HLEIVAV EY Nath Eyritediris aoMeoreeN PLAIN MILA AOL LE NE PR ENA OM pert git PEPIN PUT el ES a NPT eRe ety gretit ana ri) “ Te eT er hind " Ae has . ARIA ag eT 4? aa ° Need neck a SEIT PREIS ,. " a = x. Pe AoA aN I, Nag Tt ialhibaba lh oak aol aaa . sf mi ~ sabia Dab 2 ih neta boo ad ce a e . HAL % herded hake nae a NE sen w eN \ AION adel, DINE Nod THT OT DEI Pores T CNM aI ND SNA EF IR I, es Lhe he : nae WA a etnies id adhndadhond Ahad tsbatecea ate Nw She IRE Pm Pibddadhcdittendiaoton PEN PEN ial NERS Tar padi te KEEL NY, try is Aue PR ol inh adudiodinteednne@entcamtaneratin erates planiieliaeiighed Subtineek na tira ele eee SL La nM ttn LI NTT a Dd WII ETON Fed a E Spicieinkedieie nite nse EERILY) Part LN. Pen ee > : RA ee ee PP MEE oY ss oo ett eT ET ue selpei-f athewb aed tigt Bakadull dkiedic Ghee nian alder eG tee tit - iiaibtieupadh od aad aiteen tea ee Ld Fiat if zs ee eS tt Nee ag" INE DL LEAS LEAP Nic a Boe NAB PPS = PM dein Daa tine find RAH Tas 5 Ph daa ase padatbedndiniind chained edt aed Nera EedodadintinkeDe nied ceri ee eee % : _ VL Td MET Ree ee er ees Snlbadevehniederd alae mute + PENS RP I ell 9) PLAIN IN IN Penile a eee A tu, g a a eer ee ete SSabnaiiaiiietin teethhn tne ae ae 2 PEP ie PN, a NN oe PPO Nee Me) aiidbedinni na td ea ana ee bactheninalie Pa Fan ma PS Me Nie b-bebatieiedoban deta Mona. PTT ie Pe NDNA lt niibeihndncd ates totais x Sed ON OR ILRI IN PPD oS OS Minune Teateaes TN IE te. : ee . a apie eatin Canasta Riididbahedatatasdin tetera orate ee SE Tee ae Sliced inode ke peaheane eat oe Peale Ete has ae MP Son en ee eee! a ae ot Mee r ae re rece Wont sets Nn ee an Nea aiehditiniokea see snes eadecnabib crete aaron $ : : * Me Lindy rated Natalee Wriireraiabhes tie c a “ Tanta Neen z s lienditse-dhedind nik olen deceit aaa Te ee MeN a Fat Rawat 2 oP. s widaicdiaakeedepibtdc tain odie tae ns _ : : 4 PPI ME Fen f ee eR Nd ee NP Stig. headland aed af a e a ae gee Ly iendiep-cbethals heed orotate are PAR ALE INTC IAN Perdana Wi Py oe tat tte Pd eas : m = 4 a ta ora ~ oe! ota eh : Pn tS oS Arion ar eee a a ih tailed ab ee ca rN co Pais Bi, ~ < Tang, 7, Sea elas ear ETN Nt Na Mae hit al ol pal ep Fa, ee hake ss ON eedrnat oS tae an en ac Maun est 3 a at a , _ 2H gh rage ws oy" : oe finer " no? : ; : x Z is Pirkei ag an Scaled Sond . Te Pett rae - s site ™, .. 7 ENN lit or Pode inne Ki Pes re . . 4 * me, NaS gta Reha: We Mec tas, é 5 ‘ “ 5, ate . — ~ anne ; : > meaty re : in = Sth Banners aed . . ee ree ete ee Vee Ne, Pe ee Sethaiedameiena theta NS Spel a Mantle ate, end, ~ euilenienietiedindined «leak oka ae eh ieee aah en aera Fee” S % Pim Pr Sa cans al iene 8 aL yey Pid ket ds a ae, sav a a 3 ‘ “ ieee ena ane elieibieiie sal in ee eee Mectintnepdiedi edie eae ren Saliamdiiedin ein die ech Pet tn, inhi te hee Pd Ma Peay enna YY, Wy tenis Ve ae & {le 75 MWC OE ‘ey aX J Ez CAINE a - “if pe ay “ ns = S Q Cred (7p) 2 —— NVINOSHLIWS Sa IuYvud Mtoe BRARI ES , SMITHSONIAN = / = as = < | 7 fos: , % = omeff z 3 OY 3 z Ss o Gy, = 2 = = a » 3 = .. MITHSONIAN INSTITUTION | NOLLALILSNI_NVINOSHLINS $3 lyvyugit +o Wem & Ce = \ = o =. |< = < = os y= we 5 ea 3 = 2 at 3 2 ney LIBRARIES SMITHSONIAN INSTITUTION z of es cea eae : . = wo == Yjfp OD Ys 2 > 5 £42 WN = 20 a Vf ki SNS 2 le, PRE 2 wz a Z a NOILNLILSNI INSTITUTION NVINOSHLINS S3SIY¥VUYUaIT x= ® ep) = wv“ d <= = < JONG = > 4 Zz = “ cae © os mf © a (7p) ~ w : 8 = ue) 8 s >” s Nexo” 5° 77) za (Tp) a ie LIBRARI ES SMITHSONIAN _INSTITUTION te) ~ rT) e a p a = Uf fe ) < lit = co Ge A La," /. Cc co oo” Y = a 5 S NOILNLILSNI INSTITUTION NOILNLILSNI INSTITUTION NVINOSHLINS S3Iu¥VYSIT a . is z “ OO oo = oa > > re “ z - m m (ap) w w . & S3SIYVYEIT LIBRARIES SMITHSONIAN n z on w” oa = =. «., = < =a 5. = S B 2 g 2 oS? i = = > = > = = w” Pane Tp) NOILALILSNI NVINOSHLINS z “ > oe 7) a” = 77) LN XR i FT or a WRAY raed Sal < =) -_ > - a = Wet ia a \ — pane SaIYVUYdIT LIBRARIES SMITHSONIAN saiuvugi S3aluyv NOILALI. NVINOSHLINS NOILNLILSNI NOILNLI. INSTITUTION LIBRAR 7 NOILAL LIBRAR NOILALI. NVINOSHLINS NOILALI NOILALILSNI STITUTION - . o a e = ey Xe) < . — 0) ALS ee 9) — OATS = yk ~ - \ y ~ @ aus m LOND. Yo m JON BS S on EBRD ! = 7) = a) ca SONIAN NOILNLILSNI SHLINS S3JIYVHYSIT LIBRARIES = 22) | = 2] z aah = = .< = < \ oe = z — =m S 5 a re) eS O NS oo 3 2 SY a Ce e 2 E 2 E i > > = : rs Zz a Zz 77) Sey SMITHSONIAN INSTITUTION NOILNLILSNI 4 - ISONIAN INSTITUTION. NOILNLILSNI NVINOSHLINS S3IUVYHSIT . wn ue Ww? — (2p) ; N us a tu .o | NY ox = ao. a a KN o< Se 4 lyvug IT bs BRARIES yg = O Seah fe) o = ( Mp? ee] = (os) ‘ — wo ; Ly, 2 = 2 AN 5 2 ! Pie La ie \s INS 5 > aia = = YAY eE oe | m = m \ \S - an na = in aN: Zz : mn , S3IYVYE!IT LIBRARIES SMITHSONIAN INSTITUTION '¢2) i z “@ w < @ wn = =e = — = = _ = = z 4 z @ @ )& & 2. = = = 2 = a 3 Z 5 SONIAN INSTITUTION iy i 2 rr ul a ul on Neu a = a = fe) = oO SS = re) Zz —J z J z E SHLINS SSINVHSEIT LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI c. z E = a = - ° o = ‘w oS : “y = Se) Al Ee > Ne = > = 5 > = S = se a = res ee: 7 Z iE Z | SONIAN INSTITUTION NOIJLNLILSNI S3SIY¥VYEIT LIBRARIES z FS 2) rs ” = ees ( < Ne = = = Sop NS ; 5 . SAS ae 5 = Bi 5B WY: NR 2 2 8 GIT 2 E Z = 2 Wy «© 2 s > = > Ff = ' on As) 7) z 7) Lee SHLIWS —S3 byVvVegiy_ pM ITHSONIAN INSTITUTION NOILNLILSNI -_ « G i 7 . a = i < = y= : b 2 5 i“ Zz a w 9) > A lYVYaITLIBRARIE TITUTION NOILNLILSNI YVYUGit LIBRARIES STITUTION &, Pip ISSN: 0098-4590 Florida Scientist Volume 56 Winter, 1993 Number | CONTENTS Movement Patterns of Translocated Big Cypress Fox Squirrels RMN CAA OLCON FING) 8 a6. oe wlaveutdvntedsuadacl st anghetaesnevesneshmonennedesobelis’s Randy S. Kautz 7 The Ecological Basis of the Kissimmee River Restoration Plan .................4. . Louis A. Toth 25 Low Clutch Viability of American Alligators on Lake Apopka...........0...0000 Alan R. Woodward, H. Franklin Percival, Michael L. Jennings, and Clinton T. Moore 52 First Record of the Eastern Big-eared Bat (Plecotus rafinesquii) in Southern Florida QUARTERLY JOURNAL OF THE FLORIDA ACADEMY OF SCIENCES FLORIDA SCIENTIST QUARTERLY JOURNAL OF THE F'LORIDA ACADEMY OF SCIENCES Copyright© by the Florida Academy of Sciences, Inc. 1993 Editor Dr. DEAN MARTIN Co-Editor: Mrs. BARBARA B. MARTIN Institute for Environment Studies Department of Chemistry University of South Florida 4202 East Fowler Avenue Tampa, Florida 33620-5250 THE FLORIDA SCIENTIST is published quarterly by the Florida Academy of Sciences, Inc., a non-profit scientific and educational association. Membership is open to indi- viduals or institutions interested in supporting science in its broadest sense. Applica- tions may be obtained from the Executive Secretary. Both individual and institutional members receive a subscription to the FLORIDA SCIENTIST. Direct subscription is avail- able at $40.00 per calendar year. Original articles containing new knowledge, or new interpretation of knowledge, are welcomed in any field of Science as represented by the sections of the Academy, viz., Biological Sciences, Conservation, Earth and Planetary Sciences, Medical Sciences, Physical Sciences, Science Teaching, and Social Sciences. Also, contributions will be considered which present new applications of scientific knowledge to practical problems within fields of interest to the Academy. Articles must not duplicate in any substantial way material that is published elsewhere. Contributions are accepted only from members of the Academy and so papers submitted by non-members will be accepted only after the authors join the Academy. Instructions for preparation of manuscripts are inside the back cover. Officers for 1992-93 FLORIDA ACADEMY OF SCIENCES Founded 1936 President: Dr. THomas L. CrisMAN Treasurer: Dr. MarVIN IvEY Department of Environmental 14452 Hillview Drive Engineering Sciences Largo, Florida 34644 University of Florida Gainesville, Florida 32611 Executive Secretary: Mrs. BETTY PREECE P.O. Box 033012 President-Elect: Dr. Patrick J. GLEASON Indialantic, Florida 32903-0012 1181 Palmway Tel: 407:723-6835 Lake Worth, Florida 33460 Program Chair: Dr. DEL DELUMYEA Secretary: Dr. Patricia M. Dooris Millar Wilson Laboratory P.O. Box 2378 for Chemical Research St. Leo, Florida 33574 Jacksonville University Jacksonville, Florida 32211 Published by the Florida Academy of Sciences, Inc. P.O. Box 033012 Indialantic, Florida 32903-0012 Printing by C & D Printing Company, St. Petersburg, FL Florida Scientist QUARTERLY JOURNAL OF THE FLORIDA ACADEMY OF SCIENCES DEAN F. Martin, Editor BARBARA B. MartTIN, Co-Editor Volume 56 Winter, 1993 Number 1 Conservation MOVEMENT PATTERNS OF TRANSLOCATED BIG CYPRESS FOX SQUIRRELS (SCIURUS NIGER AVICENNIA) Patrick G. R. JopIcE! Department of Wildlife and Range Sciences University of Florida, Gainesville, Florida 32611 ABSTRACT: Five Big Cypress fox squirrels (Sciurus niger avicennia) were radio-collared and translocated from Naples, Florida, to Big Cypress National Preserve (BICU). Fox squirrels survived trapping and immediate release. Site fidelity was inconsistent and movements may be attributed to dispersal, post release investigative behavior, or food availability. Habitat use of translocated fox squirrels shifted among seasons. Translocated fox squirrels sought out suitable habitat and resident populations. TuE Big Cypress fox squirrel (Sciurus niger avicennia), listed as threatened by the Florida Game and Fresh Water Fish Commission, is 1 of 3 subspecies of fox squirrel which occur in Florida. The distribution is restricted to southwestern Florida, south of the Caloosahatchee River and west of the Everglades, and is disjunct from the range of other subspecies of fox squirrel. The subspecific name and often-used common name of mangrove fox squirrel incorrectly imply restriction to one habitat. To reflect its approximate distribution, the common name Big Cypress fox squirrel (BCFS) is preferable (Humphrey and Jodice, 1992). The subspecies was originally described by Howell (1919)and little is known about its ecology (Moore, 1956; Williams and Humphrey, 1979). Extirpation of BCFS from some areas and a probable decline in population in BICY have increased the need for general ecological information. Surveys have documented the difficulty of locating fox squirrels in BICY (Williams and Humphrey, 1979; Jodice, 1990). The objectives of this study were to use translocated BCFS to 1 Present Address: Florida Game and Fresh Water Fish Commission, Nongame Section, Rt. 7 Box 440, Lake City, FL 32055 9 FLORIDA SCIENTIST [VOL 56 supplement information on habitat use of fox squirrels in BICY and use this information to improve survey techniques. Additionally, the feasibility of using translocation for re-stocking would be assessed. STrupy AREA—Sources for translocated fox squirrels were residential areas and golf courses in Naples, Collier County, Florida, where a concurrent project exam- ined behavior and diet of BCFS in an urban setting (Jodice and Humphrey, 1992). A description of those areas appears therein. BICY (230,000 ha) is located in southwestern Florida, Collier and Monroe Counties, and has a tropical savanna climate with spring droughts, heavy summer rains, and mild, dry winters (Hela, 1952). Release sites were located north of U.S. Highway 41 along Burns Road or State Road 839 (Turner River Road) and were characterized by cypress (Taxodium distichum) domes and strands, second-growth pinelands of south Florida slash pine (Pinus elliotti var. densa), mixed swamps of cypress with associated hardwoods, and prairies. BICY is a multiple-use area with hunting, fishing, off-road vehicle use, and oil production occurring. There are more than 200 privately owned inholdings in the form of back-country camps. Duever and co-workers (1986) give an extensive description of the natural history of BICY. MeEtTHODs—Five BCFS were captured in Naples, Florida, from September 1989 through February 1990. Bird seed, peanuts, and bananas were used as bait. Individuals were captured in either a live trap or cloth sack and were transported to BICY, where they were weighed, sexed, measured, and aged as either juvenile or adult (aging based on examination of the ventral surface of the tail; Larson and Taber, 1980). BCFS were fitted with radio collars weighing 14 - 18 g (about 2.2% of average body weight of the 5 captured squirrels) and were released within 2 hours of capture in areas of recent fox squirrel observations. I attempted to visually locate translocated BCFS daily, recording locations using universal trans- verse mercator system (UTMs) coordinates and plotting them on 1:24,000 USGS quadrangle maps. At each location I recorded habitat type, understory density (visually estimated), squirrel position in the vegetation strata, time of day, weather, surrounding habitat type, and ground water level. Home ranges were estimated using computer program HOME RANGE (Ackerman et al., 1990). The measures used were: 1) 95% minimum convex polygon (MCP; Mohr, 1947), 2) 95% weighted bivariate normal ellipse (WBIV; Jennrich and Turner, 1969), and 3) harmonic mean (HM) and corresponding core area (CA; Dixon and Chapman, 1980). ResuLts—Mortality did not occur during trapping, transportation, collaring, or release. I obtained 206 radio locations for 5 individuals from 12 September 1989 - Ly April 1990 (Table 1). M1 remained within 100 m of the release site for 5 days. He then travelled an elliptical path with a maximum distance from the release site of 4.7 km, returning to within 200 m of the release site after 50 days. No signal was received after his return. Mean distance travelled was 270 m/day. M2 was released 28 days later than M1 at the same site. Except for a5 day, 7.5 km movement, he remained within 1.5 km of the release site for 108 days. He was recaptured and moved 4.5 km NW when he began frequenting a house where food was provided. M2 then established a central nest and remained within 1 km of the nest for 48 days. He then moved 10 km NE where he was observed courting a resident female. M2 slipped his collar soon after this event. Estimated home range No. 1, 1993] JODICE—MOVEMENT PATTERNS OF TRANSLOCATED BIG CYPRESS FOX SQUIRRELS 3 TABLE 1. Movements of Big Cypress fox squirrels translocated from Naples to Big Cypress National Preserve, Florida, September 1989 - April 1990. Maximum Final Squirrel dist. from dist. from sex and Date Days of No. of release release number released contact locations site (m) site (m) M1 09/12/89 50 40 4,700 200 M2 10/10/89 108 i 7,400 750 M2» 02/01/90 76 70 11,000 11,000 Fl 01/13/90 2 3 — OD) M3 01/14/90 i 6 3,000 3,000 M4 02/19/90 tat) 12 32,000 32,000 4M = males, F = females > Record of second release size during the first 108-day period (excluding locations after frequenting the house) was: 95% MCP = 128.7 ha, 95% WBIV = 66.4 ha, 95% HM = 223.6 ha, and CA (66.1% of utilization volume) = 67.0 ha. Estimated home range size for the second 48-day period was: 95% MCP = 71.9 ha, 95% WBIV = 96.8 ha, 95% HM = 270.6 ha, and CA (62.1% of utilization volume) = 52.3 ha. F 1 slipped her radio collar within 2 days of release while still at the release site. M3, the only immature individual, was released where F 1 had been released the previous day. M3 left the release site on the second day and was found dead 5 days later, apparently from predation. M4 remained within 50 m of the release site for 4 days. His last location, 51 days later, was 32 km NW of the release site (i.e., in the direction of the capture site). No signal was received after this time. Home ranges were not estimated for these three individuals. Seasonal trends in habitat use are based on pooled data for all translocated BCFS (Fig. 1). Translocated BCFS spent much time in the trees foraging on cypress cones during the late wet season. During the early and late dry seasons, as water levels receded, BCFS were often observed foraging on the ground in pinelands and digging shallow pits at the base of pine trees. This probably indicates foraging for hypogeous fungi, an important winter and early spring food for southeastern fox squirrels (Weig] et al., 1989). Cypress habitat was also used during the early and late dry seasons for nesting, mid-day inactivity, and travelling. Fourteen nests of translocated BCFS were found in BICY. All were in co- dominant or dominant cypress trees in cypress or mixed-swamp habitat. Two nests were stick structures and 12 were in a locally common bromelliad, quill-leaf or stiff- leafed wild pine (Tillandsia fasciculata). BCFS nested among the long leaves of the bromelliad and on the organic matter where the plant attaches to the tree and often added stripped cypress bark to the plant. Mean nest height was 9.4 m (S.D. = 1.8) and mean distance from nest to tree top was 3.6 m (S.D. = 1.9). 