. Hendricks, S. P., M. R. Luttenton, and S. W. Hunt. 2006. Benthic diatom species list and environmental con- ditions in the Little River Basin, western Kentucky, USA. Journal of the Kentucky Academy of Science 67:22-38. Hendricks, S. P., and D. S. White. 2000. Streams and groundwater influences on phosphorus biogeochemis- try. Pages 221-235 in J. B. Jones and P. J. Mulholland Journal of the Kentucky Academy of Science 69(2) (eds). Streams and groundwaters. Academic Press, San Diego, CA. Jarrett, G. L., and J. M. King. 1989. The diatom flora (Bacillariophyceae) of Lake Barkley. Final report to the U.S. Army Corps of Engineers, Nashville District, Nashville, TN. Krammer, K., and H. Lange-Bertalot. 1986. Bacillario- phyceae 1. Teil: Naviculaceae. in Siisswasserflora von Mittleuropa Band 2/1. Gustav Fischer Verlag, Stuttgart. Krammer, K., and H. Lange-Bertalot. 1988. Bacillario- phyceae 2. Teil: Bacillariaceae, Epithemiaceae, Surir- ellaceae. in Siisswasserflora von Mittleuropa Band 2/2. Gustav Fischer Verlag, Stuttgart. Krammer, K., and H. Lange-Bertalot. 1991. Bacillario- phyceae 3. Teil: Centrales, Fragilariales, Eunotiaceae. in Siisswasserflora von Mittleuropa Band 2/3. Gustav Fischer Verlag, Stuttgart. Krejci, M. E., and R. L. Lowe. 1987. The seasonal occurrence of macroscopic colonies of Meridion circulare (Bacillariophyceae) in a Spring-Fed Brook. Transactions of the American Microscopial Society 106:173-178. Leira, M., and S. Sabater. 2005. Diatom assemblages distribution in catalan rivers, NE Spain, in relation to chemical and physiographical factors. Water Research 39: 10-82: Leland, H. V., and S. D. Porter. 2000. Distribution of benthic algae in the upper Illinois River basin in relation to geology and land use. Freshwater Biology 44:279-301. Lowe, R. L. 1974. Environmental requirements and pollution tolerance of freshwater diatoms. Office Research and Development, U.S. Environmental Pro- tection Agency, Cincinnati, OH. EPA 670/4-74-005. Mauer, D. 1974. Kentucky moonshine. University of Kentucky Press, Lexington, KY. Mayer, B., J. B. Shanley, S. W. Bailey, and M. J. Mitchell. 2007. Identifying sources of stream water sulfate after a summer drought in the Sleepers River watershed (Vermont, USA) using hydrological, chemical, and isotope approaches. Geophysical Research Abstracts 9:09694. Pan, Y., J. Stevenson, B. H. Hill, A. T. Herlihy, and G. B. Collins. 1996. Using diatoms as indicators of ecological conditions in lotic systems: a regional assessment. Journal of North American Benthological Society 15:481—495. Patrick, R., and C. W. Reimer. 1966. The Diatoms of the United States. Potapova, M., and D. F. Charles. 2003. Distribution of benthic diatoms in U.S. rivers in relation to conductivity and ionic composition. Freshwater Biology 48: 1311-1328. Prokopy, W. R., and K. S. Switala. 1994. QuikChem Automated Jon Analyzer Methods Manual. Rimet, F., L. Ector, H. M. Cauchie, and L. Hoffman. 2004. Regional distribution of diatom assemblages in the headwater streams of Luxembourg. Hydrobiologia 520:105-117. Diatoms and Springs—Hunt and Hendricks 151 USGS Geologic Quadrangle: Birmingham Point KY GQ. 1966. USGS Geologic Quadrangle: Eddyville KY GQ, 1963. Benthological Society 15:496-507. USGS Geologic Quadrangle: Hamlin and Paris Landing Yurista, P. M., D. S. White, G. W. Kipphut, K. Johnston, GQ. 1966. USGS Geologic Quadrangle: Tharpe TN G. Rice, and S. P. Hendricks. 2004. Nutrient patterns in GQ, 1967. a mainstem reservoir, Kentucky Lake, USA, over a 10- Wellnitz, T. A., R. B. Rader, and J. V. Ward. 1996. Light year period. Journal of Lake and Reservoir Manage- and grazing mayfly shape periphyton in a Rocky ment 20:148—163. Mountain stream. Journal of the North American J. Ky. Acad. Sci. 69(2):152-163. 2008. Assessing the Impacts of Coal Waste on Residential Wells in the Appalachian Region of the Big Sandy Watershed, Kentucky and West Virginia: An Exploratory Investigation Andrew Wigginton! Department of Biology, University of Kentucky, Lexington, Kentucky 40506 Jay Mitchell Kentucky Department for Natural Resources, Division of Abandoned Mine Lands, Frankfort, Kentucky 40601 Ginger Evans* Department of Geography, Eastern Kentucky University, Richmond, Kentucky 40475 and Stephanie McSpirit Department of Anthropology, Sociology and Social Work, Eastern Kentucky University ABSTRACT This paper examined issues surrounding coal waste and its potential impacts on residential private wells by reviewing the existing literature to identify the possible issues and parameters associated with coal waste and its possible effects on private wells. Using well water data from the Big Sandy Region of Kentucky and West Virginia (n=42), drinking water quality was examined using standard heavy metal parameters associated with coal waste: arsenic, cadmium, chromium, iron, lead, manganese, zinc, and sulfate. Findings showed significantly more wells in sub-watersheds with coal waste impoundments with iron levels above secondary drinking water standards. A review of similar wells from the Kentucky Groundwater Database Repository showed a similar trend. This pattern warranted further study of Fe as a possible coal slurry waste marker. Other general findings revealed high concentrations of manganese, lead, and arsenic across our sampling of cases. Levels of these metals were high in Appalachian rock, so linking their levels to coal mining is problematic. Overall, findings suggested that residential well water in the coal mining area of the Big Sandy region of Appalachia may be of variable and sometimes unhealthy quality. KEY WORDS: Coal waste, impoundments, well water, drinking water quality INTRODUCTION The Appalachian region of eastern North America is a major coal producing region (Figure 1; US DOE 2006). On numerous occasions, residents of this region have voiced concern for the effects of coal mining on various aspects of their lives, especially the ground water used by many residents (McSpirit et al. 2005, 2007; Scott et al. 2005). Mining activities that have generated the most concern include mountain top removal, coal slurry impoundments, and the injection of coal wastes into deep mines (Stout and Papillo 2004; WV Joint Judiciary Sub- committee 2006). ' Corresponding author e-mail: ajwigg0@email.-uky.edu Coal wastes can contain a variety of metals that can cause health effects, including arsenic, barium, cadmium, chromium, lead, selenium, silver and copper, iron and manga- nese. The US EPA concluded that the coal wastes injected into deep mines could be a source of metal contamination if they entered an aquifer (US EPA 1985). Poorly managed coal waste impoundments may generate acid which can increase the solubility of heavy metals in coal waste (Brady 1998). In West Virginia, concerns over deep-mine disposal methods have lead to legislative hearings and a directive to study these concerns more deeply (WV Joint Judiciary Subcommittee 2006; HNN Staff 2007). In October, 2000 a 68-acre reservoir released impounded slurry and coal waste 152 Assessing the Impacts of Coal Waste—Wigginton et al. 153 The Big Sandy Region of the Southern, Eastern Appal R § : Westem, Central, and Eastern | Legend Southern Coal Field Region : Appalachian Coal Field Region |—#* Counties ||] states Study Counties ‘Tennessee Figure 1. into underground mines that then blew-out two mine portals, releasing an estimated 300 million gallons of slurry and sludge waste into area creeks and water ways in Martin County, KY (US DOI 2002). Environmental settlement money from the spill was used for assessments of the local reservoir, the local water company, and metal levels in resident’s homes. Several wells were examined in the latter study that led to our greater interest in the possible effects of the Martin County spill and other coal mining activities on well water quality. The US EPA (1985) and USGS (1989) in past site assessments have identified the following constituents as associated with coal slurry waste: arsenic, cadmium, chromium, iron, lead, zinc, manganese and sulfate. In general, we were interested in exploring whether some of these chemicals could be used in monitoring and tracing possible coal waste impacts and differences between sites in coal impacted areas. Therefore, it might be West Virginia Mirginia North Carolina 40 Miles The Big Sandy Region of the Southern, Eastern Appalachian Coal Field Region. expected that more downstream wells in sub- watersheds with coal waste impoundments would experience lower drinking water quality than other wells in our study area in Southern Appalachia. METHODS Considering the possible impacts of coal waste on the environment and human health, our principle research question was to explore possible coal waste effects on drinking water from residential wells. While our previous findings reinforced past US EPA and ATSDR assessments that indicated no long term effects of the 2000 event on the county public water system (Lasage and Caddell 2006; McSpirit and Wigginton 2006; Wigginton et al. 2007), the next question was whether private, residential wells had been affected. Although early sampling and testing by state regulatory agencies had shown tested wells to be within acceptable drinking water standards (KY DOW 2000), many residents in the 154 impacted area of Martin County have re- mained concerned about the possible slow migration of heavy metals and other constit- uents into their aquifers from coal waste and coal waste impoundments. Each of our three study sites was within the Big Sandy watershed region of Kentucky and West Virginia (Figure 1) in an area with coal waste impoundments and other coal mining activities such as surface mining or deep mine slurry injection. The Big Sandy region starts the area