4 FLORIDA SCIENTIST [VOL 56 80 - Seasons Z Late wet (Sep-Nov) 70> |_| Early dry (Dec-Feb) aad Late dry (Mar-May) % OF LOCATIONS & fo) ANY TN \ ANY \ \\\\\ \\ LLL LZ Cypress Prairie Pineland HABITAT TYPE \\ \\\\ \\ ‘ FIG. 1. Locations of translocated Big Cypress fox squirrels by habitat type and season in Big Cypress National Preserve, Florida, September 1989 - April 1990. Discussion—Translocated BCFS travelled widely and release site fidelity was inconsistent. Distances travelled by M1, M2, and M4 were substantially further than the 1.5 km and 4 km movements reported for translocated red squirrels (S. vulgaris) in Great Britain (Bertram and Moltou, 1986) and translocated Delmarva fox squirrels (S.n. cinereus) in Maryland (Theres, 1992), respectively. Movement away from release sites may be attributed to dispersal (M2, M3, and M4) and post-release investigative behavior (M1 and M2). Food resources may have also influenced movements of translocated BCFS. BCFS were released when food crops in BICY were at an annual low, and this may have affected movement patterns and site fidelity. In a naturally patchy landscape such as BICY, where food items are variable among habitats, a fox squirrel may range over a large area and variety of habitats to obtain the necessary amount of food. Home-range size and habitat use of other southeastern fox squirrels are influenced in this manner (Weig] et al., 1989; Kantola and Humphrey, 1990). Use of bromelliad nests by BCFS is previously unreported. Bromelliad nests are structurally similar to leaf and stick nests, yet require only minimal maintenance or construction. Bromelliads are acommon perennial on cypress trees in BICY and may provide virtually limitless nesting sites, allowing squirrels to travel without construct- No. 1, 1993] _JODICE—MOVEMENT PATTERNS OF TRANSLOCATED BIG CYPRESS FOX SQUIRRELS — 5 ing nests. One goal of this translocation project was to improve survey techniques for resident fox squirrels in BICY. Currently, biologists in BICY are opportunistically surveying for bark-stripped cypress trees, a sign of nesting activity and a field sign previously unused. Habitat use data are also being used to develop future surveys. Although translocated fox squirrels were able to find resident populations (as illustrated by M2 courting a resident female), translocation techniques need to be improved for successful re-stocking. Site fidelity may be improved by releasing squirrels during summer food peaks (ripening of pine seeds) and into sites with access to a variety of habitats. Survival of BCFS during capture and release indicates these techniques were satisfactory. Minimum survival of 3 BCFS of 50, 55, and 184 days implies habitat was suitable near release sites, or surrounding habitat was at least able to support a travelling fox squirrel. ACKNOWLEDGMENTS—This study was funded by the National Park Service, Big Cypress National Preserve. J. R. Snyder and D. K. Jansen at Big Cypress National Preserve facilitated all phases of this study. Other Park Service staff assisting were R. Beymer, G. Greco, D. Teague, G. C. Ward, D. Weeks, and volunteer B. Rogers. K. Meyer (University of Florida, Gainesville) built the radio collars and gave valuable advice throughout all stages of the project. B. A. Millsap, D. E. Runde, L. L. Williams, J. B. Wooding, and an anonymous reviewer provided useful comments on the manuscript. S. R. Humphrey, Florida Museum of Natural History, served as my major advisor and provided excellent guidance throughout all phases of this work. LITERATURE CITED ACKERMAN, B.B., F.A. LEBAN, M.D. SAMUEL, AND E.O. Garton. 1990. User’s manual for program home range. Second edition. Tech. rept. 15, Forestry, Wildlife, and Range Experiment Station, Univ. of Idaho, Moscow. 79pp. Bertram, B.C.R., AND D.P. Ma.tou. 1986. Reintroducing red squirrels into Regent’s Park. Mammal Review 16:81-88. Dixon, K.R., AND J.A. CHAPMAN. 1980. Harmonic mean measure of animal activity areas. Ecology 61:1040- 1044. Duever, M.J., J.E. Carison, J.F. MEEDER, L.C. Duever, L.H. Gunperson, L.A. RIOPELLE, T.R. ALEXANDER, R.F. MYERS, AND D.P. SPANGLER. 1986. The Big Cypress National Preserve. Research report no. 8 of the National Audubon Society. National Audubon Society. New York. 444pp. He a, I. 1952. Remarks on the climate of southern Florida. Bull. Marine Sci. Gulf Caribb. 2:438-447. Howe tL, A.H. 1919. Notes on the fox squirrels of the southeastern United States, with description of a new form from Florida. J. Mammal. 1:36-38. Humpneey, S.R. anpD P.G.R. Jopice. 1992. Big Cypress fox squirrel. Pp. 224-233. In S.R. Humphrey (ed.), Rare and Endangered Biota of Florida. Mammals. Univ. Florida Presses, Gainesville. JENNRICH, R.I. AND F.B. Turner. 1969. Measurement of non-circular home range. J. Theo. Biol. 22:227- 2a. Jopice, P.G.R. 1990. Ecology and translocation of urban populations of Big Cypress fox squirrels (Sciurus niger avicennia). M.S. Thesis, Univ. Florida, Gainesville. 89 pp. AND S.R. Humphrey. 1992. Activity and diet of an urban population of Big Cypress fox squirrel. J. Wildl. Manage. 56:685-692. KanTOoLa, A.T., AND S.R. Humpurey. 1990. Home range and mast crops of Sherman’s fox squirrel in Florida. J. Mammal. 71:411-419. Larson, J.S., AND R.D. Taser. 1980. Criteria of sex and age. In: ScHEmNitTz, S.D. (ed.), Wildlife Management Techniques Manual (4th ed.), The Wildlife Society, Washington, D.C. Moore, J.C. 1956. Variation in the fox squirrel in Florida. Am. Midl. Nat. 55:41-65. Mour, C.O. 1947. Table of equivalent populations of North American small mammals. Am. Midl. Nat. 37:223-249. 6 FLORIDA SCIENTIST [VOL 56 THERES, G.D. 1992. Maryland Department of Natural Resources, Wye Mills, MD, Pers. Commun. WEICL, P.D., M.A. STEELE, L.J. SHERMAN, J.C. HA, AND T.S. SHARPE. 1989. The ecology of the fox squirrel in North Carolina: Implications for survival in the Southeast. Tall Timbers Research Station Publ. No. 24. Tallahassee, Florida. 93pp. Wiuiams, K.S., AND S.R. Humpurey. 1979. Distribution and status of the endangered Big Cypress fox squirrel (Sciurus niger avicennia) in Florida. Florida Scient. 42:201-205. Florida Scient. 56 (1): 1-6.1993 Accepted: June 5, 1993. REVIEW Alice Lefler Primack, Journal Literature of the Physical Sciences, Scarecrow Press, Inc., Metuchen, NJ, 1992. Pp. x + 22. 5.5 x 8.75. Price: $29.95 ON-LINE searching for literature is a current trend that will become increas- ingly important in the coming years. Before one can have a good appreciation of on-line or other electronic database searching, it is useful to have an understand- ing of the basic literature. This book describes in order: the nature of scientific literature, search strategies, some useful information about on-line searching, use of index and abstracting sources, obtaining the articles of interest, writing scientific journal articles, and it concludes with a directory of “core” journals in the physical sciences. The directory is based on Science Citation Index Impact Factors for 1985 and 1986. That presumes, of course that Impact Factors are a valid means of comparing dicerse journals. Much useful information is packed into this book by the author, who is a librarian in the Science Library of the University of Florida. Those who use the literature of the physical sciences or who are considering publishing in this field wil find the book useful. — Dean F. Martin, University of South Florida No. 1, 1993] a Conservation Sciences TRENDS IN FLORIDA WILDLIFE HABITAT 1936-1987 RANDY S. KAUTZ Florida Game and Fresh Water Fish Commission, 620 S. Meridian Street, Tallahassee, FL 32399-1600 ABSTRACT: Since Europeans first arrived in Florida, at least 9 vertebrate taxa have been driven to extinction, 5 of which disappeared in the last 50 years. In addition, Florida is home to 56 taxa listed as threatened and endangered, and 44% of all vertebrate taxa in Florida are probably declining. The reason most often cited for the problems now facing Florida wildlife is habitat destruction. Data from the U.S. Forest Service were reviewed to quantify gross changes in land use, land cover, and wildlife habitat over time. During the period 1936-1987, Florida lost 1.74 million ha of forest and 1.57 million ha of herbaceous wetlands to development, predominantly to the expansion of urban areas, agriculture, and range lands. The preponderance of forest land lost was covered by pines (Pinus sp. ). Longleaf pine (P. palustris) forests suffered the most, declining 88% in 51 years: At observed rates of loss, longleaf pine forests will disappear from unprotected lands in Florida by 1995, sand pine (P. clausa) scrub forests by 2047, and pond pine (P. serotina) wetlands by 2024. Since 1970, upland hardwood forests have decreased 27%, and forested wetlands have decreased 17%. PONCE de Leon first sighted land in Florida in 1513 (Tebeau, 1971). Between then and the early 1900s, at least four vertebrate taxa disappeared from Florida at the hand of European man. The plains bison (Bison bison bison) was extirpated from Florida in the late 1700s or early 1800s as a result of wanton slaughter by the early settlers (Ehrhart, 1978). The Florida red wolf (Canis rufus floridanus) disappeared from the state in the early 1900s (Layne, 1978). The Carolina parakeet (Conuropsis carolinensis) was virtually extinct by 1900, exterminated by man as agricultural pests, for food, and for sport (Hardy, 1978a). The passenger pigeon (Ectopistes migratorius), a common winter visitor in the northern third of Florida, was driven to extinction by 1914, the victim of mass slaughter for food and sport (Hardy, 1978b). Since the 1930s, the rate of vertebrate extinctions in Florida has risen dramati- cally. The ivory-billed woodpecker (Campephilus principalis), last sighted in Florida in 1969, met its demise due to the logging of mature stands of lowland hardwoods. The last dusky seaside sparrow (Ammodramus maritimus nigrescens) died in 1987, a casualty of the destruction of salt marshes along the central east coast of Florida for mosquito control. The Chadwick Beach cotton mouse (Peromyscus gossypinus restrictus), which inhabited coastal dunes in Sarasota County, has not been seen since 1938 (Repenning and Humphrey, 1986). The pallid beach mouse (P. polionotus decoloratus), which occurred in sand dunes along the coasts of Flagler and Volusia counties, has not been seen since 1946 (Humphrey and Barbour, 1981). Both taxa are presumed extinct as a result of residential and commercial development. Goff's pocket gopher (Geomys pinetus goffi), which has not been seen since 1955, is presumed extinct as a result of conversion of its habitat to urban uses (Humphrey, 1981). 8 FLORIDA SCIENTIST [VOL 56 Florida is also home to a large number of taxa whose continued survival is in jeopardy. The Florida Game and Fresh Water Fish Commission’s current list of endangered and threatened species contains 56 vertebrate taxa, excluding fishes and whales (Wood, 1990). The number of taxa on Florida’s list of endangered and threatened species is second only to that of California (U.S. Fish and Wildlife Service, 1989). Among Florida’s endangered species are several that are perilously close to extinction. For example, the Florida panther (Felis concolor coryi) numbers fewer than 50 individuals (U.S. Fish and Wildlife Service, 1987), and the Key deer (Odocoileus virginianus clavium) and American crocodile (Crocodylus acutus) each number fewer than 400 individuals (Klimstra and Hardin, 1978; Ogden, 1978). In addition to the obvious concern for species listed as endangered and threatened, declining trends have been observed in species not yet listed and heretofore presumed safe. Millsap and co-workers (1990) found that 44 percent of Florida’s 668 vertebrate taxa are probably declining based on a broad sampling of expert opinion. Cox (1987) analyzed U.S. Fish and Wildlife Service Breeding Bird Survey data for Florida and found that, of 85 species sampled often enough to support quantitative analysis, 15 species of breeding birds in Florida showed decreasing trends. Among those decreasing are the red-headed woodpecker (Melanerpes erythrocephalus), northern flicker (Colaptes auratus), brown-headed nuthatch (Sitta pusilla), and eastern bluebird (Sialia sialis), all of which nest in cavities in or on the edges of Florida’s increasingly fragmented forests. Declining species more commonly associated with open rural areas include loggerhead shrike (Lanius ludovicianus), common ground dove (Columbina passerina), eastern mead- owlark (Sturnella magna), and eastern kingbird (Tyrannus tyrannus). The reasons for decline among these latter species are not readily apparent. Numerous studies have identified forest fragmentation and loss of habitat as the major causes of declining wildlife populations (Galli et al., 1976; Robbins, 1979; Matthiae and Stearns, 1981; Whitcomb et al., 1981; Harris and Wallace, 1984; Cox, 1988). Unfortunately, no published reports exist which quantify temporal changes in land use and land cover patterns in Florida. Therefore, it is difficult to put into perspective the extent to which, and how rapidly, the Florida landscape and its wildlife habitats have changed. The purpose of this paper is to provide quantitative information on these subjects using the best available data. METHODS—Between 1936 and 1987, the U.S. Forest Service (USFS) conducted six systematic inventories of forest resources in Florida. The results of these inventories appear in a series of USFS technical reports (Eldredge, 1938a,b; Ineson and Eldredge, 1938; McCormack, 1950; Larson and Goforth, 1961; Knight and McClure, 1971; Bechtold and Knight, 1982; Brown and Thompson, 1988). Gross changes in land use and forest types are characterized over time (Table 1). All data in the reports were collected using a combination of field survey and air photo interpretation techniques. For this paper, land use and forest type data indicative of major changes in wildlife habitat were extracted from the USFS reports and analyzed for trends. Additional unpublished data from the USFS Southeastern Forest Experiment Station also were used in the analyses. USFS definitions (Brown and Thompson, 1988) of the major land use, land cover, and forest types inventoried appear in the Glossary. Several minor corrections have been made to the USFS data reported herein. First, each of the USFS reports contains a different estimate of the total land area of Florida. To correct for minor differences, all of the land use and forest type data in each report were normalized to 14,127,100 ha, the figure reported by Anderson and Fernald (1981) for the total land area of Florida. No. 1, 1993] KAUTZ—TRENDS IN FLORIDA WILDLIFE HABITAT 1936-1987 9 TABLE 1. Tabular report of the data used to describe changes in land use, forest types, and wildlife habitat in Florida, 1936-1987. Values in parentheses were estimated. 1936 1949 1959 1970 1980 1987 Population 1,725,700 2,683,900 4,733,500 6,791,400 9,740,000 12,042,800 Land Use (Ha) Forest 8,461,800 8,468,300 7,967,600 7,263,600 6,959,800 6,722,300 Marsh (2,821,400) 2,145,100 2,454,000 1,476,900 1,229,500 1,251,300 Ag. & Range 2,449, 500 2,726,100 3,286,000 4,128,100 4,377,200 4,168,200 Urban & Other (294,400) 432,600 719,600 1,158,500 1,460,600 1,885,200 Total Area 14,027,100 14,027,100 14,027,100 14,027,100 14,027,100 14,027,100 Commercial vs. Noncommercial Forest Lands (Ha) Commercial 7,639,700 7,658,300 6,813,600 6,574,500 6,362,800 6,086,100 Noncommercial 822,200 810,000 753,900 689,100 597,000 636,300 Commercial Pine Forests (Ha) Longleaf Pine 3,090,800 2,476,400 1,312,400 607,400 504,800 386,300 Slash Pine 2,324,200 2,047,000 1,754,800 2,143,700 2,151,900 2,111,900 Loblolly Pine 259,000 260,600 147,400 140,000 167,300 235,000 Sand Pine 162,300 157,900 146,200 196,400 218,300 247,900 Pond Pine 154,800 155,500 131,900 134,500 94,700 64,100 Total Pine 6,002,500 5,109,300 3,006,900 3,236,200 3,152,100 3,057,400 Commercial Hardwood Forests (Ha) Oak-pine 339,100 621,800 978,500 491,800 Oak-hickory 608,300 998,400 1,098,300 1,104,600 866,600 767,900 Total Upland 608,300 998,400 1,437,500 1,726,400 1,445,100 1,259,700 Lowland 1,303,300 I TOUL ACO 1,973,900 1,738,000 1,761,800 1,773,000 Tot. Hardwood — 1,911,500 2,709,600 3,411,300 3,464,400 3,206,800 3,032,700 Cropland and Pasture & Range (Ha) Cropland 1,426,600 1,495,500 1,537,300 1,599,300 Pasture & Range 1,859,400 2,632,600 2,840,000 2,568,900 Second, Knight and McClure (1971) point out that 0.9 million ha of Florida were classified as forest land in the 1936, 1949, and 1959 surveys but were reclassified as rangeland for the 1970, 1980, and 1987 surveys. The lands that were reclassified supported no timber during any of the surveys nor did they show evidence of regeneration. By and large, the reclassified lands are dry prairies. Dry prairies are treeless savannahs dominated by saw palmetto (Seronoa repens) and native grasses and forbs (Abrahamson and Hartnett, 1990), and they are commonly used as native rangeland. To account for this change, normalized estimates of the 0.9 million ha of reclassified lands were subtracted from the forest land category and added to the agriculture and rangeland category for the area estimates reported for the 1936, 1949, and 1959 surveys. Third, USFS separated marsh lands from other cover types in all but the 1936 survey during which marsh lands were lumped with the “urban and other lands” class. To estimate the area of marsh lands present in Florida in 1936, a linear regression equation was developed to predict the area of “urban and other lands” using the size of the human population in the 1949-1987 surveys. The resulting equation (7 = 0.99, P < 0.01) was used to estimate the area of “urban and other lands” in 1936 using the known size of the Florida population at that time. The area of marsh lands in Florida in 1936 was calculated as the aah between the known area of the marsh/urban/other class and the estimated area of the urban/ other class. 10 FLORIDA SCIENTIST [VOL 56 Fourth, USFS stafflumped their area estimates for cropland and rangeland together in the 1936 and 1949 surveys. For this reason, it has been necessary to report these two major land use classes in the category of agricultural land for the entire 1936-1987 study period. However, USFS reported separate estimates of cropland and rangeland in the last four surveys, and these data are also reported. RESULTS—Land Use and Land Cover 1936-1987—Forest land—Forests cov- ered 8.46 million ha, or 60 percent of the land area, of Florida in 1936 (Fig. 1). By 1987, forest area had declined to 6.72 million ha, or 48 percent of Florida. This is a total loss of 1.74 million ha of forest, a 21 percent decline, over the 51-year period covered by the USFS reports. The average annual rate of loss between 1936 and 1987 was approximately 34,100 ha per year. Forest area remained roughly constant between 1936 and 1949 such that all of the decline in forest area actually occurred after 1949 (Fig. 2a). Thus, since 1949, the average rate of loss of forest has been 45,700 ha annually. Marsh land—The total area of marsh land in Florida declined from 2.82 million ha (or 20 percent of the land area of Florida) in 1936 to 1.25 million ha (9 percent of Florida) in 1987 (Fig. 1). This is a total loss of 1.57 million ha, or a 56 percent decline. The average annual rate of loss over this period was 30,800 ha per year. The period of most rapid decline in marsh area occurred between the 1959 and 1970 surveys when over 0.97 million ha of marsh were destroyed (Fig. 2b). The area of marsh land in Florida has increased slightly since 1980 (Fig. 2b), but this increase is within the range of sampling error reported by Brown and Thompson (1988). Agriculture and rangeland—Cropland and rangeland combined increased from 2.45 million ha, or 17 percent of the land area of Florida, in 1936 to 4.17 million ha, or 30 percent of Florida, in 1987 (Fig. 1). Rangeland accounted for 2.57 million HECTARES (Millions) FOREST MARSH AGRICULTURE URBAN @ 1936 [3 1987 Fic. 1. Gross changes in major land cover and land use in Florida between 1936 and 1987. The numbers above each bar are percentage of total land area. No. 1, 1993] KAUTZ—TRENDS IN FLORIDA WILDLIFE HABITAT 1936-1987 11 ha (62 percent) and cropland for 1.60 million ha (38 percent) of the lands in this class in 1987. Cropland and rangeland increased a total of 1.72 million ha over the 51-year period, an average annual increase of 33,600 ha per year. Eighty-two percent of this increase occurred between 1949 and 1970 (Fig. 2c). Most of the increases since 1959 were due to increases in rangeland rather than cropland (Fig. 2c). Croplands increased a total of only 174,100 ha between 1959 and 1987. Rangeland area decreased by 271,300 ha between 1980 and 1987. Urban and other land—In 1936 urban and other land covered only 294,400 ha, or 2 percent, of the land area of Florida (Fig. 1). However, by 1987 urban and other lands covered 1.89 million ha, or 12 percent, of Florida. This is an increase of 1.59 million ha, or 538 percent, in 51 years and an average annual increase of over 31,200 ha per year. Forest Cover 1936-1987—Commercial and noncommercial forests-On the average, commercial forest lands comprised 90 percent of all forest land in Florida during the study period (Fig. 3a). Commercial forest lands declined from 7.64 million ha in 1936 to 6.09 million ha in 1987. Noncommercial forest lands declined from 0.82 million ha in 1936 to 0.64 million ha in 1987. USFS collects detailed stand information only on lands categorized as commer- cial forest. Therefore, the following discussions track changes only in forest types growing on commercial forest lands. Data that would depict changes in forests growing on unproductive sites or on lands reserved from commercial timber production are not available and are not included. Pine and hardwood forests—Pine forests declined sharply from 6.00 million ha in 1936 to 3.51 million in 1959, and they continued to decline, although more slowly, to 3.06 million hain 1987 (Fig. 3b). This is a total loss of 2.94 million ha (or 49 percent) in 51 years, an average rate of loss of 57,700 ha per year. Over the same period, hardwood forests increased sharply from 1.91 million ha in 1936 to 3.41 million ha in 1959, peaked at 3.47 million ha in 1970, and have declined slowly since. By 1987 hardwood forests had declined to 3.03 million ha, a 13 percent loss between 1970 and 1987. Longleaf pine and slash pine forests—Longleaf pine forests covered 3.09 million ha of Florida in 1936 (Fig. 3c). At the time, longleaf pine forests were the dominant forest type in the state, comprising 51 percent of all commercial pine forests, 40 percent of all commercial forest land, and 22 percent of the entire Florida landscape. Longleaf pine forest area declined steadily over the entire study period. By 1987, it occurred on only 0.38 million ha of land. This is an 88 percent reduction in longleaf pine forest area in only 51 years, a loss of 52,600 ha annually. If this rate of decline continues into the future, longleaf pine forests will disappear from Florida on all but public lands by the year 1995. In 1936, forests dominated by slash pine covered 2.32 million ha of Florida (3c). At the time slash pine was Florida’s second most abundant forest tree, accounting for 39 percent of all commercial pine forests, 30 percent of all commercial forest land, and 17 percent of the land area of Florida. Slash pine forests declined to a low of 1.75 12 FLORIDA SCIENTIST [VOL 56 9 3 FOREST LAND MARSH LAND _ 8.5 o @ 2.5 7 ea 7.5 : = Ww Ww BS r 1.5 7 6.5 1 1930 1940 1950 1960 1970 1980 1990 1930 1940 1950 1960 1970 1980 1990 (a) (b) 5 2 ae CROPLAND AND RANGELAND URBAN AND OTHER LAND a 4 @ 1.5 cS 2 2 = 3.5 = = = ~ 3 palegelane o 1 a < = 2.5 | wy wy xr 2 r 0.5 TS mo) Oe (ee pT ec nee cese ae 1 0 1930 1940 1950 1960 1970 1980 1990 1930 1940 1950 1960 1970 1980 1990 (c) (d) Fic. 2. Temporal changes in major land use and land cover in Florida between 1936 and 1987. million ha in 1959. However since 1959, many previously cleared forest lands have been replanted to slash pines by the forest industry. As a result, slash pine forests increased to 2.14 million ha by 1970, and have remained fairly constant since then, covering 2.12 million ha in 1987. In 1987, slash pine was clearly the dominant forest tree in Florida, accounting for 69 percent of all commercial pine forests, 35 percent of all commercial forest land, and 15 percent of the land area of Florida. Loblolly pine and sand pine forests—Loblolly pine forests declined in area from 0.26 million ha (or 2 percent of the land area of Florida) in 1936 to 0.14 million ha (or 1 percent of the land area) in 1970 (Fig. 3d). However, since 1970, loblolly pine forests have steadily increased in area. By 1987 loblolly pine forests had increased to 0.23 million ha. In 1936 forests dominated by sand pines covered 0.16 million ha, or 1 percent, of the land area of Florida (Fig. 3d). Between 1936 and 1959 sand pine forests declined gradually to 0.15 million ha (Fig. 3d). However, since 1959, sand pine forest area has increased steadily. By 1987 sand pine had increased to 0.25 million ha, an No. 1, 1993] KAUTZ—TRENDS IN FLORIDA WILDLIFE HABITAT 1936-1987 13 10 7 ALL FOREST LANDS ALL PINE FORESTS 6 ~~ 8 ~~ ” ” § S5 = . Cc ial F: t = r ommercia oresis o 4 ea a aq 4 < © 53 Ww wi BS xr 2 e Noncommercial Forests 2 -" ALL HARDWOOD FORESTS 0 1 1930 1940 1950 1960 1970 1980 1990 1930 1940 1950 1960 1970 1980 tT990 (a) (b) 3.5 0.3 : LONGLEAF PINE LOBLOLLY PINE mm oO =o =a SLASH PINE § o2 = 2 Mire \aaih| ees = g en ale ae a 1.5 cc SAND PINE BE Ee oO oO w 1 wi z= Ir 0.5 0 0 1930 1940 1950 1960 1970 1980 1990: 1930° 1940 1950 1960 1970 1980 1990 (c) (d) 2 2.5 UPLAND HARDWOOD FORESTS OAK - GUM - CYPRESS = 2 2 1.5 2 = il 1.5 a 1 < c (6) = tl 5 xr 70.5 Oak-Pine 0.5 POND PINE 0 1930 1940 1950 1960 1970 1980 1990 (e) Le) 1930 1940 1950 1960 1970 1980 1990 (f) Fic. 3. Temporal changes in major forest types categorized as commercial forests in Florida between 1936 and 1987. area greater than at any other time in the 51-year study period. Oak-hickory and oak-pine forests—Upland hardwood (i.e., oak-hickory) forests covered 0).61 million ha, or 4 percent, of the land area of Florida in 1936 (Fig. 3e). Between 1936 and 1970, upland hardwood forests increased steadily to a high of 1.72 14 FLORIDA SCIENTIST [VOL 56 million ha, or 12 percent of the state. Since 1970, upland hardwood forest area has declined 27 percent, falling to 1.26 million ha, or 9 percent of the state, in 1987. In 1987 oak-hickory forests comprised 0.77 million ha, or 61 percent, and oak-pine forest comprised 0.49 million ha, or 39 percent, of the Florida’s upland hardwood forests. Lowland hardwood and pond pine forests—Lowland hardwood forests covered 1.30 million ha, or 9 percent of the land area of Florida, in 1936 (Fig. 3f). Lowland hardwood forests increased sharply to a high of 1.97 million ha, or 14 percent of the state, in 1959. By 1970, lowland hardwood forests had declined to approximately 1.74 million ha. Since 1970, lowland hardwood forests appear to have increased slightly, reaching 1.77 million ha by 1987. However, the apparent recent increase is within the range of sampling error reported by USFS and probably is not significant. In 1936 and 1949 pond pine wetlands covered only 0.15 million ha (or 1 percent) of the land area of Florida (Fig. 3f). However, since 1949, pond pine wetlands have decreased steadily. By 1987 they covered only 0.06 million ha. Although pond pine wetlands probably never occurred in great abundance in Florida, they declined 58 percent in the 38-year period between 1949 and 1987, disappearing at the rate of 2,400 ha per year. If this rate continues into the future, pond pine forests will be lost from Florida in 27 years, or by the year 2014. Forests on hydric sites—Changes in the area of commercial forests on hydric sites are indicative of changes in the area of forested wetlands in Florida. Unfortu- nately, data on the area of forest land on hydric sites are available only for the 1970, 1980, and 1987 surveys. In 1970, commercial forests on hydric sites covered 1.53 million ha of Florida and accounted for 23 percent of all commercial forest land (Fig. 4). Between 1970 and 1987, these forested wetlands declined to 1.37 million ha and accounted for 21 percent of all commercial forest land. This is a loss of 264,700 ha of forested wetlands in 17 years, a decline of 17 percent. The vast majority of this loss occurred between the 1980 and 1987 surveys. Population—Between 1830, the date of the first U.S. census in Florida, and 1936, the year of the first USFS forest survey, Florida’s population increased gradually from 34,730 to 1.73 million people (Fig. 5). The trend continued through 1950. Then, beginning around 1950, Florida’s population began to increase at a faster rate than before, and the higher rate of growth has persisted to the present day. Florida’s population reached 12.94 million in 1990, and population growth shows no signs of slowing. DiscussioN—Land Use and Land Cover 1936-1987—Forest land—The USFS data show that 1.75 million ha of forest land have disappeared from Florida since 1949. As shown (Fig. 5) forest lands began to be lost from the state at the same time that the rate of human population growth suddenly increased. The forest lands that were lost were converted to agricultural and urban uses. One aspect of changes in forest lands not discernable from the USFS data is that of changes in upland versus wetland forests. Fortunately, other information aids in separating upland forest from wetland area, at least for the latest survey. Kautz and co-workers (in review), using Landsat satellite imagery to inventory Florida vegeta- No. 1, 1993] KAUTZ—TRENDS IN FLORIDA WILDLIFE HABITAT 1936-1987 15 Lf FORESTED WETLANDS HECTARES (Millions) IN) oo b on ron) cook, — 1970 1980 1987 Fic. 4. Temporal changes in forested wetlands categorized as commercial forests in Florida between 1970 and 1987. USFS Forest Inventory Study Period ————B>] POPULATION (Millions) o fee) -_ N 0 : 1800 1850 1900 1950 2000 Fic. 5. Human population growth in Florida from the time of the first U.S. census projected through the year 2000. The shaded box shows population growth between 1936 and 1987, the period during which land cover and land use data were collected by the U.S. Forest Service. 16 FLORIDA SCIENTIST [VOL 56 tion types, recorded 2.03 million ha of forested wetlands in Florida based on 1985- 1989 data. If this number is accurate, then the total area of upland forest in Florida in 1987 was approximately 4.69 million ha. Marsh land—A few major drainage projects account for 24 percent of the 1.57 million ha marsh lands lost from Florida between 1936 and 1987. For example, 283,400 ha of marsh have been lost from the Everglades alone (VanArman et al., 1984). In addition, 81,000 ha of marsh land have been lost from the Upper St. Johns River basin (Campbell et al., 1984), 12,100 ha from the Kissimmee River system (Montalbano et al., 1979), and 6,500 ha along the north shore of Lake Apopka (Elert, 1990). These wetlands drainage projects have generally resulted in the wholesale destruction of major marsh land ecosystems. In each of these cases, marsh lands were converted to agricultural use. It seems likely that the remaining loss of marsh land resulted from extensive drainage of the wet prairies of central and southern Florida for conversion to improved pasture, although I have no substantiating data. The marsh land figures reported by USFS, at least since the 1970 survey, are comparable to those reported by others during this period. Hampson (1984) identified 1.21 million ha of marshes in the period 1972-1974 based on early satellite imagery. The Florida Department of Community Affairs estimated that there were 1.32 million ha of marsh in Florida in the 1982-1984 time period based on Landsat Thematic Mapper imagery (Robert Groce, unpublished data). Kautz and co-workers (1993) found 1.29 million ha of marsh lands in Florida (i.e., 1.10 million ha of freshwater marsh and 0.19 million ha of salt marsh) based on 1985-1989 satellite imagery. Although a tally of marsh area was not a high priority in the USFS inventories, it appears that the techniques USFS used for inventorying marsh lands were sound. Agriculture and rangeland—The majority of the increase in the agriculture and rangeland category since 1959 is attributable to large increases in rangeland (Fig. 2c). The dramatic increase in rangeland during this period was the result of widespread clearing of forest lands in central and southern Florida for cattle production. A recent decline of 0.27 million ha, or 10 percent, in the area of rangeland between 1980 and 1987 (Fig. 2c) probably indicates conversion of rangeland to urban areas. Forest Cover 1936-1987—Pine and hardwood forests—Although the overall downward trend in pine and hardwood forests is due to conversion to other land uses, some of this change can also be attributed to forest management practices and USFS definitions. McCormack (1950) and Larson and Goforth (1961) point out that, through 1960, pines were selectively logged from hundreds of thousands of hectares of predominantly pine forests, leaving only hardwoods. Also, many pine stands were clearcut but not replanted, and hardwoods were allowed to invade and dominate former pine sites. At the same time, widespread fire suppression allowed hardwoods to successfully invade many areas where frequent, naturally occurring fires had previously excluded them. The net result was conversion of some forest lands classified as pine types in the early surveys to hardwood sites in subsequent surveys. No. 1, 1993] KAUTZ—TRENDS IN FLORIDA WILDLIFE HABITAT 1936-1987 iY Thus, the apparent increase in hardwood forest area between 1936 and 1959 is in part a result of hardwood dominance on former pine sites in response to forest manage- ment practices during the period. Another confounding problem is that no distinction was made between pine and oak-pine forests in the 1936 and 1949 surveys. Rather, because