of southern Appalachia, which, according to USGS reports, is less prone to acid mine drainage (AMD) due to more calcareous and carbonate-based strata than northern Appalachia (USGS 1989). However, the Big Sandy region has been the site of the three previously cited coal waste cases that have been handled by the US EPA or ATSDR. Accordingly, we focused our analysis on three sites within this region. Site 1, Martin County, Kentucky, was the site of the 2000 coal waste release, where the bottom of the Big Branch Slurry impound- ment breached into underground mine works, releasing over 300 million gallons of coal slurry and sludge material into area waters. Martin County has a small land area and population of only 12,093 people (U.S. Census Bureau, State and County Quick Facts, http:// www.census.gov/). The principal economic activity is coal mining. In 2005, there were 11 active mining operations with 3 surface permits and 8 underground mining operations in 2005. Together, they produced 5.3 million - tons of coal that year (U.S. Department of Energy 2006). Besides the Big Branch Slurry impoundment, there are 4 other coal waste impoundments in the county based on online data from the Coal Impoundment Project (Coal Impoundment, Location and Information Sys- tem (LIS) http:/www.coalimpoundment.org/). Whereas other studies had used more focused approaches, we applied more random methods in data collection to avoid any bias in our resulting data. During late summer (August and September), well water samples were collected based on door to door solici- tations by our field from 24 wells across the county (Figure 2). Given the length of time since the 2000 coal slurry release, it is possible that this event or other smaller incidences of “black water” have impacted the aquifers Journal of the Kentucky Academy of Science 69(2) throughout the county and hence, the ratio- nale for our focus on a wide geographic scope of the county. Site 2, Mingo County, West Virginia, also is heavily involved in coal production and lies parallel to Martin County, on the other side of the Tug Fork of the Big Sandy River. In 2005, there were 28 active mining operations, 15 deep mines and 13 active surface or moun- tain-top removal sites that produced 14.5 mil- lion tons of coal (U.S. Department of Energy 2006), as well as 12 coal waste impoundments (Coal Impoundment, LIS). The population of Mingo is higher than that of Martin County with 27,100 residents (US Census Bureau). The Rawls, Merrimac and Lick Creek area (near Williamson, WV) was the site of the most recent ATSDR investigation into the effects of deep mine slurry injection methods on domestic wells which resulted in a warning for infants and those with liver and gastroin- testinal disease to refrain from drinking from domestic supplies due to high concentrations of manganese (ATSDR 2004, 2005). In this case, we refrained from sampling in the Rawls, WV, area due to a pending civil suit between residents and the coal company. For a discussion of data and findings from this area, see ATSDR (2004); Stout and Papillo (2004), and ATSDR (2005, -a response to Stout). In collecting data on other places in the county, we relied on a field guide who was a member of a concerned citizen’s group. This more focused method allowed us to collect from 11 wells in mid- December 2006 (Figure 2). Site 3 in Pike County, Kentucky, has a population of 66,860 people (U.S. Census). It is the principal coal-producing county in Kentucky. In 2005, there were 112 active mine operations in Pike County (68 under- ground, 44 surface mines) and they produced 28.4 million tons of coal (U.S. Department of Energy 2006). There are 17 coal waste impoundments in this county (Coal Impound- ment, LIS). It also was the county where the 1985 coal waste injection case took place in which the US EPA concluding that an ‘imminent and substantial” threat existed to groundwater as a consequence of these deep- mine disposal methods. Our specific study site was a far distance from that site. In Pike County, we collected well water samples from homes that were Assessing the Impacts of Coal Waste—Wigginton et al. 155 Ball Watersheds @ Coal!lmpoundments & Sample Well Sites by Sub-Watershed and Coal Waste Impoundments { 10 1.475,000 Figure 2. Sample Well Sites by Sub-Watershed and Coal Waste Impoundments. directly down gradient from an active moun- tain-top removal site. Residents in this area have complained about damage and contam- ination from these operations to their private wells, but these complaints do not revolve specifically around coal waste. Well water samples, in this case, were collected based on communication with a concerned resident who then served as our field guide. With their assistance, we collected samples from 8 wells in and around a small hamlet in mid October 2006. Water samples were collected from a total of 42 domestic wells from Martin, Mingo, and Pike counties during low-flow hydrolog- ical conditions of mid to late summer and mid to late fall (early winter). Five samples were to be collected on each home: a well sample for detailed analysis, a second well sample for portable test kit analysis, cold water tap, hot water tap, and water heater drain valve. Some homes could not be sampled at all locations because of poor access, especially for water heaters. In this research, only well water data are reported in detail. Wells were sampled by choosing the outside tap closest to the well pump. Care was taken to ensure that the tap exited the line before any water softener. Water was allowed to run for several minutes until a change in temperature was noticed, indicating that any pressure tanks had been flushed and new water was being drawn from within the well casing (Haman and Bottcher 1986). In the case of one well with no pump apparatus, it was sampled with a_ bottle attached to a weighted line. New polyethylene (PE) bottles were used to collect all samples. After collection, sample bottles were placed in plastic, resealable bags and kept on ice until they were acidified using trace metal grade nitric acid (HNO3; US EPA 2005). Date and time that samples were collected was recorded in our field logs as well as other descriptive data. Global Positioning System (GPS) coordinates were logged on general vicinity (generally within 30 m) rather than exact well location. A more rigorous study of specific hydrological conditions would imply collecting GPS data on the exact location of 156 each domestic well and the mapping of other land and hydrological characteristics of inter- est on site. However, such mapping would necessitate formalizing additional consent and confidentiality agreements between research- ers and university institutional review boards before such information could be recorded and catalogued on homeowners. Gathering more specific participant information that could potentially link participants to particular wells would have likely reduced the number of residents willing to work with us and hence, our rationale for collecting GPS data within the general (30 m) vicinity. Field data were also collected on well-depth and whether wells were drilled or hand-dug. In this exploratory study, well-depth was collected based on homeowner estimates. Past USGS studies have used either driller logs and/or field probes to better measure well depth, but for our purposes residential self- reports were adequate because we were pursuing citizen based science and the cost and potential for damaging residential wells if more accurate measurement were made would have discouraged participation. We also collected background data about house- hold size (number of people present in the household) to serve as a proxy assessment of heavy versus light water use. Information was also collected on the age of the home and plumbing type: copper, iron, galvanized steel, or polyvinyl chloride (PVC). Analytical Parameters and Laboratory Methods In its investigative report after the 2000 coal sludge event, the National Research Council stated, as one of its recommendations, the need to characterize the constituents of coal slurry to better “aid in monitoring schemes” (NRC 2002, p. 127). In our efforts to start to characterize the constituents of coal slurry, metals were initially selected for analysis based on their prevalence in early US EPA and ATSDR reports and known presence in coal. They included arsenic, barium, cadmi- um, chromium, cobalt, copper, iron, lead, manganese and selenium (U.S. EPA 1985; Booth et al. 1999; ATSDR 2001; Huggins 2002; Wagner and Hlatshwayo 2005). Sulfate was also selected for analysis because it can be found in acid mine drainage or water pumped Journal of the Kentucky Academy of Science 69(2) from flooded mines (Grey 1998; Batty et al. 2005; Tiwary 2001). All samples were analyzed for metals at the Ecotoxicology and Environmental Assessment Laboratory (EEAL) at the University of Kentucky using a Varian Vista MPX Simulta- neous Inductively Coupled Plasma Optical Emission Spectrometer (ICP-OES) and meth- ods of the APHA (1995). Calculated metal concentrations were corrected back to check standards. Samples collected during the field sweeps were analyzed for sulfate during May, 2007 at the Environmental Research and Training Laboratory (ERTL), also at the University of Kentucky using a Dionex ICS- 2500 Ion Chromatograph and methods devel- oped by the U.S. EPA (Method 300.0; 1993). Geographic Information System Analysis Prior to statistical analysis, additional geo- coding methods were used in a Geographic Information System (GIS) to code wells based on sub-watershed and whether the well was in a sub watershed with an active or reclaimed impoundment. Other GIS data were also compiled based on well proximity