of their economically valuable pine component, USFS stafflumped oak-pine forests with other pine types. However, in later surveys USFS recognized oak-pine forests as a distinct type, but decided that the oak-pine type more appropriately belonged in the hardwood category and so lumped them with hardwood types for reporting purposes. Thus, the precipitous drop in pine forests and dramatic increase in hardwood forests between 1936 and 1959 can be explained partially by USFS’s eventual recognition of oak-pine forests as a separate type and including them with hardwood rather than pine forests. Longleaf pine and slash pine forests—Historically, longleaf pine and slash pine have been the two most abundant and economically important trees in the Florida landscape. The fire-tolerant longleaf pine was the dominant tree of Florida’s pine forests prior to the advent of European man (Abrahamson and Hartnett, 1990; Myers, 1990). Longleaf pine occurred extensively in sandhill, clayhill, and flatwoods communities where frequent, lightning-set fires eliminated hardwoods and other species of pines from competition. On the other hand, slash pine, a species less tolerant of frequent fire, was relegated to low areas, wetlands, and scattered uplands that were naturally protected from fires. As discussed previously, the total area of Florida covered by pines of all types has declined markedly since 1936, and previously dominant longleaf pine forests have all but disappeared. On the other hand, the total area of slash pine forest has remained roughly constant, largely as a result of conversion of vast areas of mature pine forests to young, even-aged plantations. It is well documented that the wildlife habitat values of plantations are inferior to those of natural pine stands (Harris et al., 1974; Umber and Harris, 1974; Kautz, 1984; Repenning and Labisky, 1985; McComb et al., 1986). Hence, the net result is that a large percentage of Florida’s remaining pine forests now provide poor quality habitat for many formerly abundant species of wildlife. Bechtold and co-workers (1990) provide data on the status of pine plantations in Florida in 1987. This data serves as an overall index of the wildlife habitat value of Florida’s remaining pine forests. USFS defines “pine plantations” as “stands that have been artificially regenerated by planting or direct seeding and with a southern yellow pine, white pine-hemlock, or other softwood forest type.” USFS defines “natural pine” forests as “stands that have not been artificially regenerated and with a southern yellow pine, white pine-hemlock, or other softwood forest type.” Bechtold and co-workers (1990) reported that in 1987 1.63 million ha, or 53 percent, of Florida’s pine forests were in plantations. This amounts to 27 percent of all commercial forest land in Florida. Approximately 1.30 million ha, or 63 percent, of all slash pine forests are in pine plantations. On the other hand, 0.35 million ha, or 91 percent, of Florida’s remaining longleaf pine forests are in natural stands, and only 36,400 ha of longleaf pines are in plantations. 18 FLORIDA SCIENTIST [VOL 56 The remaining area of mature pine forest is of particular interest to the conservation of the red-cockaded woodpecker (Picoides borealis) in Florida. The red-cockaded woodpecker is listed by the Florida Game and Fresh Water Fish Commission as a threatened species and by the U.S. Fish and Wildlife Service as an endangered species. The survival of the red-cockaded woodpecker is in jeopardy because this species nests almost exclusively in longleaf and slash pines greater than 70 years old (Lennartzet al., 1983), and most mature pine timber has long since been cut. Bechtold and co-workers (1990) report that in 1987 there were no pine plantations in Florida over 70 years of age and that natural pine stands older than 70 years totaled only 30,800 ha. Wood (1983) showed that the mean diameter breast height (dbh) and age of red-cockaded woodpecker cavity trees in longleaf pines were 39.4 cm and 86 years, respectively, and in slash pines were 40.6 cm and 70 years, respectively. Bechtold and co-workers (1990) report that in 1987 there were a total of 4,594 longleaf pines and 10,188 slash pines with a dbh greater than 38 cm in all of Florida. Thus, in 1987 the entire future of the red-cockaded woodpecker in Florida depended on 15,000 mature pine trees growing on 30,800 ha of forest land. Loblolly pine and sand pine forests—Where they occur, loblolly pine and sand pine tend to be canopy dominants, but forests dominated by these pines never have covered extensive areas of Florida. In many respects, the changes in loblolly pine and sand pine forest area parallel those of slash pine forests. Loblolly pine is a relatively fast-growing species that readily invades old fields in the northern half of the state. Between 1936 and 1970 loblolly pines declined gradually as pine forests were cleared but not replanted. However, by 1970, loblolly pine plantations began to appear with increasing frequency on the clayhills of northern Florida. As a consequence the total area of loblolly pine forests in Florida has steadily increased since 1970. By 1987, 0.13 million ha, or 55 percent, of all loblolly pines in Florida were in short-rotation plantations. Historically, sand pines occurred as the dominant canopy tree only in Florida's rare scrub communities, which are found on nutrient-poor, excessively drained soils of ancient sand dunes. Scrub communities are of special biological importance because they support a variety of plants and animals that occur nowhere else in the world (Myers, 1990). Harlow and Jones (1965) reported that sand pine and oak scrubs covered only 0.21 million ha (or 1.5 percent) of pre-settlement Florida. In the past, sand pines have been of little interest to foresters because they exhibit poor form. Sand pine forests were in a state of gradual decline through 1959 (Fig. 3d). However, in the 1960s the forest industry began planting dense stands of sand pines on sandhill sites because they grow faster than either longleaf or slash pines on such sandy, well-drained soils and because they have value as pulpwood. Thus, by 1987, the area of sand pine forest in Florida was greater than at any other time in the 51- year study period. Bechtold and co-workers (1990) report that 0.13 million ha, or 52 percent, of all stands of sand pine were in pine plantations in 1987. Between 1936 and 1959, Florida lost an average of 700 ha of sand pine scrub each year, largely a result of conversion to citrus groves and urban areas. Since 1959, No. 1, 1993] KAUTZ—TRENDS IN FLORIDA WILDLIFE HABITAT 1936-1987 19 the scrub community has continued to disappear from the state, but the rate of loss of the scrub community has been masked by increases in the area of sand pine due to commercial plantings on sandhill sites. If the rate of loss observed between 1936 and 1959 can be used to estimate the current status of the scrub community in Florida, native sand pine habitats would have declined to approximately 126,500 ha by 1987. This estimate compares favorably with the 117,100 ha of sand pine that Bechtold and co-workers (1990) report was in natural stands in 1987. Since approximately 84,600 ha of all the sand pine forest in Florida occurs on the Ocala National Forest, these numbers suggest that no more than 41,900 ha of sand pine scrub exist in Florida outside of Ocala National Forest. If the rate of loss that existed between 1936 and 1959 remained constant into the future, all sand pine scrub outside of the Ocala National Forest would be eliminated from Florida in 60 years, or by the year 2047. In reality, the rate of loss of scrub habitat in Florida is undoubtedly much higher than 700 ha per year, as it is well known that scrubs are one of the most rapidly disappearing community types in Florida (Myers, 1990). Oak-hickory and oak-pine forests—Together, oak-hickory and oak-pine forests account for most of Florida’s upland hardwood forests. The major Florida plant community types represented by USFS’s oak-hickory and oak-pine forest classes are hardwood hammocks, the mixed pine-hardwood forests of the panhandle, live oak (Quercus viginiana)-cabbage palm (Sabal palmetto) hammocks in the southern portion of the peninsula, and the tropical hardwood hammocks of the Florida Keys. There was a rapid increase in the area of upland hardwood forests between 1936 and 1970 (Fig. 3e). As discussed earlier, this increase was largely due to young hardwood forests invading pine sites that were logged but not replanted. To a lesser extent, changes in USFS views of the oak-pine class also influenced the rapid increase in upland hardwood forest area through 1970. Since 1970, upland hardwood forests have declined 27 percent, probably because upland hardwood forests are being replaced with urban areas and converted to pine plantations. Lowland hardwood and pond pine forests—The area of lowland hardwood forest increased markedly between 1936 and 1959 (Fig. 3f), an unexpected result. It seems most likely that this increase is another case of the influence silvicultural practices have had on the forest types reported by USFS. The USFS definition of oak-gum-cypress (i.e., lowland hardwood) forests indicates that pine trees may be present in this type. It also states that, if pines constitute 25-50 percent of stocking, a stand would be classified as oak-pine rather than oak-gum-cypress. Inall likelihood, the apparent increase in the area of lowland hardwoods is probably the result of the selective removal of pines (probably slash pines) such that the remaining forests were dominated by hardwoods and were therefore reported as lowland hardwood forests in later surveys. Since 1970, however, lowland hardwood forest area has remained roughly constant, suggesting that few if any loggable pines remain in lowland forests. Forests dominated by pond pine typically occur in low, wetland sloughs interspersed within the flatwoods community. Since 1949, pond pine wetlands have decreased steadily, largely due to the silvicultural practices of draining flatwoods and bedding in low areas to make more land available for the commercial production of 20 FLORIDA SCIENTIST [VOL 56 slash pines. Between 1949 and 1987 pond pine forests disappeared from Florida at the rate of 2,400 ha per year. If this rate of loss continues into the future, pond pine forests will disappear from Florida in 27 years, or by the year 2014. Forests on hydric sites—The lowland hardwood forest type includes both upland and wetland forests and cannot be used to estimate area of forested wetlands. However, since the 1970 survey, USFS categorized all commercial forest stands according to physiographic class, largely on the basis of soil moisture conditions. The USFS definition for the hydric physiographic class is similar to the Cowardin and co- workers (1979) definition of wetland. Therefore, changes in the area of commercial forest land on hydric sites are assumed to be indicative forested wetlands trends in Florida. As noted previously, Kautz and co-workers (in review) reported that there were 2.03 million ha of forested wetlands in Florida based on 1985-1989 satellite data. Assuming that their figure is correct, the 1.27 million ha of forested wetlands on commercial forest land accounted for 62 percent of all forested wetlands in Florida in 1987. The remaining 0.76 million ha were on reserved or unproductive sites. With such a large area of forested wetlands on reserved or unproductive sites, it is tempting to suggest that the loss of forested wetlands may be attributed to the reclassification of the lost area from commercial forest to reserved forest as a result of extensive land acquisitions by public agencies. However, a review of the USFS reports indicates that the loss of 264,671 ha of forested wetlands between 1970 and 1987 is an actual decline in forested wetland area and not an artifact of reclassifica- tion. Population—To provide habitat for growth of the human population in Florida between 1936 and 1987, it has been necessary to convert millions of ha of productive wildlife habitat to other uses, placing the survival of many species in doubt. Since the current rate of growth of the human population is expected to persist into the foreseeable future (Duda, 1987), Floridians can expect continued erosion of the wildlife habitat base and increasing jeopardy for more and more species. ACKNOWLEDGMENTS—I would like to thank Noel Cost for generously providing unpublished Forest Service data used in this paper and for reviewing an earlier draft of the manuscript. Jim Cox, Perry Oldenburg, David Kelly, Brad Hartman, and Mark Kopeny also provided comments on previous drafts. I am especially grateful to the Florida Game and Fresh Water Fish Commission and the Nongame Wildlife Program for supporting this work. LITERATURE CITED ABRAHAMSON, W. G. anD D. C. HARTNETT. 1990. Pine flatwoods and dry prairies. Pp. 103-149. In: MYERS, R. L. AND J. J. EWEL (eds.), Ecosystems of Florida. University of Central Florida Press, Orlando, FL. ANDERSON, J. R. JR. AND E. A. FERNALD. 1981. Introduction. Pp. 1-3. In: FERNALD, E. A. (ed.), Atlas of Florida. Florida State Univ. Foundation, Inc., Tallahassee, FL. BECHTOLD, W. A. AND H. A. KNIGHT. 1982. Florida’s forests. U.S.D.A. Forest Service, Southeastern Forest Experiment Station, Resource Bull. SE-62, Asheville, NC. 48 pp. .. M. J. BROWN anD R. M. SHEFFIELD. 1990. Florida’s forests, 1987. U.S.D.A. Forest Service, Southeastern Forest Experiment Station, Resource Bull. SE-110, Asheville, NC. 83 pp. BROWN, M. J. aND M. T. THOMPSON. 1988. Forest statistics for Florida, 1987. U.S.D.A. Forest Service, No. l, 1993] KAUTZ—TRENDS IN FLORIDA WILDLIFE HABITAT 1936-1987 WA \\ Southeastern Forest Experiment Station, Resource Bull. SE-101, Asheville, NC. 61 pp. CAMPBELL, D., D. A. MUNCH, R. JOHNSON, M. P. PARKER, B. PARKER, D. V. RAO, R. MARELLA AND E. ALBANESI. 1984. St. Johns River Water Management District. Pp. 158-177. In: FERNALD, E. A. AND D. J. Patron (eds.), Water Resources Atlas of Florida. Florida State Univ., Tallahassee, FL. COWARDIN, L. M., V. CARTER, F. C. GOLET and E. T. LAROE. 1979. Classification of wetlands and deepwater habitats of the United States. U.S. Fish and Wildlife Service, FWS/OBS-79/31, Washington, D.C. 103 pp. Cox, J. 1987. The breeding bird survey in Florida: 1969-1983. Florida Field Natural. 15(2):29-56. . 1988. The influence of forest size on transient and resident bird species occupying maritime hammocks on northeastern Florida. Florida Field Natural. 16(2):25-56. Dupa, M. D. 1987. Floridians and wildlife: sociological implications for wildlife conservation in Florida. Florida Game and Fresh Water Fish Commission, Nongame Wildlife Program Technical Report No. 2, Tallahassee, FL. 130 pp. EHRHABRT, L. M. 1978. Recently extirpated plains bison. P. 48. In: LAYNE, J. N. (ed.), Rare and Endangered Biota of Florida. Volume One. Mammals. University Presses of Florida, Gainesville, FL. ELDREDGE, I. F. 1938a. Forest resources of northwest Florida, a progress report. U.S.D.A. Forest Service, Southern Forest Experiment Station, Forest Survey Release No. 33, New Orleans, LA. 33 pp. F OAD. Forest resources of central and south Florida, a progress report. U.S.D.A. Forest Service, Southern Forest Experiment Station, Forest Survey Release No. 38, New Orleans, LA. 30 pp. ELERT J. 1860. Florida Game and Fresh Water Fish Commission. Tallahassee. Pers. Comm. GALLI, A. E., C. F. LECK anp R. T. T. FORMAN. 1976. Avian distribution patterns.in forest islands of different sizes in central New Jersey. Auk 93:356-364. HAMPSON, P. S. 1984. Wetlands in Florida. Florida Department of Natural Resources, Bureau of Geology, Map Series No. 109, Tallahassee, FL. 1 p. . HARDY, J. W. 1978a. Recently extinct Carolina parakeet. P. 120. In: KALE, H. W., II (ed.), Rare and Endangered Biota of Florida. Volume Two. Birds. University Presses of Florida, Gainesville, FL. . 1978b. Extinct passenger pigeon. Pp. 120-121. In: KALE, H. W., II (ed.), Rare and Endangered Biota of Florida. Volume Two. Birds. University Presses of Florida, Gainesville, FL. HARLOW, R. F. AND F. K. JONES, JR. 1965. The white-tailed deer in Florida. Florida Game and Fresh Water Fish Commission, Technical Bull. No. 9, Tallahassee, FL. 240 pp. HARRIS, L. D. AND R. D. WALLACE. 1984. Breeding bird species in Florida forest fragments. Proc. Annu. Conf. Southeast. Assoc. Fish and Wildl. Agencies 38:87-96. .. L. D. WHITE, J. E. JOHNSTON AND D. G. MILCHUNAS. 1974. Impact of forest plantations on North Florida wildlife and habitat. Proc. Annu. Conf. Southeast. Assoc. Game and Fish Comm. 28:659-667. HUMPHREY, S. R. 1981. Goff’s pocket gopher (Geomys pinetis goffi) is extinct. Florida Scient. 44(4):250- 252. and D. B. BARBOUR. 1981. Status and habitat of three subspecies of Peromyscus polionotus in Florida. J. Mamm. 62(4):840-844. INESON, F. A. AND I. F. ELDREDGE. 1938. Forest resources of northeastern Florida. U.S.D.A. Forest Service, Southern Forest Experiment Station, Miscellaneous Publ. No. 313, Washington, D.C. 40 Pp. KAUTZ, R.S. 1984. Criteria for evaluating impacts of development on wildlife habitats. Proc. Annu. Conf. Southeast. Assoc. Fish and Wildl. Agencies 38:121-136. , D. T. GILBERT AND G. M. MAULDIN. 1993. Mapping Florida Wildlife Habitats using Landsat Thematic Mapper imagery. Florida Scient. In Press. KLIMSTRA, W. D. AND J. W. HARDIN. 1978. Endangered Key deer. Pp. 15-17. In: LAYNE, J. N. (ed.), Rare and Endangered Biota of Florida. Volume One. Mammals. University Presses of Florida, Gainesville, FL. KNIGHT, H. A. AND J. P. MCCLURE. 1971. Florida’s timber, 1970. U.S.D.A. Forest Service, Southeastern Forest Experiment Station, Resource Bull. SE-20, Asheville, NC. 48 pp. LARSON, R. W. AND M. H. GOFORTH. 1961. Florida’s timber. U.S.D.A. Forest Service, Southeastern Forest Experiment Station, Forest Survey Release No. 57, Asheville, NC. 32 pp. LAYNE, J. N. 1978. Recently extinct Florida red wolf. Pp. 48-49. In: LAYNE, J. N. (ed.), Rare and Endangered Biota of Florida. Volume One. Mammals. University Presses of Florida, Gainesville, FL. 29 FLORIDA SCIENTIST [VOL 56 LENNARTZ, M. R., H. A. KNIGHT, J. P. MCCLURE AND V. A. RUDIS. 1983. Status of red-cockaded woodpecker nesting habitat in the South. Pp. 13-19. In: Woop, D. A. (ed.), Red-cockaded Woodpecker Symposium II, Proceedings, Florida Game and Fresh Water Fish Commission, Tallahassee, FL. MATTHIAE, P. E. aND F. STEARNS. 1981. Mammals in forest islands in southeastern Wisconsin. Pp. 55- 66. In: BURGESS, R. L., anD D. M. SHARPE (eds.), Forest Island Dynamics in Man-Dominated Landscapes. Springer-Verlag, New York, NY. McComB, W.C., S. A. BONNEY, R. M. SHEFFIELD AND N. D. COST. 1986. Den tree characteristics and abundance in Florida and South Carolina. J. Wildl. Manage. 50(4):584-591. McCorMACcK, J. F. 1950. Forest statistics for Florida, 1949. U.S.D.A. Forest Service, Southeastern Forest Experiment Station, Forest Survey Release No. 36, Asheville, NC. 73 pp. MILLSAP, B. A., J. A. GORE, D. E. RUNDE AND S. I. CERULEAN. 1990. Setting priorities for the conservation of fish and wildlife species in Florida. Wildl. Monogr. No. 111. 57 pp. MONTALBANO, F., III, K. J. FOOTE, M. W. OLINDE AND L. S. PERRIN. 1979. The Kissimmee River channelization: a preliminary evaluation of fish and wildlife mitigation measures. Pp. 508-515. In: SWANSON, G. A. (tech. coord.), The Mitigation Symposium: A National Workshop on Mitigating Losses of Fish and Wildlife Habitats. U.S.D.A. Forest Service, Gen. Tech. Rep. RM-65, Rocky Mountain Forest and Range Experiment Station, Ft. Collins, CO. MYERS, R. L. 1990. Scrub and high pine. Pp. 150-193. In: MYERS, R. L. and J. J. EWEL (eds.), Ecosystems of Florida. Univ. of Central Florida Press, Orlando, FL. OGDEN, J. C. 1978. Endangered American crocodile. Pp. 21-22. In: MCDIARMID, R. W. (ed.), Rare and Endangered Biota of Florida. Volume Three. Amphibians and Reptiles. University Presses of Florida, Gainesville, FL. REPENNING, R. W. and S. R. HUMPHREY. 1986. The Chadwick Beach cotton mouse (Rodentia: Peromyscus gossypinus restrictus) may be extinct. Florida Scient. 49(4):259-262. AND R. F. LABISKY. 1985. Effects of even-age timber management on bird communities of the longleaf pine forest in northern Florida. J. Wildl. Manage. 49(4): 1088-1098. ROBBINS, C. S. 1979. Effect of forest fragmentation on bird populations. Pp. 198-212. In: DEGRAAF, R. M. AND K. E. EVANS (eds.), Workshop proceedings: management of North Central and Northeast- ern forests for nongame birds. U.S.D.A. Forest Service, Gen. Tech. Rep. NC-51, St. Paul, MN. TEBEAU, C. W. 1971. A history of Florida. University of Miami Press, Coral Gables, FL. 527 pp. UMBER, R. W. anp L. D. HARRIS. 1974. Effects of intensive forestry on succession and wildlife in Florida sandhills. Proc. Annu. Conf. Southeast. Assoc. Game and Fish Comm. 28:686-693. U.S. FISH AND WILDLIFE SERVICE. 1987. Florida panther (Felis concolor coryi) recovery plan. Prepared by the Florida Panther Interagency Committee for the U.S. Fish and Wildlife Service, Atlanta, GA. 75 pp. U.S. FISH AND WILDLIFE SERVICE. 1989. Endangered Species Tech. Bull. 14(8):10-11. VAN ARMAN, J: D. NEALON, S. BURNS, B. JONES, L. SMITH, T. MACVICAR, M. YANSURA, A. FEDERICO, J. Bucca, M. KNAPP, AND P. GLEASON. 1984. South Florida Water Management District. Pp. 138- 157. In: FERNALD, E. A., AND D. J. PATTON (eds.), Water Resources Atlas of Florida. Florida State University, Tallahassee, FL. WHITCOMB, R. F., C.S. ROBBINS, J. F. LYNCH, B. L. WHITCOMB, M. K. KLIMKIEWICZ and D. BYSTRAK. 1981. Effects of forest fragmentation on avifauna of the eastern deciduous forest. Pp. 125-206. In: BURGESS, R. L. AND D. M. SHARPE (eds.), Forest Island Dynamics in Man-Dominated Landscapes. Springer-Verlag, New York, NY. Woop, D. A. 1983. Foraging and colony habitat characteristics of the red-cockaded woodpecker in Oklahoma. Pp. 51-58. In: Woop, D. A. (ed.), Red-cockaded Woodpecker Symposium II, Proceedings, Florida Game and Fresh Water Fish Commission, Tallahassee, FL. . 1990. Official lists of endangered and potentially endangered fauna and flora in Florida. Florida Game and Fresh Water Fish Commission, Tallahassee, FL. 19 pp. GLossaRY. U.S. FOREST SERVICE DEFINITIONS OF MAJOR LAND USEAND FOREST TyPES INVENTORIED IN FLORIDA (Brown and Thompson, 1988). No. 1, 1993] KAUTZ—TRENDS IN FLORIDA WILDLIFE HABITAT 1936-1987 23 MAJOR LAND COVER AND LAND USE Forest land—Land at least 16.7 percent stocked by forest trees of any size, or formerly having had such tree cover, and not currently developed for nonforest use. Marsh—Low, wet areas characterized by a heavy growth of grass and reeds and an absence of timber (McCormack 1950). Agriculture (or Cropland and Rangeland)—The following USFS types were summed to yield an estimate of area in agricultural use: Cropland—Land under cultivation within the past 24 months, including or- chards and land in soil-improving crops but excluding land cultivated in developing improved pasture. Also includes idle farmland. Idle farmland—Land including former cropland, orchard, improved pasture, and farm sites not tended within the past 2 years, and currently less than 16.7 percent stocked with live trees. Rangeland—Land on which the natural vegetation is predominantly native grasses, grasslike plants, forbs, or shrubs valuable for forage, not qualifying as timber land and not developed for another use. Rangeland includes natural grassland and savannah. Improved pasture—Land currently improved for grazing by cultivation, seed- ing, irrigation, or clearing of trees or brush. Urban and other areas—Areas developed for residential, industrial, or recreational purposes, school yards, cemeteries, roads, railroads, airports, beaches, powerlines and other rights-of-way, or other nonforest land not included in any other specified land use class. FOREST RESOURCES Commercial forest land (Timberland)—Land at least 16.7 percent stocked by forest trees of any size, or formerly having had such tree cover, not currently developed for nonforest use, capable of producing 0.23 cubic meters of industrial wood per hectare (20 ft? per acre) and not withdrawn from timber utilization by legislative action. Noncommercial forest land—Land comprised of the following USFS forest types. Reserved timberland—Forest land sufficiently productive to qualify as timber- land, but withdrawn from timber utilization through statute or administrative designation. Woodland—Forest land incapable of producing 0.23 cubic meters per hectare per year of industrial wood under natural conditions, because of adverse site conditions. Pine forests—An estimate of the area of Florida in forests dominated by pines was developed by summing the following USFS forest types: 94 FLORIDA SCIENTIST [VOL 56 Longleaf pine—Forests in which longleaf pine (Pinus palustris) constitutes a plurality of the stocking. Slash pine—Forests in which slash pine (P. elliottii) constitutes a plurality of the stocking. Loblolly pine—Forests in which loblolly pine (P. taeda) constitutes a plurality of the stocking. Sand pine—Forests in which sand pine (P. clausa) constitutes a plurality of the stocking. Pond pine—Forests in which pond pine (P. serotina) constitutes a plurality of the stocking. Hardwood forests—An estimate of the area of Florida in forests dominated by hardwoods was developed by summing the following USFS forest types: Oak-hickory—Forests in which upland oaks (Quercus spp.) or hickory (Carya spp.), singly or in combination, constitute a plurality of the stocking, except where pines account for 25 to 50 percent, in which case the stand would be classified oak-pine. Oak-pine—Forests in which hardwoods (usually upland oaks) constitute a plurality of the stocking but in which pines account for 25 to 50 percent of the stocking. Oak-gum-cypress—Bottomland forests in which tupelo (Nyssa aquatica), blackgum (N. biflora) sweetgum (Liquidambar styraciflua), oaks, or southern cypress (Taxodium distichum, T. ascendens), singly or in combination, consti- tute a plurality of the stocking, except where pines account for 25 to 50 percent, in which case the stand would be classified oak-pine. (USFS also refers to this type as lowland hardwoods). Palm, other tropical—Forests in which palms and other tropicals constitute a plurality of the stocking. Florida Scient. 56(1): 7-24.1993 Accepted: June 5, 1992. No. 1, 1993] 25 Biological Sciences THE ECOLOGICAL BASIS OF THE KISSIMMEE RIVER RESTORATION PLAN Louis A. TOTH Division of Kissimmee & Okeechobee Systems Research, Department of Research, South Florida Water Management District, P.O. Box 24680, 3301 Gun Club Road, West Palm Beach, F lorida 33416-4680 ABSTRACT: This review synthesizes over 40 years of studies on the ecological resources of the Kissimmee River. Prior to 1962 the Kissimmee River ecosystem supported diverse fish and wildlife populations including waterfowl, wading birds, and a nationally recognized fishery. The historic floodplain consisted of a mosaic of broadleaf marsh, shrub and prairie wetland communities. Between 1962 and 1970 the river was channelized and transformed into a series of impounded reservoirs. The physical impacts of channelization, including alteration of the system’s unique hydrologic characteristics, largely eliminated the wetland and fish and wildlife values of the river and floodplain. Attendant “restoration” studies, including the recently completed demonstration project which documented habitat, water quality, avian, fish, and invertebrate responses to water level manipulations, reestablished floodplain inundation, and reintroduced flow, confirmed the feasibility of restoring both the structure and functions of the historic Kissimmee River ecosystem. The integration of available data on pre-channelization resources, impacts of channelization, and restoration-related studies forms the basis of the current plan to restore the ecological integrity of the Kissimmee River. ENVIRONMENTAL resources of Florida’s Kissimmee River basin (Fig. 1) have been studied for over 40 years. The primary impetus for these efforts has been the Central and Southern Florida Flood Control Project (U.S. Army Corps of Engineers, 1956) which resulted in channelization of the river. A river restoration plan has been developed from the integration of studies of pre-channelization ecology, environ- mental impacts of channelization, and theoretical and empirical analyses of the ecological benefits of various restoration proposals (Loftin et al., 1990b). This paper reviews the ecological basis of the Kissimmee River restoration plan, with emphasis on hydrology, floodplain vegetation, and avian and fish communities. Included are key results of a recently completed demonstration project which documented the feasibility of reestablishing both structural and functional aspects of river ecosystem integrity. HISTORICAL BIOLOGICAL RESOURCES AND ECOLOGY—Prior to channelization the Kissimmee river/floodplain ecosystem supported a diverse complement of fish, wildlife and wetlands. Although biological characteristics, as well as other structural and functional aspects of the pre-channelization ecosystem, were shaped by com- bined effects of numerous physical, chemical and biological processes, the system’s unique hydrological characteristics were a prominent driving force in the ecosystem (Toth, 1990b). The pre-channelization floodplain was a haven for avian wildlife (Table 1). White ibis (Eudocimus albus) was the most numerous wading bird species, but 10 26 FLORIDA SCIENTIST [VOL 56 Ses — a a a KISSIMMEE a u I B 6 ees aae ASK Ee | | OKEECHOBEE 0 8 16 KILOMETERS Fic. 1. Map of Kissimmee River basin showing C-38, S-65 water control structures and headwater chain of lakes. other species were common: great blue herons (Ardea herodias), great egrets (Egretta alba), snowy egrets (E. thula), Louisiana herons (E. tricolor), little blue herons (E. caerulea), green herons (Butorides virescens), glossy ibis (Plegadis falcinellus), limpkins (Aramus guarauna), sandhill cranes (Grus canadensis), and black-crowned night herons (Nycticorax nycticorax)(Perrin et al., 1982). The Kissimmee River basin’s lakes, sloughs and marshes also were important habitat for No. 1, 1993] TOTH—THE ECOLOGICAL BASIS OF THE KISSIMMEE RIVER RESTORATION PLAN PAT TABLE 1. List of waterfowl and wading bird species found in the pre-channelization (1949-57) Kissimmee River ecosystem (from Florida Game and Fresh Water Fish Commission, 1957). Waterfowl Green-winged teal Anas crecca American wigeon Anas americana American black duck Anas rubripes Mottled duck Anas fulvigula Blue-winged teal Anas discors Gadwall Anas strepera Mallard Anas platyrhynchos Northern shoveler Anas clypeata Pintail Anas acuta Wood duck Aix sponsa Canvasback Aythya valisineria Redhead Aythya americana Ring-necked duck Aythya collaris Scaup Aythya affinis Bufflehead Bucephala albeola Common goldeneye Bucephala clangula Hooded merganser Mergus cucullatus Red-breasted merganser Mergus serrator Ruddy duck Oxyura jamaicensis Wading Birds White ibis Eudocimus albus Glossy ibis Plegadis falcinellus Great egret Egretta alba Snowy egret Egretta thula Louisiana heron Egretta tricolor Little blue heron Egretta caerulea Reddish egret Egretta rufescens Great blue heron Ardea herodias Great white heron Ardea occidentalis Green heron Butorides virescens Wood stork Mycteria americana Black-crowned night heron Nycticorax nycticorax Yellow-crowned night heron Nyctanassa violacea Limpkin Aramus guarauna Sandhill crane Grus canadensis American bittern Botaurus lentiginosus Least bittern Ixobrychus exilis wintering waterfowl populations (Florida Game and Fresh Water Fish Commission, 1957). Approximately 20,000-25,000 waterfowl, including 19 species, typically utilized the basin from November-March, and about 20% of these wintering populations occurred on the wetlands of the river/floodplain (U.S. Fish and Wildlife Service, 1959). The river and floodplain supported at least 39 fish species (Table 2) which typically were concentrated in littoral river habitat, adjacent backwater sloughs, and floodplain marsh (Florida Game and Fresh Water Fish Commission, 1957). Marsh 28 FLORIDA SCIENTIST [VOL 56 TABLE 2. List of fish species found in the pre-channelization (1956-57) Kissimmee River ecosystem (from Florida Game and Fresh Water Fish Commission, 1957). Redfin pickerel Chain pickerel Redbreast sunfish Warmouth Bluegill Redear sunfish Spotted sunfish Largemouth bass Black crappie Everglades pygmy sunfish Bluespotted sunfish Dollar sunfish White catfish Yellow bullhead Brown bullhead Tadpole madtom Channel catfish Bowfin American eel Gizzard shad Lake chubsucker Florida gar Pirate perch Threadfin shad Blackbanded darter Swamp darter Golden topminnow Seminole killifish Mosquitofish Least killifish Flagfish Sailfin molly Bluefin killifish Brook silversides Tidewater silversides Golden shiner Pugnose minnow Taillight shiner Coastal shiner Esox americanus Esox niger Lepomis auritus Lepomis gulosis Lepomis macrochirus Lepomis microlophus Lepomis punctatus Micropterus salmoides Pomoxis nigromaculatus Elassoma evergladei Enneacanthus gloriosus Lepomis marginatus Ictalurus catus Ictalurus natalis Ictalurus nebulosus Noturus gyrinus Ictalurus punctatus Amia calva Anguilla rostrata Dorosoma cepedianum Erimyzon sucetta Lepisosteus platyrhincus Aphredoderus sayanus Dorosoma petenense Percina nigrofasciata Etheostoma fusiforme Fundulus chrysotus Fundulus seminolis Gambusia affinis Heterandria formosa Jordanella floridae Poecilia latipinna Lucania goodei Labidesthes sicculus Menidia beryllina Notemigonus crysoleucas Notropis emiliae Notropis maculatus Notropis petersoni habitats had large numbers of small fish, including young game fish, illustrating that the floodplain provided important production areas for forage fish and nursery habitat for game