to a deep mine site or surface mine site, but preliminary analysis showed these other factors to be statistically insignificant in explaining well water quality perhaps due, in part, to the variability in data mapping of active deep mines and surface mine sites. Also, slurry injection point data for Mingo County was viewed in a GIS, but the limited number of cases (n = 11) prevented systematic analysis on this dimension. In the conclusion of this paper, future research needs are mentioned with emphasis on developing GIS methods. For the following analysis, the main spatial tool in explaining water quality was whether a well was in a sub-watershed with either an active or reclaimed impoundment and down stream of it (Figure 2). Statistical Methods Although somewhat different methods were used to determine which sites to sample, it was not anticipated that this would hinder data analysis since guides could not know whether given sites were contaminated or not. Effectively, the more focused approach used in Pike and Mingo Counties simply allowed us Assessing the Impacts of Coal Waste—Wigginton et al. 157 Table 1. General overview of findings on concentrations for heavy metals by site: findings for those metals that produced notable results reported: arsenic, iron, lead, and manganese (ppb) Non-detects set to half detection limit. N Min Ql Mdn Q3 Max Mean Std Dev Std Err US EPA # Ex Arsenic (As) Site 1+ 24 2.50 2.50 2.50 2.50 14.5 342 2.81 0.57 10 1 Site 211 2.50 2.50 2.50 5.33 7.19 3.50 1.76 0.53 10 (0) Site 3° 8 2.50 2.50 4.25 8.97 1340 La 470 166 10 1 Site 3£ 7 2.50 2.50 2.50 7.10 9.60 4.67 2.91 1.09 10 0 Iron (Fe) Site 1 9A 95 957 1210 4220 17,600 3540 5100 121 300 17 Site 2 1 ees 53.9 4120 12,400 = 31,100 6590 9350 2820 300 8 Site 3 S~ 25 25 25 209 3000 367 1040 367 300 i| Lead (Pb) Site 1 94 0.50 0.50 2.61 992 527 B02 107 21.9 15 6 Site 2 11 0.5 1.15 1.55 9.59 29 5.83 8.87 267 15 1 Site 3 8 0.50 0.50 1.20 29.4 347 51.2 120 42.5 15 3 Manganese, (Mn) Site 1 94 025- 222 70 959 870 192 255 52 50 15 Site 2 ll 0.25 0.90 234 826 1300 366 934 133 50 5 Site 3 8 1.20 44 220 1290 30,500 4120 10,700 3770 50 v4 Sulfate (SO, 2>) Site lt 313 1.13 2.0 4.68 35.6 147 29.2 50.9 14.1 250 0) Site 2 9 2.06 10.0 28.6 86.6 132 46.6 46.3 15.4 250 ) Site 3° a 8.18 14.0 37.0 81.4 93.7 47.2 36.7 13.8 250 0 * Site 1: Martin County, KY. » Site 2: Mingo County, WV. * Site 3: Pike County, KY. * Site 3: Arsenic concentrations for Pike County with extreme case, unused well (1335.4 ppb) deleted from analysis. to find homes with wells more quickly than in Martin County. For further analysis data for all three counties was pooled. To explore our hypotheses, well data for each of the above heavy metals was also coded based on being above or below safe and secondary drinking water standards. These data were then compared for downstream wells located in an impoundment sub- watershed versus other wells in non impound- ment sub-watersheds. Standard nonparamet- ric cross tab and Chi Square tests of statistical significance were used to determine differ- ences in the proportion of wells below or above safe or secondary drinking water standards based on sub-watershed compari- sons. To confirm these findings, well data from the Kentucky Groundwater Database Repository (KGDR) were analyzed in a similar way. Wells from the same counties and from approximately the same time frame (2003-2007) were selected for analysis. To avoid artificial bias based on sampling fre- quency, if multiple samples were taken on a given day, one sample was selected at random and any others discarded. Unfortunately, finer level comparison of means tests and regres- sion analyses could not be performed on the data due to the extreme variability in well water findings and due to the relatively small case base. RESULTS AND DISCUSSION Arsenic (As). Two wells exceeded the 10 ppb US EPA drinking water guideline for arsenic (Table 1), one each in _ site 1 (max=14.5 ppb) and site 3 (max=1340 ppb). The extreme value for site 3 was a sample drawn from a well that is no longer in use (other wells on this property were within drinking water standards for As.) With this unused well dropped from the analysis, findings show one other well at site 1 (14.5 ppb) with As concentrations above US EPA 10 ppb safe standards. Samples taken from cold water taps were also analyzed and while results for most residences were within acceptable ranges (=1.85 ppb + SD =3.07), one location (Mingo County) had cold tap water with levels above safe drinking water 158 standards for arsenic (11.4 ppb). This may suggest the need for further study of well water for arsenic exposures within this partic- ular area, especially in light of extreme As results from site 3, which are discussed in more detail below. High levels of As in roundwater are common in Appalachia. Shiber (2005) found that 6% of tap water samples from residences with private wells had As levels in excess of 10 ppb. To confirm the extreme findings from the location in site 3, the well was sampled a second time six months later. Arsenic levels from this unused source still exceeded the guideline value, but at 12.2 ppb. During the second site visit, it was noted that the water level in the 55 m deep well had changed dramatically from nearly dry during the first sampling (October) to having a water table 21 m deep (May). Based on independent well water tests conducted during the 1980s provided by the home owner, water levels were relatively stable at approximately 8 m deep. Though arsenic concentrations are not recorded on these logs, since the 1980s, this area of Pike County has experienced heavy surface (Mountaintop removal) and continued deep mining which has likely altered the hydrology of the area greatly over the past 20 years. Moreover, Kentucky coals typically exceed mean arsenic concentrations for U.S. coals and this area of Pike County has recorded high As concentrations in stream sediment (58-110 ppm) (Tuttle et al. 2002). While other wells at site 3 were within safe drinking water standards, the reported excee- dances and human health effects associated with As exposure (Smith 1992, 1998; Gomez- Caminero et al. 2001; Kinniburgh and Kosmus 2002) suggests that further study may be warranted. Iron (Fe). Median iron values for site 1 (1210 ppb) and site 2 (4120 ppb) far exceeded US EPA secondary standards (300 ppb) (Ta- ble 1). In fact, 17 of 24 wells (71%) for site 1 and 8 of 11 wells (73%) for site 2 exceeded secondary Fe standards (300 ppb). In com- parison, only 1 of 8 wells was in exceedances (max=3000 ppb) at site 3. Maximum excee- dances at site 1 (17,600 ppb) and site 2 (31,000 ppb) ranged from 58 to over 100 times secondary drinking water standards. While all sites (1,2 and 3) are known for Journal of the Kentucky Academy of Science 69(2) above average iron concentrations in soils and sediments (Gustavsson et al. 2001), high Fe might not only be associated with area soils, but also pyrite (FeSg), a natural component of coal, which may also contribute to water bourn levels of Fe. Water from coal seams is sometimes high in Fe (Wunsch 1992). More- over, magnetite (Fe30,4), a major component of the coal cleaning process and coal waste impoundments, may also be a source contrib- utor. In the next set of analyses, the possible effects of coal waste on well water and iron levels is modeled through GIS analysis of sub watersheds. Lead (Pb). Mean concentrations for lead are higher in KY counties (Table 1, site 1 and 3) and were at least twice the level US EPA safe drinking water standards (US EPA = 15 ppb; site 1 X = 35 ppb; site 3 x = 51 ppb). Maximum exceedances across both sites (site 1 = 527 ppb and site 3 = 347 ppb) were well over 20 times safe drinking water standards set for lead. Due to these extraordinarily high exceedances, and the well-substantiated health effects of lead exposure on human health, cold water tap results were also examined. Analyses showed coldwater sam- ples to be within more or less acceptable ranges (x = 4.23, SD = 14.7). However, two coldwater (drinking water) samples reported lead levels above safe standards (25.3 ppb; 92.6 ppb respectively). Manganese (Mn). As with iron, elevated levels of Mn can be associated with coal seams and mine drainage, possibly because of sulfide mineral oxidation (O’Steen and Rauch 1983; Wunsch 1992). Some parts of Appalachia have Mn soil concentrations well above the national average (Gustavsson et al. 2001). Across all three sites (Table 1), more than 60% of tested domestic wells exceeded US EPA secondary standards set for manganese (50 ppb). Median concentrations were in exceedance and were nearly 8X secondary standards across two of the three sites (site 1, mdn = 70, site 2, mdn = 234; site 3, mdn = 220). Median values and exponentially higher mean concentrations (site 1 x = 870 ppb; site 2 x = 1300; site 3 x = 30,500) indicate manganese drinking water exposures of possible health concern as addressed by ATSDR in its 2005 site assess- ment of Mingo County (ATSDR 2004). In recent West Virginia hearings, health officials Assessing the Impacts of Coal Waste—Wigginton et al. Table 2. 159 Comparison of several heavy metal parameters below or above drinking water standards for reference wells in sub-watersheds with no coal waste impoundments (reference wells) versus wells in sub-watersheds with one or more coal waste impoundments (n = 42).