fish species (Miller, 1990). Despite extremely limited access for fisherman, the historic Kissimmee River had a nationally renowned largemouth bass (Micropterus salmoides) fishery (U.S. Fish & Wildlife Service, 1959) and two other game fish species, bluegill (Lepomis macrochirus) and black crappie (Pomoxis nigromaculatus), were abundant (Florida Game and Fresh Water Fish Commission, 1957). The floodplain was covered primarily by three vegetation community types: 1) willow (Salix caroliniana) and buttonbush (Cephalanthus occidentalis) woody shrub wetlands, 2) broadleaf marshes dominated by Sagittaria lancifolia and Pontederia No. 1, 1993] TOTH—THE ECOLOGICAL BASIS OF THE KISSIMMEE RIVER RESTORATION PLAN 29 TABLE 3. Wetland communities of the pre-channelization (1952-54) Kissimmee River ecosystem (from Pierce et al., 1982). Community Type Area (ha) River Channel/Open Water 1977 Broadleaf Marsh 8665 Wet Prairie Maidencane 1468 Mixed species 2450 Wetland Shrub Willow 746 Buttonbush 1319 Wetland Forested Cypress 89 Other Hardwood 50 Switchgrass 238 Other Wetland 240 Total Wetland e238 cordata, and 3) maidencane (Panicum hemitomon), beakrush (Rhynchospora inundata) and mixed species wet prairies (Pierce et al., 1982)(Table 3). The distribution of these communities generally reflected inundation characteristics along an increasing elevation gradient from the river channel to the periphery of the floodplain; broadleaf marsh and wetland shrub communities occurred in perma- nently and frequently flooded areas adjacent the river, while wet prairie was dominant along the drier, peripheral margins of the floodplain. This generalized distributional pattern of plant communities across the floodplain was modified by natural levees, abandoned river channels (oxbows), cattle and agricultural activity, dikes and drainage projects, the presence of other minor plant community types, the widely meandering river channel, hydrologic inputs from tributary sloughs, and the highly variable species composition of the broadleaf marsh, wetland shrub and wet prairie communities. As a result of these factors, the pre-channelization floodplain was composed of a mosaic of hundreds of distinct patches of intermingled vegetation types. Most of these pre-channelization studies indicate hydrologic features greatly influenced biological characteristics of the ecosystem. The key hydrologic determi- nants of pre-channelization ecology (Toth, 1990b) included several unique charac- teristics, such as continuous river discharge, frequent overbank flow, slow stage recession rates, and prolonged floodplain inundation (Table 4). IMPACTS OF CHANNELIZATION—The Kissimmee River was channelized between 1962 and 1970 as part of a Federally authorized flood control project. As a result of 30 FLORIDA SCIENTIST [VOL 56 TABLE 4. Key characteristics of pre-channelization Kissimmee River hydrology. Stage and discharge regimes with high year-to-year variability Continuous river flow Highest discharge during late summer and fall Lowest discharge from January through June Average channel velocities between 0.25 and 0.6 m sec’ Frequent overbank flow Slow rates of drainage and water level decline on the floodplain Prolonged flooding on most (approximately 80%) of the floodplain but seasonal wet-dry cycles along peripheral elevations Water depths on the floodplain that typically ranged between 0.3 - 0.6 m but > 0.9 m near the river channel channelization, most of the river flow formerly carried by the 166 km meandering channel and its 1.5 - 3 km wide floodplain now is confined to a 90 km long, 9 m deep, and 64 - 105 m wide canal (C-38). Beginning at the outlet of Lake Kissimmee, six water control structures (S-65, S-65A, S-65B, S-65C, S-65D, and S-65E) with tieback levees were constructed along the length of the canal. These structures and levees transformed the contiguous river/floodplain ecosystem into five stair-step impoundments or pools (Fig. 2). Channelization physically impacted the Kissimmee River ecosystem in two ways: (1) through destruction of large portions of the river channel and floodplain and, (2) through elimination of the system’s unique hydrologic characteristics (Toth, 1990a). Excavation of the canal and deposition of spoil had the most direct effect, obliterating approximately 56 km of river channel (Toth, 1990a) and 2,800 ha of C-—38 IMPOUNDMENTS 15.9 GROUND ELEVATION 13.4 11.0 8.5 ELEVATION (m — N.G.V.D.) a 6.1 EXISTING WATER LEVEL LIMIT OF IMPOUNDED MARSHLAND UNDER EXISTING STABILIZED WATER LEVELS 3.7 8 16 24 32 40 48 56 64 72 LINEAR DISTANCE FROM S—65 (KM) Fic. 2. Diagram of C-38 system showing stabilized water surface profiles relative to floodplain ground elevations along the Kissimmee River valley. No. 1, 1993] TOTH—THE ECOLOGICAL BASIS OF THE KISSIMMEE RIVER RESTORATION PLAN Z| floodplain wetlands (Milleson et al., 1980). However, transformation of the interact- ing river/floodplain ecosystem into a series of impoundments and deep canal drained much of the floodplain, stabilized water levels, and greatly modified flow character- istics. These physical alterations led to significant biological impacts. Between 12,000 - 14,000 ha of the pre-channelization floodplain wetlands were either completely drained, covered with spoil, or transformed into canal (Pruitt and Gatewood, 1976). As a result, the broadleaf marsh, wet prairie and wetland shrub communities that once dominated the floodplain were largely eliminated, and drained areas were converted to unimproved and improved pasture (Milleson et al., 1980). Moreover, stabilized water levels led to reduced diversity of plant species within remaining floodplain wetlands (Dineen et al., 1974; Goodrick and Milleson, 1974), which occur in the lower, impounded portions of each pool, and at the confluence of large tributary systems. Reduced diversity is most apparent over the floodplain landscape where broad expanses of relatively homogeneous plant communities have replaced the pre-channelization wetland mosaic. Drainage and degradation of floodplain wetlands eliminated fish and wildlife habitat and virtually destroyed the complex food web that the floodplain once supported. The number of active bald eagle (Haliaeetus leucocephalus leucocephalus) territories decreased by 74%, wintering waterfowl declined by 92%, and wading bird use of the lower Kissimmee River basin is extremely limited (Perrin et al., 1982; Shapiro et al., 1982). The naturalized cattle egret (Bubulcus ibis) a species that occurs primarily in association with cattle on pastures and other ruderal, terrestrial habitats has replaced the diverse complement of wading bird species that once used Wading Bird Densities 1978 - 1980 120 100 & Other Waders OO Cattle Egrets Mean Annual Density ( Birds per km? ) fon o Fic.3. Mean annual wading bird densities in the channelized Kissimmee River/C-38 system during 1978-80 (from Toland, 1990). 39 FLORIDA SCIENTIST [VOL 56 the floodplain wetlands (Fig. 3). Effects on the floodplain food base were even more striking. Based upon average densities in remaining marshes (Milleson, 1976), a constant source of over five billion forage fish and six billion freshwater shrimp (Palaemonetes paludosus) was lost as a result of floodplain drainage (Toth, 1990a). These small fish and invertebrates were a primary food source for wading birds and river fish species in the pre-channelization ecosystem. Post-channelization sampling of canal and remaining river habitat documented other impacts to fish and invertebrate communities. Two species, coastal shiner (Notropis petersoni) and blackbanded darter (Percina nigrofasciata), may have been extirpated (Perrin et al., 1982; Wullschleger et al., 1990a; Florida Game and Fresh Water Fish Commission, 1991), and in some remnant river channels game fish have declined from 43 to 28 percent of the total catch and the total number of fish species utilizing these habitats has fallen from 39 to 17 (Davis et al., 1990). Benthic invertebrate communities of the canal and remaining river runs (Table 5) display low densities and diversity, are dominated by taxa that are tolerant of degraded habitat conditions, and are characteristic of a reservoir rather than riverine environment. These impacts are linked to the destruction and degradation of the river channel and to unfavorable characteristics of canal habitat. Dissolved oxygen regimes (see Federico, 1982; Perrin et al., 1982; Rutter et al., 1986, 1989; Wullschleger et al., 1990a) are indicative of habitat degradation in the channelized system and low dissolved oxygen concentrations are a major cause of deterioration of biological communities. During summer and fall months, dissolved oxygen concentrations in the river and canal consistently fall below 2 mg L" - levels that are stressful, if not lethal, to most fish and invertebrate species (Fig. 4). Dissolved oxygen concentrations during this bottleneck period appear to be leading to long-term degradation of the system’s remaining fishery resources by limiting production of intolerant game fish species, and shifting the competitive environment in favor of tolerant, and typically, less desirable species, like gar (Lepisosteus spp.) and bowfin (Amia calva) (Florida Game and Fresh Water Fish Commission, 1991; Miller, 1992). Other proximate factors that have contributed to the deterioration of biological resources and ecological values of the channelized system also are linked to hydrologic alterations (Toth, 1990a). These include reduced substrate, depth, and flow diversity, and sedimentation, which, like unsuitable dissolved oxygen regimes, have resulted largely from altered flow characteristics. Maintenance of stable water levels has impacted food supplies of river or canal fish species by precluding the import of floodplain invertebrates which used to occur in the pre-channelization system when water drained off the floodplain during receding hydrographs. The food or energy base was impacted further by floodplain drainage, which led to replacement of the wetland plants, particularly riparian or near-bank species, that formerly fueled invertebrate food chains. Stabilized water levels also interfered with use of the floodplain as nursery or breeding areas by game fish species (Toth, 1991). Another floodplain function was lost as flow was confined within canal banks, preventing floodplain wetlands from filtering sediments and associated nutrients No. 1, 1993] |. TOTH—THE ECOLOGICAL BASIS OF THE KISSIMMEE RIVER RESTORATION PLAN 333 TABLE 5. Mean monthly (March - August 1984) densities of benthic macroinvertebrates in Ponar grab samples taken at two C-38 sites (n=12) and three remnant river sites (n=16)(from Rutter et al., 1986). ibaxa C-38 Remnant River Oligochaeta Aulodrilus piquetia 1.0 0.4 Branchiura sowerbyi 1.3 = Haemonais waldvogeli 0.8 0.4 Limnodrilus hoffmeisteri 5.6 0.6 Other 0.4 0.1 Amphipoda Hyalella azteca 4.2 3.6 Hydracarina 0.7 0.9 Ephemeroptera 2.0 0.1 Odonata 0.2 0.1 Coleoptera 0.3 = Hemiptera — 0.7 Trichoptera The 0.3 Chironomidae Chironomus spp. 1.3 4.6 Cladopelma sp. 0.3 Fo Cladotanytarsus spp. 2.9 - Clinotanypus sp. 1.6 - Coelotanypus scapularis 1.0 - Cryptochironomus fulvus 0.8 - Dicrotendipes spp. 1.2 0.9 Einfeldia natchitocheae - 2.3 Glyptotendipes spp. Su 5.9 Lauterborniella sp. Sell By Microtendipes sp. 15 = Polypedilum spp. ett = Procladius sp. 1.8 0.8 Tanytarsus sp. 8.3 0.6 Other 1.1 0.9 Ceratopogonidae 2.9 0.3 Chaoboridae Chaoborus punctipennis 0.9 10.1 Mollusca Corbicula fluminea 0.8 - Other Tes 0.6 Mean Sample Density 64.4 40.3 Mean # of Taxa per Sample 11.8 4.6 Mean Sample Diversity 2.83 1.26 from river discharge. These changes in physical and chemical attributes, biological resources, and functional processes, demonstrate that channelization destroyed the ecological integrity of the Kissimmee River ecosystem. EARLY RESTORATION STUDIES—Ecological impacts of channelization led to evaluations of the utility of various restoration measures. These studies began in the 34 FLORIDA SCIENTIST [VOL 56 C—38 REMNANT RIVER CHANNEL Dissolved Oxygen Profiles RS FRG FU Fd FAS FAS FS Fah Ph FAS Fh Fd Pd Pad Fes Fag Fa Fag RS RS RS RS PRS RS FS Re FS FS PRS es ss Fes Ps Surface hey 1m 0.8 2m 0.6 3m 0.4 4m 0.3 5m 0.2 6 m 0:2 Flocculent organic ; deposits 7m 0.0 8m 0.0 9m 0.0 Fic. 4. Typical summer-fall dissolved oxygen profiles in C-38 and remnant river channels. Data show mean dissolved oxygen concentrations (mg L"') of six canal sites and four remnant river sites on 29 August 1989. early 1970s and initially had two, somewhat independent areas of emphasis. One line of investigation considered the value of Kissimmee River restoration as a means to prevent or reverse degradation of water quality in Lake Okeechobee. Other studies evaluated measures to restore lost biological resources of the river and floodplain. At least part of the early impetus for restoration was derived from concern that construction of C-38 bypassed nutrient absorption processes on the Kissimmee River floodplain and thereby created a conduit for rapid downstream transport of nutrient loads into Lake Okeechobee (Governor’s Conference on Water Manage- ment in South Florida, 1971; Central and Southern Florida Flood Control District, 1972; Marshall et al., 1972). Sources of water quality degradation were clarified by the legislatively created “Special Project to Prevent the Eutrophication of Lake Okeechobee” (Huber et al., 1976; Florida Department of Administration, 1976; McCaffrey et al., 1977) and subsequent studies (Federico et al., 1978; Federico, 1982; Goldstein, 1986), which identified agricultural runoff, particularly from improved pasture and dairy operations, and associated secondary drainage systems, as the primary causes of elevated nutrient loads in the channelized C-38 system. Nutrient removal capabilities of Kissimmee marshes also were verified (Florida Department of Administration, 1976; Federico et al., 1978; Davis, 1981; Goldstein, 1986, 1990), but nutrient uptake “efficiency” (i.e., nutrient removal relative to inputs) appears to be dependent upon hydrology and nutrient loading rates. Davis (1981, 1992) found that re-establishment of floodplain wetlands with hydrologic characteristics and plant species composition resembling pre-channelization condi- tions reduced total phosphorus and inorganic nitrogen concentrations of river water by at least 40%, and that this “cleansing effect” persisted for over 10 years. However, a partially drained floodplain wetland which was subjected to seasonal wet-dry cycles, alternated over time as a source and sink for nutrients (particularly No. l, 1993] TOTH—THE ECOLOGICAL BASIS OF THE KISSIMMEE RIVER RESTORATION PLAN OD nitrogen)(Federico et al., 1978), while marshes exposed to high nutrient loads eventually reached an equilibrium or steady state in which nutrient input and output were equivalent (Goldstein, 1986, 1990). High nutrient loads also may lead to wetlands that are structurally and functionally different from pre-channelization swamps and marshes. Although these studies indicate that a restored Kissimmee River ecosystem can reduce nutrient loads, reestablishment of this function, and other environmental values of the former system, may be achieved only if key hydrologic characteristics of floodplain wetlands are restored, and elevated nutrient inputs generated by intensive agricultural land uses are reduced at their source. Initial studies on the potential for restoration of biological resources showed that reflooding of drained marsh resulted in rapid wetland vegetation growth and recolonization by small fish and invertebrates, including crayfish (Procambarus fallax) and freshwater shrimp (Palaemonetes paludosus) (Milleson, 1976). Another study (Goodrick and Milleson, 1974) provided early, post-channelization data on the viability of seed banks in floodplain soils, and showed that short-term drying (drawdown) of a remaining floodplain marsh produced explosive growth of wild millet (Echinochloa walteri), an annual plant which produces seed that is an important food source for waterfowl. Production of waterfowl food plants on the Kissimmee River floodplain also was enhanced by reintroduction of seasonal water level fluctuations (Perrin et al., 1982). These studies indicated that water level fluctuations can be effective in revitalizing floodplain wetland resources. Increased public outcry in the 1970s led to State and Federal legislation that provided mandates for evaluation of restoration plans. In 1976, the Florida legisla- ture passed the Kissimmee River Restoration Act (Chapter 76-112, Florida Statutes) which created the Coordinating Council on the Restoration of the Kissimmee River Valley and Taylor Creek-Nubbins Slough Basin. During the ensuing seven years, Coordinating Council activities provided a forum for technical evaluation and public review of Kissimmee River restoration needs, concepts and options. In addition to clarifying water quality issues in