* SDWS / DWS? Reference Iron (Fe) Below SDSW(%)" 300 ppb 15:54) Above SDWS(%) 3 (46) Lead (Pb) Below DWS (%) 15 ppb 21475) Above DWS (%) TAD) Manganese (Mn) Below SDWS (%) 50 ppb 11 (39) Above SDWS (%) 17 (61) Impoundment Sub-watershed 2 (14) X = 5.98, df= 1, sig = .015* 12 (86) 11 (79) XxX’? = .066, f = 1, sig = .798 3 (21) 5 (36) X = .05, df = 1, sig = .822 9 (64) * Site 1: Martin County, KY - Site 2: Mingo County, WV - Site 3: Pike County, KY. » SDWS = US EPA secondary drinking water standard; DWS = US EPA safe drinking water standard. * p =.05. have raised similar concerns regarding high Mn concentrations in exceedance of second- ary standards (WV Joint Judiciary Subcom- mittee Hearing 2006, testimony of toxicolo- gist, Dawn Seeburger). Manganese findings across all three active coal mining sites, not only site 2 (Mingo County), indicate Mn concentrations to be far in exceedance of secondary standards and therefore, coldwater (drinking water) samples were also checked for Mn levels. While median levels were within acceptable levels (11 ppb) for all three sites, the overall mean concentration was high suggesting wide variability in cold water results (x = 136 ppb + SD = 261 ppb). Thirty-six percent of cold water, drinking water samples were above 50 ppb guidelines; 5 cold tap samples were 10 to 20 times in exceedance of secondary standards (537 to 1245 ppb). Unfortunately, little work has been done on the possible health effects of such high levels of manganese in drinking water in Appalachia, although the ATSDR did recommend that certain groups avoid drinking groundwater with high levels of Mn (ATSDR 2004, 2005). Sulfate (SO,4?, Table 1). Wells across all three sites had average sulfate concentrations below US EPA drinking water secondary standards (250 ppb). Moreover, maximum values were also well below US EPA standards at each site (site 1, max = 147 ppb; site 2, max = 132 ppb; site 3, max = 94 ppb). It was predicted that wells in sub-watersheds with impoundments (Table 2) would have a higher proportion of wells above US EPA safe (DWS) or secondary (SDWS) drinking water standards than other reference wells that are outside impoundment sub-watersheds (see Figure 2). Yet, due to the nature of the statistical tests (cross tab and Chi Square tests), analyses could not be conducted on arsenic, or sulfate due to an insufficient number of cases above safe or secondary standards. Therefore, iron, lead and manganese are used as possible parameters, with concentrations collapsed into those above and below US EPA primary or secondary drinking water standards to compare across impoundment and non-impoundment sub-watersheds. Finally, only those residential wells expressly used for drinking water were used in this analysis (n = 42). Findings are first reported for iron (Ta- ble 2). It was expected that iron might be more prevalent in wells in sub-watersheds with coal waste impoundments due to iron being a possible by-product of coal waste because various iron compounds are present in coal naturally (e.g., pyrite, FeS2) and due to the coal preparation process (Magnetite, Fe30,). Findings (Table 2) suggest some support for this proposition insofar as a higher proportion of wells (86%) in coal waste impoundment sub watersheds had iron con- centrations exceeding US EPA secondary standards (300 ppb). In comparison, a lower proportion (46%) of wells outside impound- ment sub-watersheds reported iron concen- trations above secondary standards. This difference was statistically significant (X? = 5.98, df = 1, sig = 0.015) suggesting that wells near coal waste impoundments are more likely to experience poor drinking water quality because of iron exceedances. This conclusion is supported by the fact that well data from 160 the Kentucky Groundwater Database Repos- itory show a similar pattern with 84% of wells in sub-watersheds with impoundments having iron levels above the US EPA secondary standard. In sub-watersheds with no im- poundments, 66% of wells exceeded the same limits. Findings are next reported for lead and manganese (Table 2): Here too it was pre- dicted that lead and manganese might be more prevalent in wells in sub-watersheds with coal waste impoundments due to both being a notable parameter of coal waste in past regulatory assessments. However, on these two parameters, findings suggest the need for further exploration and study. Sub- watershed comparisons do not reveal either parameter to be a significant predictor of possible coal waste effects. For lead, a similar proportion of wells in impoundment sub- watersheds (21%) in comparison to other wells (25%) reported levels above safe drink- ing water standards and thus, there was no statistically significant difference between wells near impoundments and other wells with respect to lead levels above safe drinking water standards (X* = 0.066, df = 1, sig = 0.798). Similar non significant findings (X? = 0.05, df = 1, sig = 0.822) are reported for manganese. Findings show a similarly high proportion of wells in both impoundment (64%) and non impoundment (61%) sub- watersheds reporting manganese levels above secondary drinking water standards. Yet, interestingly, when well data from the Ken- tucky Groundwater Database Repository were compared for manganese, 90% of wells from sub-watersheds with coal waste impound- ments were above the US EPA secondary drinking water limit while 69% of sub- watersheds lacking impoundments had high manganese levels. The rock found in conjunction with coal, often called over burden, if it must be removed in the course of coal mining, may be a source of heavy metal in addition to the coal itself. The shale and sandstone of Appalachia have higher concentrations of Cd; Cu, Co; Cr Fe: Ky Mn. Mo, Ni“ Pbs S. Ti, and Zn than the national average for these rock types (Bogner and Sobek 1979). The occurrence of various metal sulfides is very common in Appalachian rock. The greater the Journal of the Kentucky Academy of Science 69(2) availability of oxygen to these sulfides in a wet environment, the greater their solubility often is (Morgan et al. 1992). Settling pond materials are especially susceptible to disso- lution because they are fine grained and homogeneous (Bogner and Sobek 1979). Additionally, metal sulfide particles can accu- mulate in the lighter waste portion of the coal cleaning process (Finkelman 1979). Given that much rock in the Appalachian region is higher than average in various heavy metals, it is possible that the higher levels of Mn and Pb found in the present study were not related to coal mining and simply reflect the influence of the rock through which the water passed. However, coal mining creates conditions that optimize for the mobilization of metals, either from coal, coal waste, or overburden, by crushing these materials into smaller particles, thus increasing surface area and access to oxygen. Bienkowski indicated that wells with their bottoms above the level of the local stream had lower pH and alkalinity, and higher concentrations of Mn, Fe, sulfate, and total dissolved solids, due largely to increased oxygen availability (1990). The same author indicated that the source for the Fe and sulfate in such cases was pyrite from coal (Bienkowski 1990). Especially interesting is the trend of increased Mn levels in wells from the Kentucky Groundwater Database Repos- itory in watersheds with coal waste ponds. Manganese is often highly mobile and can correlate with Fe levels in ground water, so it too may be useful as an indicator of mining activity (Bogner and Sobek 1979; Bienkowski 1990). These factors indicate that while this study cannot offer conclusive evidence about the role of coal mining in the high levels of some metals found in Appalachian water, mining activities should be considered an important candidate for their source. A more rigorous study covering a larger number of wells would be needed to confirm these possibilities as well as offer insight into less commonly encountered metals such as Pb. SUMMARY In its report after the 2000 coal sludge event, the National Research Council stated, as one of its recommendations, the need to characterize the constituents of coal slurry to better assist in monitoring and regulating coal Assessing the Impacts of Coal Waste—Wigginton et al. waste and coal waste impoundments (2002). In our efforts to start to characterize coal slurry constituents, metals were initially se- lected for analysis based on their prevalence in early US EPA and USGS coal slurry reports (US EPA 1985; USGS 1989). Our original research question was to identify possible markers associated with coal waste by com- paring well water in known coal waste impoundment sub-watersheds_ with other wells in our study area of the Big Sandy Watershed. Our findings did show significant- ly more wells in impoundment sub-water- sheds with iron levels above secondary drink- ing water standards. This significant pattern may warrant further study of iron as a possible marker of coal slurry waste because various iron compounds are present in coal naturally (e.g., pyrite, FeSs) and magnetite (Fe304) is a major component of the coal cleaning and preparation process and a known waste by- product of coal slurry. Well data from the Kentucky Groundwater Database Repository (KGDR) showed the same overall pattern and also showed a similar pattern for manganese that our data did not. Further work with manganese may be justified as well because of conflicting findings. Yet, irrespective of significant differences between sub-watersheds, overall high concen- trations of iron, as well as manganese and lead and sometimes arsenic suggests that residen- tial well water in the Big Sandy region of Appalachia may sometimes be of unhealthy quality. This, in itself, warrants further study. Therefore, efforts should be made to more systematically investigate, model and compare domestic well water quality in areas of the Big Sandy less impacted by mining with more impacted areas. To this end, more collabora- tion between states as well as research and citizen groups in data sharing is warranted. Already, there has been some productive discussion across borders between citizen groups and research teams to share data and GIS techniques. But more formal collabora- tions and working partnerships are required to develop more systematic comparisons and controls in order to isolate and test the effects coal slurry waste and other coal mine activities on well water. More systematic data collec- tion, comparisons and controls will allow us to better determine whether or not poor well 161 water quality is associated with active coal mining, an association that many in the region assert to be true, but that has so far been difficult to “prove.” Systematic analysis through collaborative development of geo- coding and GIS methods, such as_ the additional overlay of deep-mine injection points and surface and underground mine activities with well water data, will allow us to more objectively determine the association between coal mine activities and possible impacts on private drinking water within the Big Sandy Region. Future research would be greatly facilitated by interested univer- sity or private sector scientists or interested citizen groups seeking a collaborative part- nership with any of several state agencies, such as the Kentucky Geological Survey, which maintains the Kentucky Groundwater Database Repository, the Kentucky Division of Water, or the University of Kentucky’s Kentucky Water Resources Research Insti- tute. This is in accord with emerging col- laborative decision-making models that are designed to monitor and protect water re- sources. LITERATURE CITED ATSDR (Agency of Toxic Substances and Disease Registry). 2001. Record of Activity. Name: Martin County Coal Slurry, February 16. ERS Log #01-2117. Atlanta, Georgia. ATSDR (Agency of Toxic Substances and Disease Registry). 2004. Health Consultation. Private Well Water Quality. Williamson, WV Sites (a/ka/ Williamson Area). February 13. Prepared by: West Virginia Department of Health and Human Resources, Bureau of Public Health under a cooperative agreement with the Agency for Toxic Substance and Disease Registry, Division of Health Assessment and Consultation, Atlanta, Georgia. ATSDR (Agency of Toxic Substances and Disease Registry). 2005. Health Consultation. Private Well Water Quality. Williamson, WV Sites (a/ka/ Williamson Area). April 6. Prepared by: West Virginia Department of Health and Human Resources, Bureau of Public Health under a cooperative agreement with the Agency for Toxic Substance and Disease Registry, Division of Health Assessment and Consultation, Atlanta, Georgia. APHA. 1995. Standard Methods for the Examination of Water and Wastewater. 19th ed. American Water Works Association and Water Pollution Control Fed- eration, Washington, DC. Batty, L. C., L. Atkin, and D. A. C. Manning. 2005. Assessment of the ecological potential of mine-water 162 treatment wetlands using a baseline survey of macro- invertebrate communities. Environmental Pollution 138:412-419. Bienkowski, L. S. 1990. Delineation and characterization of the aquifers of the Eastern Kentcky Coal Field. Ph.D. Dissertation. University of Kentucky. Bogner, J. E., and A. A. Sobek. 1979. Distribution of selected metals in classic overburden units of the Appalachian and interior coal basins: Water quality implications. Compte Rendu 4:479-490. Booth, C. A., D. A. Spears, P. Krause, and A. G. Cox. 1999. The determination of low level trace elements in coals by laser ablation-inductively coupled plasma-mass spectrometry. Fuel 78:1665—1670. Brady, K. 1998. Groundwater chemistry from previously mined area as a mine drainage quality prediction tool. Pages 9:19:21 in K. Brady, M. Smith and J. Schueck (eds). Coal mine drainage prediction and _ pollution prevention in Pennsylvania, 5600-BK-DEP2256 8/98. Pennsylvania Department of Environmental Protection, Harrisburg, Pa. Finkelman, R. B. 1979. Mode of occurrence of accessory sulfide and selenide minerals in coal. Compte Rendu 4:407-412. Gomez-Caminero, A., P. Howe, M. Hughes, E. Kenyon, D. R. Lewis, J. Ng Moore, A. Aitio, and G. Besking. 2001. Arsenic in drinking water. World Health Orga- nization, Geneva. Grey, N. F. 1998. Acid mine drainage composition and the implications for its impact on lotic streams. Water Research 32:2122-2134. Gustavsson, N., B. Bglviken, D. B. Smith, and R. C. Severson. 2001. Geochemical landscapes of the conter- minous United States- New Map Presentations for 22 Elements. U.S. Geological Survey Professional Paper 1648. U.S. Department of the Interior. Haman, D. Z., and D. B. Bottcher. 1986. Home water quality and safety. Circular 703, Florida Cooperative Extension Service, Institute of Food and Agricultural Sciences, University of Florida. HNN Staff. 2007. Senate, House agree to study coal slurry and groundwater contamination: Sludge Safety Project wants independent verification of study. The Hunting- ton News, March 9. Huggins, F. E. 2002. Overview of analytical methods for inorganic constituents in coal. International Journal of Coal Geology 50:169-214. (KDOW) Kentucky Division of Water. 2000. Martin County Coal, Well-Water Testing Reports: DOW File: 00548 10-680-8002. Kinniburgh, D. G., and W. Kosmus. 2002. Arsenic contamination in groundwater: Some analytical consid- erations. Talanta 58:165-180. LaSage, D., and M. J. Caddell. 2006. Chemistry in bottom sediment of Crum Reservoir, Martin County, Eastern Kentucky compared to a reference reservoir in central Kentucky. http:/www.anthropology.eku.edu/martincounty/ PDF/res_study.pdf. Journal of the Kentucky Academy of Science 69(2) McSpirit, S., S. Scott, D. Gill, S. Hardesty, and D. Sims. 2007. Risk perceptions after a coal waste impoundment failure: A survey assessment. Southern Rural Sociology 22:88—110. McSpirit, S., S. Scott, S$. Hardesty, and R. Welch. 2005. EPA actions in post disaster Martin County, Kentucky: An analysis of bureaucratic slippage and agency recreancy. Journal of Appalachian Studies 11:30-58. McSpirit, S., A. Wigginton, and D. Sims. 2006. Assess- ment of finished water, the public water system. Martin County, KY. McSpirit, S., S. Scott, D. Gill, S. Hardesty, and D. Sims. 2007. Risk perceptions after a coal waste impoundment failure: A survey assessment. Southern Rural Sociology 22:88-110. Morgan, J. R., R. W. Holman, and D. J. Wilson. 1992. Oxidation of heavy metal sulfides in relations to the environment. Journal of the Tennessee Academy of Science 67:90-92. NRC (National Research Council), Committee on Coal Waste Impoundments. 2002. Coal waste impound- ments: Risks, responses and alternatives. National Academy Press, Washington, DC. O’Steen, W., and H. W. Rauch. 1983. Effects of coal mining on ground-water quality in West Virginia. Pages 355-363 in 1983 Symposium of Surface Mining, Hydrology, Sedimentology and Reclamation, Nov 27- Dec 2. University of Kentucky, Lexington, KY. Scott, S., S. McSpirit, S. Hardesty, and R. Welch. 2005. Post disaster interviews with Martin County citizens: “Gray clouds” of blame and distrust. Journal of Appalachian Studies 11:7—29. Shiber, J. G. 2005. Arsenic in domestic wellwater and health in central Appalachia, USA. Water, Air, and Soil Pollution 160:327-341. Smith, A., C. Hopenhayn-Rich, M. Bates, H. Goeden, I. Hertz-Picciotto, H. Duggan, R. Wood, and M. Kosnett. 1992. Cancer risks from arsenic in drinking water. Environmental Health Perspectives 97:259-267. Smith, A., M. Goycolea, R. Haque, and M. L. Biggs. 1998. Marked increase in bladder and lung cancer mor- tality in a region of Northern Chile due to arsenic in drinking water. American Journal of Epidemiology 147:660-669. Stout, B. M., and J. Papillo. 2004. Well water quality in the vicinity of a coal slurry impoundment near William- son, West Virginia. Coal Impoundment Project, Wheel- ing Jesuit University, Wheeling, WV. Tiwary, R. K. 2001. Environmental impact of coal mining on water regime and its management. Water, Air, and Soil Pollution 132:185-199. Tuttle, M., M. Goldhaber, L. Ruppert, and J. Hower. 2002. Arsenic in Rocks and Steams Sediments of the Central Appalachian Basin, Kentucky. U.S. Geological Survey Open-File Report 02-0028. US DOE (United States Department of Energy). Office of Coal, Nuclear, Electric and Alternate Fuels. 2006. Annual Coal Report 2005. Washington, DC. Assessing the Impacts of Coal Waste—Wigginton et al. US DOI (United States Department of Interior), Office of Surface Mining. 2002. Report on October 2000 breakthrough at the Big Branch Slurry Impoundment. Washington, DC. US EPA. 1985. Eastern Coal Corporation. Docket No. IV-85-UIC-101 (Proceeding under — Section 7003 of the Solid Waste Disposal Act 42 U.S.C. § 6973). US EPA. 1993. Methods for the determination of inorganic substances in environmental samples. Meth- od 300.0 EPA 600/R-93-100. Office of Research and Development. Washington DC. US EPA. 2005. Test methods for evaluating solid wastes, 3rd edition, final update 3. SW-846. Office of Solid Waste, Washington D.C. USGS (United States Geological Survey). 1989. Summary of the U.S. Geological Survey and U.S. Bureau of Land Management National Coal Hydrology Program, 1974— 1984. L. Britton, C. Anderson, D. Goolsby and B. Van 163 Havern (eds). U.S. Government Accounting Office, Washington, D.C. Wagner, N. J., and B. Hlatshwayo. 2005. The occurrence of potentially hazardous trace elements in five Highveld coals, South Africa. International Journal of Coal Geology 63:228—246. Wigginton, A., S. McSpirit, and D. Sims. 2007. Heavy metal accumulation in hot water tanks in a region experiencing coal waste pollution and comparison between regional water systems. Bulletin of Environ- mental Contamination and Toxicology 79:405—409. Wunsch, D. R. 1992. Ground-water geochemistry and its relationship to the flow system at an unmined site in the Eastern Kentucky Coal Field. Dissertation. University of Kentucky. WV Joint Judiciary Subcommittee. 