the basin (as discussed above), the principal outcome of the Coordinating Council studies and review process was endorsement of dechannelization, and specifically the “Partial Backfill Plan”, as the recommended means to restore the Kissimmee River. In 1985, the U.S. Army Corps of Engineers feasibility study and environmental impact statement on restoration alternatives quantified projected water quality and fish and wildlife benefits of the Partial Backfill Plan (U.S. Army Corps of Engineers, 1985). Water-quality modelling studies predicted that implementation of the Partial Backfill Plan would reduce river phosphorus concentrations and loadings to Lake Okeechobee by 41% by the year 2035. The U.S. Fish and Wildlife Service used the Habitat Evaluation Procedure (HEP) to evaluate effects of alternative restoration plans on availability of 17 habitat types, and the suitability of these habitats for 25 selected species or taxonomic groups. The Partial Backfill Alternative produced at least 20% more habitat units (a combined measure of habitat availability and suitability) than all other evaluated plans. KISSIMMEE RIVER DEMONSTRATION PROJECT—In response to the Kissimmee River Coordinating Council’s 1983 recommendation, the South Florida Water 36 FLORIDA SCIENTIST [VOL 56 Management District designed a demonstration project which was intended to resolve remaining technical issues regarding the Partial Backfill Plan and to evaluate the feasibility of restoring the system’s biological resources. The demonstration project was conducted in Pool B, a 19.5 km long section of canal, remnant river and floodplain, between S-65A and S-65B in Osceola, Okeechobee and Highlands Counties (Fig. 5). The project had four major components: implementation of a pool stage fluctuation schedule, construction of three weirs across C-38, creation of a “flow-through” marsh, and hydrologic and hydraulic modelling studies. Pool stage fluctuation was used as a means to counter the loss and degradation of wetland habitat that had resulted from adherence to stable pool stages. An annual 11.9 - 12.8 m (39 - 42 ft) schedule was intended to reestablish seasonal water level fluctuations over approximately 1080 ha of floodplain, including 526 ha that had been drained since channelization. The purpose of the weirs was to simulate effects of backfilling the canal by diverting flow through adjacent remnant river runs. Installation of weirs also was expected to produce additional floodplain inundation as water was detained upstream of these structures during discharge periods. Another feature of the Partial Backfill Plan, the “flow-through marsh” concept, was employed to recreate approxi- mately 121 ha of marsh in a section of the Pool B floodplain where ground elevations are higher than the peak stage of the fluctuation schedule. This impoundment was created by installing a culvert in the S-65A tieback levee and constructing a berm along C-38. Hydrologic and hydraulic modelling were used to evaluate engineering feasibility of dechannelization, flood control implications, and sedimentation issues (Loftin et al., 1990b). A multi-agency research team evaluated effects of demonstration project components on waterfowl, wading birds, fish, invertebrates, vegetation, water quality and habitat parameters. Habitat Parameters —Because channelization impacted environmental values of the Kissimmee River primarily by altering hydrologic regimes, an effective restoration plan must re-establish stage and discharge characteristics that formerly determined the ecological structure and function of the river ecosystem. Although the demonstration project was not intended to fully restore pre-channelization hydrology within the Pool B test area, hydrologic changes were expected to be the primary elicitor of biological responses. Effects of the demonstration project on floodplain hydrology varied along the length of the pool (Fig. 6). The greatest inundation occurred during 1987-88 when 80-90% of the floodplain in the lower 20% of the pool had inundation frequencies comparable to pre-channelization records; however, water level manipulations, which determined hydroperiods in this portion of the pool, did not reproduce historical water depths or year-to-year stage variability which once led to the diverse inundation and drying patterns that were a critical component of pre-channelization floodplain hydrology. In the middle 40% of the pool, the combined influence of controlled stage fluctuations and backwater effects of project weirs resulted in periodic flooding of 75% of the floodplain, but only 20% of the floodplain in this portion of the pool was exposed to prolonged inundation. Floodplain inundation in this section of the pool was limited by incongruity between the pool stage fluctuation schedule and floodplain ground elevations; the peak stage of the fluctuation sched- No. 1, 1993] _TOTH—THE ECOLOGICAL BASIS OF THE KISSIMMEE RIVER RESTORATION PLAN 37 S—65/A SENT BOCES 6 FLOW THROUGH MARSH Cus = Sle ele KILOMETERS RIVER RUN R4 FORT 0 1 2 3 KISSIMMEE C385 sii Bo C38 e—s Sin B4 RIVER RUN RS RIVER RUN R2 C35—— site B2 RIVER RUN R1 S658 Fic. 5. Map of Pool B showing demonstration project components and dissolved oxygen sampling locations. ule inundated only 40-45% of the middle Pool B floodplain. Backwater effects of weirs increased floodplain inundation primarily during January - April and Septem- ber, and only slightly increased the range of stage variability, which remained considerably lower than pre-channelization ranges during all months except March and April. In the northern 40% of Pool B, backwater effects of weirs only periodically reflooded about 30-35% of the completely drained floodplain. More consistent flooding occurred within the flow-through marsh, where pre-channelization inunda- tion frequencies were restored on approximately 35% of the floodplain and at least 38 FLORIDA SCIENTIST [VOL 56 Edge of Historic Floodplain Edge of Historic Floodplain Prolonged Inundation Periodic Inundation Drained Spoil C-38 Canal Fic. 6. Typical floodplain inundation during the demonstration project. No. 1, 1993} TOTH—THE ECOLOGICAL BASIS OF THE KISSIMMEE RIVER RESTORATION PLAN 39 55% of the area was inundated seasonally. The influence of backwater effects on floodplain inundation was limited by rates at which water drained off the floodplain. Resultant “spiked” hydrographs con- trasted sharply with the gradual rates at which water levels on the floodplain typically receded prior to channelization. Pre-channelization stage recession rates typically did not exceed 0.03 m day", whereas floodplain water levels often declined at rates > 0.08 m day’ during the demonstration project. Rapid stage recession rates were caused by the drainage capacity of C-38 and operational rules governing outflow from the upper Kissimmee lakes, which frequently produced abrupt declines in discharge. Re-introduction of flow through river runs adjacent to weirs was a primary objective of the demonstration project. Resultant river flow regimes were a function of upper basin discharge characteristics and the flow diversion efficiency of weirs. Because each weir had a central 11.9 m wide, 1.2 m deep navigation opening, a maximum of 60% of C-38 flow was diverted through adjacent floodplain and river channels during high discharge periods, and much lower proportions were diverted when discharges were < 28 m? sec’ (Loftin et al., 1990a). This limitation, coupled with the upper basin lakes regulation schedules and operation rules, produced river flow regimes that differed greatly from pre-channelization flow characteristics. Highest discharges occurred from January - April, during the drawdown phase of the upper lakes schedule, rather than during wet season months. During high flows, the drainage capacity of the canal produced an unnaturally steep gradient and velocities as high as 0.9 m sec’ in river runs adjacent to weirs. Extended no flow periods were common during June-December each year, and typical pre-channelization base flows (i.e., discharges exceeding 11 m° s') were generated through river runs adjacent to weirs only half as frequently during the demonstration project. Despite these uncharacteristic flow attributes, the demonstration project showed that re-introduction of flow and associated fluvial processes enhance diversity and quality of degraded river habitat. Natural channel morphology and a predominantly sand substrate were restored in river runs adjacent to weirs through gradual flushing and/or covering of the layer of organic deposits which had accumulated on the river bottom since channelization (Fig. 7). Water Quality—No significant changes in nutrient levels were detected over the four-year monitoring period (Rutter et al., 1989). These results largely repudi- ated concerns that flushing of bottom organic deposits from remnant river channels, and reflooding of drained floodplain which had been converted to pastureland, would result in increased nutrient loads. In fact, sediment and nutrient filtration functions were reestablished on at least part of the floodplain. During a high discharge test, virtually the entire suspended solids load was removed, and total phosphorus concentrations were 68% lower in water draining off the floodplain than in river discharge (Toth, 1991). Following the test, new organic deposits were conspicuous at mid-pool, floodplain sampling sites. Chlorophyll a data provided more evidence indicating that the channelized system functions more like a degraded reservoir than a natural river. During the demonstration project, average chlorophyll a concentrations at both canal and river sampling locations commonly exceeded levels (i.e., > 20 ug L') that are indicative of 40 FLORIDA SCIENTIST [VOL 56 MARCH 1985 DEPTH (CM) LOCATION ALONG X—SECTION (1.5 m INTERVALS) JULY 1985 NOVEMBER 1988 DEPTH (CM) LOCATION ALONG X-SECTION (1.5 m INTERVALS) Fic. 7. Changes in river bottom profiles following re-establishment of flow through remnant river channels. eutrophic conditions in lakes (Carlson, 1977). Low dissolved oxygen levels remained the most significant water quality prob- lem in the demonstration project area. Although reintroduction of flow through remnant river channels resulted in a more uniform surface to bottom distribution of dissolved oxygen (Rutter et al., 1989), river concentrations were determined prima- rily by ambient dissolved oxygen levels of diverted canal water. Dissolved oxygen No. 1, 1993] TOTH—THE ECOLOGICAL BASIS OF THE KISSIMMEE RIVER RESTORATION PLAN — 4 concentrations < 2.0 mg L" recurred consistently during summer and fall months in C-38 and hydraulically connected river runs (Fig. 8a-c). However, a remnant river channel which received regular baseflows from a tributary watershed maintained surface to bottom dissolved oxygen concentrations > 3.0 mg L? throughout the summer (Fig. 8d). These results indicate that re-establishment of more natural inflows, including continuous baseflows, would lead to improved dissolved oxygen regimes in a restored river system. 4 —— C38 - Site B4 —— River Run - R2 —— C38 - Site B2 | —+— River Run - R1 Dissolved Oxygen (mg L”") 1.6) Dissolved Oxygen (mg L") 4, 65 6 7 8 °9 Oe QOS 4G NG UT eS Depth (m) Depth (m) —— C38 - Site B6 —— River Run - R4 —— C38 - Site BS —— River Run - R3 Dissolved Oxygen (mg L") Dissolved Oxygen (mg L") ONDA 2) 3.8142 HS) Ge 7 Beg Depth (m) Depth (m) Fic. 8. Summer-fall dissolved oxygen regimes in Pool B during the demonstration project. Data show mean dissolved oxygen concentrations from four canal sites (C-38 - sites 2,4,5,6) and four remnant river sites in R1, R2, and R4 and five sites in R3 over five bi-weekly sampling dates from 15 August - 10 October 1989. Sampling locations are shown in F igure 7. 49 FLORIDA SCIENTIST [VOL 56 Vegetation—Plant community responses during the demonstration project indicate that reestablishment of appropriate hydrologic conditions can lead to rapid restoration of river and floodplain wetland vegetation characteristics. Significant changes in plant community composition occurred in time periods as short as one year, and included responses to subtle, as well as, major changes in flow, water depths, inundation frequencies, and temporal inundation patterns (Miller et al., 1990; Toth, 1991). The primary effect of reintroduced flow was to reduce encroach- ment by vegetation into the center of remnant river channels by confining growth of emergent and floating species to a littoral fringe. However, the greatest response to hydrologic changes occurred on the floodplain, where distributions and growth of hydrophytic species, particularly Alternanthera philoxeroides, Panicum hemitomon and Polygonum punctatum, expanded, while frequencies of mesophytic and xero- phytic taxa declined (Table 6). These results indicate that many of the remaining wetland species on the channelized floodplain are sensitive to hydrologic change, and have the reproductive potential, including a viable seed bank, to rapidly colonize habitats with favorable hydrology. Re-established hydrophytic species partially restored functional values of the floodplain by increasing habitat diversity (e.g., patches of Alternanthera beds) and providing potential food sources (e.g., Polygo- num seed) for waterfowl (Beckwith and Hosford, 1957; Stieglitz, 1972; Landers et al., 1976). TABLE 6. Plant species whose distributions on the Pool B floodplain were affected significantly by hydrologic changes during the demonstration project (from Toth, 1991). Species With Increased Frequencies of Occurrence Alternanthera philoxeroides Eleocharis vivipara Panicum hemitomon Polygonum punctatum Salix caroliniana Species With Decreased Frequencies of Occurrence Ambrosia artemisiifolia Axonopus affinis Axonopus compressus Boltonia diffusa Centella asiatica Eupatorium capillifolium Hydrocotyle sp. Paspalum conjugatum Sambucus canadensis Urena lobata No. 1, 1993] TOTH—THE ECOLOGICAL BASIS OF THE KISSIMMEE RIVER RESTORATION PLAN 43 While observed vegetation changes clearly demonstrated the feasibility of reestablishing vegetative characteristics of littoral and floodplain habitats, the inadequacy of demonstration project hydrology led to only temporary or incomplete restoration of these wetland communities. Prolonged periods of no or low flow, allowed nuisance exotics such as water hyacinth, Eichornia crassipes, and water lettuce, Pistia stratiotes, to form dense mats, which choked sections of river channel and impacted native littoral species (Miller et al., 1990). Weed species also persisted on drained floodplain that was subjected to periodic or seasonal inundation. These “shortcomings” of the demonstration project emphasize that restoration of biologi- cal attributes requires appropriate river and floodplain hydrology. Prior to channelization, a diverse mosaic of littoral and floodplain plant communities was maintained by continuous flow, and highly stochastic and widely varying stage and discharge regimes. Invertebrates—Reintroduction of flow through remnant river runs led to incipient changes in invertebrate communities, which at least temporarily reestab- lished species composition with rudimentary characteristics of a natural river community. Invertebrate responses to flow, or to flow-induced habitat (e.g., sub- strate and dissolved oxygen regimes) improvements, included increased densities and diversity in littoral habitats, elimination of Chaoborus punctipennis as a domi- nant species, and colonization by Sphaeriacean clams and several rheophilic taxa (Stenacron interpunctatum, Cheumatopsyche sp. and Rheotanytarsus spp.)