2006. Hearing on underground injection of coal slurry and groundwater contamination before WV Joint Judiciary Subcommit- tee, WV Legislature, October 16 Charleston, WV. J. Ky. Acad. Sci. 69(2):164-169. 2008. Building a Centralized Database for Kentucky Fishes: Progress and Future Applications Robert L. Hopkins II,’ Michael D. Burns, Brooks M. Burr, and Lisa J. Hopman Department of Zoology, Southern Illinois University, Carbondale, Illinois 62901 ABSTRACT Kentucky has a long history of ichthyological collection and study. Voucher specimens are available for collections dating back to 1870 and are housed in various museums and research collections in the United States. In this paper we present a descriptive overview and the current status of a project aimed at building a centralized database for vouchered records of Kentucky fishes. To date, we have entered over 51,000 records into the database. Each of these entries contains at least three vital pieces of information: species identification, georeferenced locality, and time of collection. With increasing recognition of the potential value of museum-based data in biological, ecological and conservation studies such a centralized database will serve as an important scientific resource for the study of Kentucky fishes. KEY WORDS: fishes, database, Kentucky, biodiversity, historical, museum INTRODUCTION Ichthyological investigations in Kentucky date back to the earliest scientific documen- tation of the state’s fish fauna in the “Ichthyo- logia Ohiensis” (Rafinesque 1820). In_ this work, Rafinesque provided natural history and descriptive information for fishes mostly from the Ohio River near the falls at Louisville. Only a small number of collections are known prior to 1820 and include anecdotal accounts of larger species, e.g., from the Ohio River (Pearson and Krumholz 1984). Woolman (1892) conducted a more comprehensive survey of Kentucky fishes and documented statewide historical distributional information prior to widespread anthropogenic influence. More recent literature on Kentucky fishes includes a distributional catalogue (Evermann 1918), state fish book (Clay 1975), a species checklist (Burr 1980), an Ohio River status and distribution update (Pearson and Krum- holz 1984) and a statewide distributional atlas (Burr and Warren 1986). The Kentucky State Nature Preserves Commission, the Kentucky Division of Water, and the Kentucky Depart- ment of Fish and Wildlife Resources regularly conduct surveys and produce reports which contribute to the understanding of fish distributions and the status of streams and rivers in the state. Other notable contributions were made by Minor E. Clark, William R. Turner, and those associated with the Ohio ' Corresponding author e-mail: rhopkins@siu.edu River Valley Water Sanitation Committee who provided detailed descriptions of watersheds in the state, pre-impoundment studies, and detailed documentation of the Ohio River ichthyofauna, respectively. The most comprehensive assessment of Kentucky fish distributions was provided by Burr and Warren (1986). They presented a detailed discussion of physiographic and hydrographic features in Kentucky, species accounts that included point locality range maps, narrative descriptions of distribution, systematics, habitat, and a recommended conservation status. The distributional data in the atlas span from 1818 to 1985 and include records from all major museum holdings in the United States as well as records from personal collecting efforts by the authors. The range maps were fashioned by hand and no computerized records were created during the project. Burr and Warren (1986) is still the principle reference for Kentucky fishes. However, taxonomic revi- sions, descriptions of new species, and nu- merous new collection records since 1986 warrant updating of the distributional data. In recent years there has been increasing interest in museum-based informatics (Gra- ham et al. 2004). Natural history collections offer a wealth of temporal and geographic data tied to animal and plant collections. Develop- ments in computing and analytical tools, especially geospatial technologies, are making these data more relevant and useful. As a result, many of the records used in producing 164 A Database for Kentucky Fishes—Hopkins et al. the distributional atlas (Burr and Warren 1986) now have been electronically cataloged. These data are available by written request to collection managers or, in some cases, via searchable internet databases. Efforts are underway by various museums and institu- tions to digitize and link databases online — allowing users to perform a single query across multiple databases. One such project for fishes is Fishnet 2 (www. fishnet2.net). This website allows a user to search fish records across 29 institutional databases with a single query, and results can be exported in a variety of formats for subsequent processing. How- ever, this system has limitations: 1) only a relatively small percentage of the data are georeferenced, 2) there is no consistent schemata to the exported data, and_ 3) taxonomy rarely is updated for historical records. In light of the current usage and potential utility of museum-based information for the study, conservation, and management of Kentucky fishes, we decided to build a centralized database for the state. Because many conservation plans and research projects are implemented within state boundaries, such a database would provide historical and contemporary data that would be readily available for analyses. The database also would include information linking records with museums containing voucher specimen(s). The intention of this project is to enhance research and conservation efforts focused on Kentucky’s ichthyofauna. The objectives of this multi-year, ongoing project are to 1) obtain records of Kentucky fish collections from all major museum and institutional holdings, 2) computerize and georeference those records and merge the data into a single dataset with a standardized information schemata, 3) complete error checking and data cleaning, and 4) update taxonomic and nomenclatural information. We provide an overview of our progress and the current status of the database. MATERIALS AND METHODS In 2005, Southern Illinois University at Carbondale (SIUC) acquired the ichthyology collection from the University of Louisville (UL). A major task was undertaken to curate, georeference, and computer catalogue the 165 estimated 13,000 lots from UL. In addition to the UL project, we also began curating, georeferencing, and computerizing —back- logged records and collections from SIUC, Kentucky State Nature Preserves Commis- sion, Kentucky Department of Fish and Wildlife Resources, and the Kentucky Divi- sion of Water. This initiative provided the impetus to expand our efforts to create a comprehensive and centralized database of Kentucky fish records. This database is hereafter referred to as the Kentucky Fishes Database (KFD). We estimate that greater than 90% of the vouchered collection records for Kentucky fishes are located in the following museums: California Academy of Sciences (CAS), Cor- nell University (CU), Illinois Natural History Survey (INHS), University of Kansas (KU), National Museum of Natural History (USNM), Ohio State Museum of Biological Diversity (OSM), Southern Illinois University Carbondale (SIUC), University of Michigan Museum of Zoology (UMMZ), and Tulane University (TU). We requested or download- ed all Kentucky fish records from these museums. Many other smaller collections also contain valuable data including Morehead State University (MoSU) and Eastern Ken- tucky University (EKU). We acquired paper copies of select records from MoSU and incorporated them into the database. The research collection at Murray State University had earlier been transferred and catalogued into the SIUC system. Records received from museums were in various processed states. For example, some records were fully georeferenced and needed only to be converted to decimal degree format and fitted to the KFD schemata — whereas others were not georeferenced and contained numerous taxonomic and geographical errors. Most of the datasets, however, contained good descriptive information on localities but were not georeferenced and had out-dated taxono- my. For each record processed, we georefer- enced the site in decimal degree format using the North American Datum of 1983, updated taxonomic and nomenclatural information, and standardized the data schemata. Formats for data and headers were specifically chosen for quick incorporation into a geographical 166 Table 1. Contributing museums and respective number of records provided. Records from SIUC include the ~13,000 lots incorporated from the University of Louis- ville. Contributed Museum Vouchered Records California Academy of Sciences 246 Cornell University 1051 Illinois Natural History Survey 4100 University of Kansas 598 Morehead State University 1261 National Museum of Natural History 908 Ohio Museum of Biological Diversity 2108 Southern Illinois University Carbondale 36,724 Tulane University 875 University of Michigan Museum of Zoology 3342 TOTAL 51,213 information system (GIS), which played a key role in data editing and enrichment. Ultimate- ly, each record contains three vital dimen- sions: specimen identification, georeferenced collection locality, and time of collection. All records in the KFD are vouchered records. The incorporation of only vouchered records was intentional and assures that each entry is backed by a physical specimen; this allows the verification of species identification, use in research projects, and increases data integrity. RESULTS AND DISCUSSION In 2005, the SIUC collection contained about 10,000 computerized records for Ken- tucky. As of early 2008, after the incorporation of the UL collection and processing of backlogged collections from state agencies, that number exceeded 36,000. At the end of 2008, SIUC will have more than 40,000 records of Kentucky fishes dating from the 1920s. We have voucher records for every extant species in Kentucky. SIUC’s records — including those incorporated from UL - account for about 70% of the records in the KFD. As of early 2008, the KFD contained over 51,000 records (Table 1). Excluding SIUC, the largest contributors of Kentucky records were INHS and UMMZ with 4100 and 3100 contributed records, respectively. The CAS and the USNM contained fewer records, but housed the majority of the A.J. Woolman and P.H. Kirsch collections, which are of historical value. Ultimately, we expect the KFD to Journal of the Kentucky Academy of Science 69(2) exceed 60,000 records. We are currently processing additional records from the Uni- versity of Florida, University of Tennessee, Academy of Natural Sciences of Philadelphia, University of Alabama, Field Museum of Natural History, and the Harvard Museum of Comparative Zoology. In addition, we continue to process records from state agen- cies in Kentucky and from our own collection efforts. The KFD includes records dating from the 1870s. Woolman’s collections represent the earliest comprehensive surveys included in the database. Fewer than 20 pre-Woolman records, from unnamed collectors, exist for the entire dataset. The spatiotemporal cover- age of the KFD (Figure 1) is statewide for selected time periods but most comprehensive for the 1970-1990 decades. The majority of vouchered specimens were added after 1950 with most specimens collected since 1980. Despite over 100 years of collection efforts, there are still some regions of the state lackin adequate coverage. Historically, the Nolin River drainage has lacked basic survey work. Likewise, additional collection efforts in the lower Green River, Tug Fork, Mississippi River, and direct Ohio River tributaries would permit a more complete documentation of Kentucky’s ichthyofauna. A consistent data schemata is being applied to all KFD records (Table 2). We encourage additional data enriching by incorporation of comments from collector’s field notes, pub- lished papers, and reports when available. Such ancillary information (e.g., general comments on habitat, unusual observations, collecting methods/techniques, information on species not vouchered, designation as type specimens, etc.) can greatly enrich the point data. To date, we have georeferenced all records acquired (Table 1), but have not completed the totality of editing and enriching the data. Updating taxonomy from historical collections has proven to be time consuming, and we have made only moderate progress. For newer collections and especially those housed at SIUC we have completed most of the taxonomic updating. Nomenclatural and taxonomic issues are being addressed using several methods. For species with contrasting biogeographies we are able to clean the data within a spatial A Database for Kentucky Fishes—Hopkins et al. Pre-1950 Collections 1970 - 1990 Collections Figure 1. format. For example, the Etheostoma specta- bile (Agassiz), E. simoterum (Cope), E. virga- tum (Jordan), and Notropis rubellus (Agassiz) complexes, each containing species with allopatric ranges are edited by plotting the target data within a GIS and assigning correct names to points using spatial-selection meth- ods. In contrast, much editing of older names is being done in a standard database format. Despite extensive cleaning, we continue to find errors within the database. Based on past experiences, we estimate that perhaps 5-7% of the entries contain errors of some nature (e.g., out-dated taxonomy, misspellings, etc.) with 1-2% containing errors that seriously compromise the integrity of the data (e.g., Table 2. Data schemata for database with examples of included information. Field Name Information Museum SIUC CatNum 64,942 Species Etheostoma derivativum Latitude 36.8822 Longitude —87.1111 Waterbody Whippoorwill Creek Drainage Red-Cumberland River County Todd State Kentucky Day 19 Month April Year 2007 Collectors = RL Hopkins, LJ Fisk Locality at SR 106 bridge; 0.5 mi S of Claymour, KY Notes riffle area now inundated due to fallen tree 167 1950 - 1970 Collections Post-1990 Collections Overview by time periods of collection records currently contained in the Kentucky Fish Database. incorrect locality information or misidentifi- cation). Focus on the three vital dimensions (identification, georeferenced localities, and time of collection) for each record is required to ensure high quality datasets. The potential applications of the KFD data are multifold. Museum-based data can be integrated with other data types to address a series of questions ranging from conservation research to the study of ecological and evolutionary processes. Other applications include spatial analysis of biodiversity, the study and modeling of species distributions, analysis of range shifts, planning and devel- oping future field projects, identification of potential areas for reintroductions, etc. To illustrate the utility of the KFD, we have completed a cursory spatiotemporal analysis of the distribution of the Trout-Perch (Per- copsis omiscomaycus Walbaum) in Kentucky (Figure 2). Some localities for this species included in Burr and Warren (1986) are not shown on the map (Figure 2) because the KFD is restricted to vouchered records. In this case, the voucher for a collection of Trout-Perch Welter (1938) in North Fork Triplett Creek (Licking River drainage) re- portedly deposited at MoSU could not be located. The overall pattern that emerges (Figure 2) is the absence of Trout-Perch in recent collections from the western portions of its range. In contrast, eastern populations in Little Sandy River, Levisa Fork, Tygarts 168 Collection Interval © Pre-1970 fF] 1971-1990 @ 1991 - present Figure 2. Creek, and Kinniconick Creek seem stable, although, no recent specimens have been captured from the East Fork of the Little Sandy River or Little Blaine Creek (Big Sandy River drainage). The apparent disappearance of this species in western areas warrants further research investigating the causes of its range contraction in Kentucky. CONCLUSIONS Since building the database we have been able to respond to questions about the Burr and Warren maps much more effectively. We are not only able to provide locality informa- tion, but also the time of collection, holding museum, a list of other species captured at the site, etc. Upon request, we can produce maps and spreadsheets containing the aforemen- tioned information. It is evident that such a database is a valuable tool for researchers and will increase the use of museum and research collections in ichthyological, ecological, and conservation studies. Furthermore, as we begin to mine the data we are finding many noteworthy records for Kentucky fishes. We have evidence of range expansions and contractions and of new drainage records for several species. The collection from UL has provided a wealth of additional historical information for Kentucky fishes. In addition, there have been multiple Journal of the Kentucky Academy of Science 69(2) N 200 Kilometers A be tt tt tH HI Spatiotemporal map of Trout-Perch (Percopsis omiscomaycus) records in Kentucky. recent collections of rare fishes — an encour- aging sign. We are preparing a manuscript providing an overview of some of the more noteworthy records stemming from the KFD project. We encourage museums of all sizes to georeference and computerize their fish collection records, so that they be more accessible for use in conservation planning and the study of evolutionary and environ- mental processes. ACKNOWLEDGEMENTS We are indebted to the following museums and their collection managers for making these data available to us for processing: CAS, CU, INHS, KU, MoSU, USNM, OSM, TU, UL, and UMMZ. The following people devoted much time and energy into curation, georeferencing, and computerization of the records: K. Butera, A. L. DeFore, D. J. Eisenhour, S. L. Flood, R. R. Hopkins, and A. Richter. Partial support for this project was provided by the Kentucky Department of Fish and Wildlife Resources. LITERATURE CITED Burr, B. M. 1980. A distributional checklist of the fishes of Kentucky. Brimleyana 3:53-84. Burr, B. M., and M. L. Warren, Jr. 1986. A distributional atlas of Kentucky fishes. Kentucky Nature Preserves Commission Scientific and Technical Series 4:1—398. A Database for Kentucky Fishes—Hopkins et al. 169 Clay, W. M. 1975. The fishes of Kentucky. Kentucky De- Pearson, W. D., and L. A. Krumholz. 1984. Distribution partment of Fish and Wildlife Resources, Frankfort, KY. and status of Ohio River fishes. ORNL/sub/79-783 1/1, Evermann, B. W. 1918. The fishes of Kentucky and Oak Ridge National Laboratory, Oak Ridge, TN. Tennessee: a distributional catalogue of the known Rafinesque, C. S. 1820. Ichthyologia ohiensis. W.G. Hunt, species. Bulletin of the Bureau of Fisheries 35:295-368. Lexington, KY. Graham, C. H., S. Ferrier, F. Huettman, C. Moritz, and — Welter, W. A. 1938. A list of the fishes of the Licking River A. Townsend-Peterson. 2004. New developments in drainage in eastern Kentucky. Copeia 1938:64-68. museum-based informatics and applications in biodi- | Woolman, A. J. 1892. Report of an examination of the versity analysis. Trends in Ecology and Evolution rivers of Kentucky, with lists of the fishes obtained. 19:497-503. Bulletin U.S. Fish Commission 10:249-288. J. Ky. Acad. Sci. 69(2):170-177. 2008. Mossbauer Study of the Half-Metallic Ferromagnet Fe,.,Co,Si Patrick Mono and Amer S. Lahamer'! Physics Department, Berea College, Berea, Kentucky 40404 ABSTRACT Mossbauer spectroscopy and X-ray diffraction measurements were performed to probe the magnetic properties of Fe,.,Co,Si (0 = x = 0.8) compounds at room temperature. X-ray diffraction measurements showed that the crystal structures of the synthesized samples existed as a single phase for all concentrations. Fittings of the Méssbauer spectra revealed that the family of Fe;_.,.Co,Si are weakly ferromagnetic for 0.1 < x < 0.5 and completely lose their magnetic properties when x = 0.5. KEY WORDS: Mossbauer, isomer shift, quadrupole splitting, half-metal, ferromagnets INTRODUCTION In 1983, de Groot et al. discovered that half-metallic ferromagnets existed when only charge carriers of one spin direction contrib- uted to the conduction band (de Groot et al. 1983). This implies that there is a band gap at the Fermi level for one spin direction. Following their discovery, there has been enormous interest in half-metals, as they form the basis of the new field appropriately termed spintronics. The bands calculation of the compound Feo.sCoo 9Si clearly shows the Fermi level in a gap of the minority band (Figure 1) (Guevara et al. 2004). These compounds had been studied even before the discovery of half-metalicity (Wertheim et al. 1966; Beille et al. 1981). The Fe ,.,Co,Si alloys form disordered solid solutions of cubic structure B20 at all concentrations (Figure 2) (Racu et al. 2007; Fanciulli et al. 1996). The Fe,.,Co,Si com- pound is remarkable in that when x = 1, CoSi is diamagnetic and when x = 0, FeSi is paramagnetic; however, there is a concentra- tion range of x where the Fe,_,Co,Si alloys are magnetically ordered. The appearance of the ferromagnetic order is due to the interaction between the moments of the cobalt and iron atoms. The range over which these alloys are supposed to be ferromagnetic is not consistent in the literature. It has been reported that the ferromagnetic range is 0.05 < x < 0.95 (Fanciulli et al. 1996), and when x = 0.5, the alloys were found to be non-magnetic (Wertheim et al. 1966). Furthermore, for x > 0.25, it is reported that the system starts to ' Corresponding author e-mail: lahamera@berea.edu segregate Co from Fe atoms leading to the disappearance of half metallicity and to a decrease of the magnetic moment (Guevara et al. 2004). To determine the range over which these Fe ,.,Co,Si alloys are magnetically ordered, alloys were synthesized in the range of 0 = x < 0.8. X-ray diffraction measurements were used to determine the crystal structure of the alloys and Mossbauer spectroscopy measure- ments were employed to search for the appearance or disappearance of magnetic order in these alloys. METHODS The alloys were prepared using the solid state method. Twenty millimoles of stoichio- metric proportions of pure Fe, Co, and Si were mixed, pulverized and cooked at 1100°C for 15 hr under the flow of argon. The samples were allowed to slow cool after which they were pulverized and X-ray diffraction measurements were taken at room tempera- ture. A Phillips X’Pert diffractrometer was used to collect the X-ray crystal structure data. Mossbauer absorption measurements were taken in excess of 1,000,000 counts with about 5 mCi of *Co-embedded in a rhodium matrix source. The data were taken in the sinusoidal mode and fitted with the commercially available Recoil fitting routine. The Moéss- bauer system (MVT 1000) used to collect data was made by the German Wessel Company. RESULTS AND DISCUSSION The X-ray diffraction patterns of all of the Fe,.,Co,Si (0 = x = 0.8) samples showed a single phase and are iso-structural with the 170 Mossbauer Study Fe ;.,Co,Si—Mono and Lahamer ' ® : Pe ee eel a ee ee ee Oe eT eR ee ee ee ee ee -2 -| 171 ie Gas aus cea eee ea — polis geese - Feo gCo 0.2Si 1 2 E (eV) Figure 1. parent compounds FeSi and CoSi (Figure 3). The Fe local symmetry is trigonal, with one Si nearest neighbor at 0.229 nm, three Si second nearest neighbors at 0.236 nm and three Si third nearest neighbors at 0.253 nm (Fig- ure 2) (Fanciulli et al. 1996). As mentioned earlier, these alloys possessed the B20 cubic structure which allowed for the existence of nuclear electric quadrupole and anisotropic magnetic dipole interactions. These effects were manifested in the resulted Mossbauer spectra of all the samples by showing a quadrupole splitting doublet (Figure 4). The Mossbauer results clearly showed that Feo 9- Density of states (arbitrary units) of Feo sCoo Si and FeSi. Cop Si and Feo sCoo25i were magnetic and had hyperfine magnetic fields (hmf) of 18.7 and 31.5 Tesla respectively which were referenced to metal iron with an hmf of 33 Tesla. The samples Feps5Coo5Si and Feo »9- Coo.sSi showed no magnetic splitting (Fig- ure 4). All of the data were fitted with only one doublet and one sextet for x = 0.1 and x = 0.2 samples as they showed magnetic splitting (Table 1). The data could be fitted with two sets of doublets, representing the different sites that the iron occupies, however, these samples were polycrystalline (powders), and 172 Journal of the Kentucky Academy of Science 69(2) Figure 2. The B20 structure of the e-FeSi compound. The numbers 1, 2, 3, and 4 denote the atoms at different positions. any differences are not resolved in the Mossbauer spectra hence only averages are reported. The resulted isomer shift values were constant (within the error bars), which sug- gested that the crystal configuration remained the same as more cobalt atoms replaced iron atoms (Figure 5). It also illustrated that the cobalt and iron atoms were equivalent. On occasions when there was a small variation in the isomer shift values, it usually was due to the charge imbalance that was screened by electrons on the iron atom itself. On the other hand, a decrease of the quadrupole splitting values was observed as the cobalt concentration was increased (Fig- ure 6), which was consistent with the assump- tion that the electric field gradient due to distant charges was suitably modified by anti- shielding factors. As iron atoms were added to the CoSi compound, the iron contribution was increased due to the presence of more nearest iron neighbors; however, the observed in- crease was not linear as there was an abrupt change in the quadrupole splitting around x = 0.2 (Figure 6). The line width was observed to experience a slight increase from 0.157 mm/sec at x = 0.8 to 0.189 mm/sec for x = 0.2 when cobalt was substituted for iron (Figure 7). This trend seemed reasonable since as cobalt replaced iron there would be fewer iron-iron neigh- bors. The result was consistent with that reported in the literature (Preston et al. 1966). The line width of the reference metal iron was 0.1478 = 0.0036 mm/sec. Finally, the measured hyperfine magnetic field increased as x increased, peaked around x = 0.2 and vanished for x = 0.5 (Table 1), which was consistent with reported increase in the magnetic moment in the range of 0.08 = x < (0.25 (Chernikov et al. 1997). Mossbauer Study Fe;.,Co,Si—Mono and Lahamer FeSi x= 0 Feo 9C09 181 x=0.1 ae Bh ae Feo gC0o 2Si x= 0.2 oe ee. | Feo 5C09 581 x=0.5 an a Feo 2Co Si x= 0.8 20 30 40 50 60 70 80 Position [°2 Theta] Figure 3. X-ray diffraction patterns of Fe,.,Co,Si(x=0, 0.1, 0.2, 0.5, and 0.8). 173 174 Journal of the Kentucky Academy of Science 69(2) FeSi x=O Feo 9COo | Si Feg 5C0O9 5S1 eaten 9 Hes ase tegen sOh Hg? ag ae Date, AP Feo 2C 09.881 x=0.8 = ere ee arr ks ae eG ea oa CS ee ‘ ee -1 0 8 6 agen -2 0 2 4 6 8 ° 40 v (mm/s) Figure 4. Mdéssbauer data fitted with one doublet for x = 0, 0.5. and 0.8 samples or with a doublet and a sextet for x = 0.1 and 0.2 samples. Table 1. and hmf is the hyperfine magnetic field. Mossbauer Study Fe;_,Co,Si—Mono and Lahamer 175 Summary of the results of the fitting routine (Recoil). I.S. is the isomer shift, A is the quadrupole splitting, Sample Linewidth (mm/sec) I.S. (mm/s) A (mm/s) hmf (mm/s) % Site Population Iron (Ref) - Sextet 0.148 + 0.004 0 0) 2.238 + 0.001 100 FeSi - Doublet 0.184 + 0.002 0.276 + 0.001 0.485 + 0.001 100 Fep gCoo {Si - Doublet 0.185 + 0.005 0.273 + 0.002 0.471 + 0.003 84 Fep gCoo 1Si - Sextet 0.172 + 0.044 0.005 + 0.041 1.268 + 0.028 16 Feo gCop 2Si - Doulet 0.189 + 0.005 0.278 + 0.002 0.489 + 0.003 62 Feo sCoo 2Si - Sextet 0.048 + 0.022 0.012 + 0.021 2.136 = 0.012 38 Fep5Cop9 5Si - Doublet 0.162 + 0.003 0.280 + 0.002 0.456 + 0.002 100 Feo Cog gSi - Doublet 0.157 + 0.006 0.278 + 0.003 0.403 + 0.004 100 CONCLUSIONS Mossbauer spectra of Feo 5Coo5Si and Feo 9- The synthesis of the Fe;.,Co,Si resulted in single phased alloys for all concentrations. Feo.gCoo.2Si exhibited a consistent x-ray dif- fraction pattern as found in the literature. Mossbauer spectroscopy results showed that Feo,sCoo2Si is magnetic and had an hmf of about 30.98 Tesla. When x = 0.5, the samples lost their magnetic ordering as illustrated by the absence of magnetic splitting in the 0.285 ereereer eee peececececce 0.28 0.275 Isomer Shift (mm/sec) 0.27 -0.2 0 0.2 Coo.85i alloys. ACKNOWLEDGEMENTS We would like to thank the Undergraduate Research and Creative Projects Program (URCPP) of Berea College for their support during the last two summers. We would like to thank the Appalachian College Association (ACA) for their financial support. We also eens eedeeweecea eee abeeweeceece ees e ses epee ceen t t 4 a t i 1 t + 1 iy ‘ 4 ty alesis eae af a eine 6 ne cinlaa aaa aaa eee Pe ee 0.4 0.6 0.8 1 Composition x in Fe Co Si Figure 5. Isomer Shift of Fe;_,Co,Si (x=0, 0.1, 0.2, 0.5, and 0.8). 176 0.55 | 2 05 DH : ‘ E $ $ ‘en A = 0.45 MN (