(Rutter et al., 1989; Toth, 1991). However, the inadequacy of river inflows precluded more complete or lasting restoration progress; Sphaeriacean clams largely disappeared and densities of Chaoborus increased following a three month period of no flow during summer 1987 (Table 7). These responses verified the importance of continu- ous flow, and demonstrated that reestablishment of this key hydrologic determinant of river invertebrate community structure will lead to restoration of a natural river community. Peak benthic invertebrate densities in remnant river channels with reintroduced flow ranged to 22,000 invertebrates m”, and typically were highest in littoral habitats and during spring sampling periods. Invertebrate colonization rates on reflooded portions of drained floodplain indicated that this trophic link in the food web (e.g., between wetland vegetation and higher level consumers like fish and waterfowl) also can be reestablished fairly rapidly (Toth, 1991). Although seasonally reflooded habitats lacked a full comple- ment of trophic guilds (e.g., shredders), representative densities of the most common floodplain invertebrate taxa were attained after about 40 days of inunda- tion. Highest densities approached 40,000 invertebrates m?, and were found in reflooded areas that were in close proximity to, or hydraulically connected with, existing aquatic habitats. These included a periodically reflooded section of formerly drained floodplain which was surrounded by broadleaf marsh, and floodplain habitats where colonization was facilitated by overbank flow. The trophic impor- tance of hydraulic connectivity between the river and floodplain also was illustrated by invertebrate export rates when water drained off the floodplain. Demonstration project sampling indicated that small rivulets draining the floodplain may carry up to 4800 small fish and invertebrates per hour to adjoining river habitats (Toth, [VOL 56 FLORIDA SCIENTIST - ajqeyynuepiuy) epodAosajeg og aepruaeyds epoddosgjeg - aepruolu yr) epoddosgjog 6 pauruny pjna1g40y epoddosgjag - g1qeyuepiuy epodo.jsey 6 aepisdyg ‘aepiqiourlg epodoisesy F ia aeplig ‘eepiqoipAy epodossesy - - - 7 - 60T om - — aeprfAouy epodo.jse5 - - - = 6 - ~ — 7 1ayIO eiaidiq PPS 6 LS 961 = SS FOI 96LT 1661 syuuadijound snsoqovy,) e1aydiq ZS €8T €8 ZS1 6E €8 L8 bP 601 oepruosodoyesaD eiaidiq yi» ell 002 96 6ST 002 OLT — 609T €19 oe PIWOUOIIY) vrajdiq — fF - P - - 6 ~ p synpy pure ovare'y v1a}doajop = GS v = = as = - - aepipodosjuso4jog v1a}doyou ~ 19 Py 6 ZC - - - - aepiia00}daq v1ajdoyou fy, fF 9% Pp - fF - - ~ - aepiyodsdospAY ‘sepyydorpAy viaydoyou J - - Pp - - ZG 6 - Fr IEPILIBPUZOD eyeUOPO - ~ uy 6 CT 9G - - Pp via}dostuy eyeUOpO = = = - es = = i i be Mee ed |) - - - - ~ - - - F a]qeyqueptiuy viaydo1smiayd - ZG pA Ee L8T OF LT ce P DINU s1UuaDD viaydo1swiaydy 0€ Ze p Srl ce El SF p = euLeorpAy - - - - LI - - ~ - snsopnjod sajauowan)pg epodeoeq - Pp - - - ~ - ~ t oPPLIRULLURS) epodiydury = FOIT = 4 O€9T FOI 6 ZS LI poaq=v Dyjajvhy epodiydury 96S 00% yy 16S ee COT, == = 1660 966 PLT PPOdRIISO) 99% 96 ZI cg LST ZG 18 19F OGF epodedo5 at LIE 0€ - SLT Lg 6 18 6E B1200PP|.) eTOI 6£9 rrS = PLGECGGET 1633 OOZE 0€8 cg Boyoos!O 181 BEE OgT LIZ FOS 191 6 6 = epoyewan ‘auIpraty PUR|Oqan,L ONV UdV AON ONV UdV OHd ONV UdV OHd UxeL L861 ee aaa es a ee 44 (166 ‘OL Wor) E¥ 1M 0} JUDOR[pe oUURYO JOAL JURUWaL oY} WO sojdures jaUURYO—prIUL UT (__U #) SATISUSP ayBIQOVOAUl URI‘) ATAVL No. 1, 1993] TOTH—THE ECOLOGICAL BASIS OF THE KISSIMMEE RIVER RESTORATION PLAN 45 1990a). Thus, floodplain invertebrate production may contribute to the river/ floodplain food web directly (via consumption of marsh invertebrates by small forage fish like Gambusia), or indirectly (as in consumption of forage fish by wading birds), and may occur through predation on the floodplain, or following export to the adjoining river channel. Monitoring data showed that floodplain invertebrate den- sities peak in spring, which can be timely for many waterfowl and river fish species that produce young during this period. Wading birds and waterfowl—The demonstration project also led to increased utilization of the Pool B floodplain by wading birds and waterfowl (Toland, 1990 and unpublished manuscript). During 1987-89, Pool B had the highest duck density (3.9 birds per km?) and wading bird and waterfowl species diversity and richness of any of the five pools within the Kissimmee River/C-38 system. Pool B wading bird densities were over two-times higher at the end of the demonstration project monitoring period than densities found in this area during 1978-80 (Fig. 9). Moreover, cattle egrets comprised only 20% of the wading bird densities in Pool B compared to 42% of the total wading bird numbers in the Kissimmee River/C-38 system. Much of the increased wading bird and waterfowl utilization of the Pool B 60 1978-80 oA >0')| | mt 1987-88 gp O 1988-89 40 : 30 20 Pool Pool Pool Pool Pool Paradise A B C D E Run Density ( wading birds per km‘? ) Fic.9. Densities of wading birds in the Kissimmee River/C-38 system before (1978-80) and during (1987-89) the Pool B demonstration project (from Toland, unpublished manuscript). 46 FLORIDA SCIENTIST [VOL 56 floodplain occurred in the flow-through marsh, where inundation patterns were comparable to pre-channelization hydroperiods. The area included within the flow- through marsh aerial surveys comprised < 40% of the Pool B floodplain, but supported 70% of the waterfowl and 66% of the wading birds and had the highest wading bird density (41.5 birds per km’) in the Kissimmee River/C-38 system. Fish—Pool B fish populations exploded in response to a prolonged period of high water levels in winter and spring of 1987-88 (Florida Game and Fresh Water Fish Commission, 1991). Increased pool stages also led to expansion of resident, marsh fish populations (Fig. 10), but it appears that water levels did not get deep enough, and/or were not deep long enough, to accomodate utilization of marsh habitat by fish species that typically occur in the adjacent river channel (Toth, 1991). Electrofishing catch rates of several game fish species, including largemouth bass, 13 (r=0.72) MEAN NUMBER OF FISH PER m® Oo = NO Ww > ul o N @ o MEAN SAMPLE DEPTH (cm) Fic. 10. Marsh fish density in relation to mean water depths at sampling locations. Data points show means of triplicate samples from 17 sampling dates during 1985-88. bluegill, spotted sunfish (Lepomis punctatus) and warmouth (Lepomis gulosus), suggest utilization of remnant river runs with reintroduced flow is increasing; however, most game fish abandon these remnant river habitats during late summer and fall when flows cease and dissolved oxygen levels plummet (Wullschleger et al., 1990a; Florida Game and Fresh Water Fish Commission, 1991). Prevailing low dissolved oxygen regimes and discontinuous flow limited any enhancement of Pool B fish populations and degradation of game fish resources has continued. For example, although mean annual abundance of largemouth bass and other game fish increased significantly between 1987 and 1990, biomass of these species remainded relatively unchanged (Florida Game and Fresh Water Fish Commission, 1991). This trend reflects a shift in population size structure from adults to young fish, and indicates that these species are reproducing successfully but No. 1, 1993] | TOTH—THE ECOLOGICAL BASIS OF THE KISSIMMEE RIVER RESTORATION PLAN 47 few recruits are surviving through early ages. Thus, the number of fish that are living to a reproductive age may be getting smaller. Low dissolved oxygen levels during summer and fall may be the key limiting factor in game fish recruitment. However, game fish also are being subjected to increasing competition and predation pressures from gar and bowfin, which showed increasing abundance and biomass from 1987-90. Conclusions—Demonstration project monitoring results confirmed the feasi- bility of restoring the structure and functions of the Kissimmee River ecosystem. Water level manipulations, increased floodplain inundation and reintroduction of flow re-established wetland plant communities, increased wading bird and water- fowl utilization, improved habitat conditions, and led to favorable responses by invertebrates, forage fish and game fish. However, because hydrologic characteris- tics were not reestablished completely, most biological components of the river and floodplain were affected temporarily and/or only partially restored. The demonstra- tion project did not restore any portion of the Kissimmee River, but provided evidence indicating that restoration of ecological integrity of this river/floodplain ecosystem is possible. KISSIMMEE RIVER RESTORATION PLAN—Restoration of the Kissimmee River must be based upon an ecosystem perspective. Lost river values clearly were dependent upon complex environmental attributes, including numerous physical, chemical and biological processes, dynamics of intricate food webs, and an array of river and floodplain habitat characteristics and interactions. However, integration of accumulated technical data on pre-channelization resources, impacts of channelization, and demonstration project results indicate that restoration of the ecological integrity of the river and floodplain can be achieved by reestablishing the physical form and hydrology of the ecosystem. Physical form requirements include reestablishment of natural river channel morphology, the lateral connectivity be- tween the river and floodplain, and longitudinal continuity of river and floodplain habitat (Egbert, 1990; Karr, 1990; Toth, 1990b). Hydrologic restoration must include both stage and flow characteristics (Egbert, 1990; Johnson and Turnbull, 1990; Karr, 1990; Toth, 1990b; Wullschleger et al., 1990b), particularly reestablish- ment of natural inflows from the headwater lakes (Loftin and Obeysekera, 1990; Obeysekera and Loftin, 1990; Toth, 1990b; Williams, 1990; Toth et al., 1992). The current restoration plan for the Kissimmee River (U.S. Army Corps of Engineers, 1991) would reestablish physical form and pre-channelization hydrologic characteristics along 70 continuous kilometers of river channel and over approxi- mately 11,000 ha of floodplain, and would restore the ecological integrity of 104 km? of river ecosystem. FUTURE STUDIES—Restoration evaluation will be the cornerstone of future environmental studies on the Kissimmee. To accurately assess the success of restoration, the scope of restoration evaluation efforts must be congruous with the goal and therefore have an ecosystem perspective. A restoration evaluation program 48 FLORIDA SCIENTIST [VOL 56 should have the following features: 1. Provide a thorough understanding of ecosystem structure and function, including predictive capability for most components—with and without restoration; 2. Show direct cause-effect relationships between restoration measures and ecological responses; 3. Include quantifiable biological responses and statistical comparisons; 4. Document ecological changes that are of both social and scientific importance. A comprehensive evaluation program for the Kissimmee River restoration project will include a combination of taxonomic, habitat and conceptual approaches which will provide solid scientific data on the processes responsible for pattern in restored areas (Karr et al., 1991). Hydrology, plant communities, birds and fish are critical components, and development of a suite of quantitative and conceptual models for these components, as well as, key water chemistry attributes such as dissolved oxygen, organic carbon and nutrients, are needed. A comprehensive restoration evaluation program will include five aspects or phases (Karr et al., 1991). First, reference conditions must be established to define expectations for each evaluation component. Reference conditions may include a reference site or location, historical data or theoretical approaches. Secondly, baseline or existing conditions must be measured to allow statistical comparisons with the expected (reference) conditions, and ultimately with actual conditions achieved by restoration. Both of these phases are integral components of the required modelling studies. Another phase of the evaluation program will involve measurements of construction impacts. Construction impact assessments will en- sure that any temporary or incidental environmental impacts are minimized and/or alleviated as the project progresses. Post-construction studies will be the most complex aspect, and phase with the longest duration. Restoration of complex components such as the floodplain food web could be a relatively slow process, but short-term trends will be evaluated and are critical to the success of the fifth evaluation phase - the adaptive management phase. This aspect of the evaluation program will use acquired information for continual, scientifically informed fine- tuning of the restoration efforts, particularly the water management component of the project. The available data base provides direction for implementation of the restoration evaluation program. Pre-channelization data are adequate to establish reference conditions for hydrology and the distribution of wetland plant communities, and an excellent baseline data base is available for most water chemistry parameters. Efforts are currently underway to fill in data gaps that are required for establishing other reference and baseline conditions, and to construct relevant models. ACKNOWLEDGMENTS—I respectfully acknowledge the work of the scientists whose contributions to Kissimmee River restoration efforts are reviewed in this paper. I thank N. Aumen, D. Black, J. Karr, G. Redfield, T. Tisdale and an anonymous reviewer for helpful comments on earlier drafts, and G. Eyeington, B. Moghaddam, and J. Noel for assistance in preparing the manuscript. No. 1, 1993] tTorH—THE ECOLOGICAL BASIS OF THE KISSIMMEE RIVER RESTORATION PLAN 49 LITERATURE CITED BECKWITH, S.L. AND H.J. HOSFORD. 1957. A report on seasonal food habits and life history notes of the Florida duck in the vicinity of Lake Okeechobee, Glades County, Florida. Am. Mid]. Nat. 57: 461- 473. CARLSON, R.E. 1977. A trophic state index for lakes. Limnol. Oceanogr. 22: 361-369. CENTRALAND SOUTHERN FLORIDA FLOOD CONTROL DISTRICT. 1972. Findings and recommendations of the governing board, Central and Southern Florida Flood Control District, as the result of public hearing concerning alleged environmental damage resulting from channelization of the Kissimmee River, November 15, 1972. Centr. South. Fla. Flood Control Dist., West Palm Beach, BE. / Davis, L.J., S.J. MILLER, AND J.G. WULLSCHLEGER. 1990. Paradise Run fisheries investigations. Pp. 149-159. In: LOFTIN, M.K., L.A. TOTH, AND J. OBEYSEKERA.(eds.), Proc. Kissimmee River Restoration Symp., October 1988, Orlando, Florida. S. Fla. Water Manage. Dist., West Palm Beach, FL. Davis, S.M. 1981. Mineral flux in the Boney Marsh, Kissimmee River. Mineral retention in relation to overland flow during the three-year period following reflooding. S. FL. Water Manage. Dist., Tech. Publ. #81-1, West Palm Beach, Fla. 54 pp. 1992. S. Fl. Water Manage. Dist., West Palm Beach, FL. Pers. Commun. DINEEN, J.W., R.L. Gooprick, D.W. HALLETT, AND J.F. MILLESON. 1974. The Kissimmee River revisited. Centr. South. Fla. Flood Control Dist., In Depth Report, Vol. 2, No. 2, 11 pp. EGBEBT, A. 1990. Fish and wildlife perspectives on Kissimmee River restoration. Pp. 295-98. In: LOFTIN, M.K., L.A. TOTH, AND]. OBEYSEKERA.(eds.). Proc. Kissimmee River Restoration Symp.., October 1988, Orlando, Florida. S. Fla. Water Manage. Dist., West Palm Beach, FL. FEDERICO, A.C., J.F. MILLESON, P.S. MILLAR, AND M. ROSEN. 1978. Environmental studies in the Chandler Slough watershed. S. Fla. Water Manage. Dist., Tech. Publ. #78-2, West Palm Beach, FL. 120 pp. 1982. Water quality characteristics of the lower Kissimmee River basin, Florida. S. Fla. Water Manage. Dist., Tech. Publ. #82-3. 107 pp. FLORIDA DEPARTMENT OF ADMINISTRATION. 1976. Final report on the special project to prevent eutrophication of Lake Okeechobee. Div. State Planning, Tallahassee, FL. FLORIDA GAME AND FRESH WATER FISH COMMISSION. 1957. Recommended program for Kissimmee River Basin. Fla. Game and Fresh Water Fish Comm. Mimeo. 38 pp. 1991. Lake Okeechobee-Kissimmee River-Everglades resource evaluation project. Federal Wallop-Breaux completion report (F-52-5). 328 pp. GOLDSTEIN, A.L. 1986. Upland detention/retention demonstration project, final report. Impacts of agricultural land use on water quality and utilization of wetlands for detention/retention in the Kissimmee River basin. S. Fla. Water Manage. Dist., Tech. Publ. #86-2, West Palm Beach, FL. 374 pp. 1990. Upland demonstration project - water quality impacts of agricultural land use and wetland BMPs in the Kissimmee River valley. Pp. 111-124. In: LOFTIN, M.K., L.A. TOTH, AND J. OBEYSEKERA.(eds.) Proc. Kissimmee River Restoration Symp., October 1988, Orlando, Florida. S. Fla. Water Manage. Dist., West Palm beach, FI. GOODRICK, R.L. AND J.F. MILLESON. 1974. Studies of floodplain vegetation and water level fluctuations in the Kissimmee River valley. S. Fla. Water Manage. Dist., Tech. Publ. #74-2. 60 pp. GOVERNOR’S CONFERENCE ON WATER MANAGEMENT IN SOUTH FLORIDA. 1971. Statement to Governor Reubin O’D. Askew. Centr. South. Fla. Flood Control Dist., Water Manage. Bull., Vol. NOT. HUBER, W.C., J.P. HEANEY, P.B. BEDIENT, AND J.P. BOWDEN. 1976. Environmental resources management studies in the Kissimmee River basin. Final Rept. Dept. Environ. Eng. Sci., Univ. Florida, Gainesville, FL. ENV-05-76-2. 279 pp. JOHNSON, F.A. AND R.A. TURNBULL. 1990. Restoration of waterfowl habitat in the Kissimmee River valley. Pp. 57-66. In: LOFTIN, M.K., L.A. TOTH, AND J]. OBEYSEKERA.(eds.) Proc. Kissimmee River Restoration Symp., October 1988, Orlando, Florida. S. Fla. Water Manage. Dist., West Palm Beach, FI. KARR, J.R. 1990. Kissimmee River: Restoration of degraded resources. Pp. 303-320. In: LOFTIN, M.K., TOTH, L.A., AND J. OBEYSEKERA (eds.) Proc. Kissimmee River Restoration Symp., October 1988, Orlando, Florida. S. Fla. Water Manage. Dist., West Palm Beach, FL. 50 FLORIDA SCIENTIST [VOL 56 , H. STEFAN, A.C. BENKE, R.E. Sparks, M.W. WELLER, J.V. McArtuur, AND J.H. Zar. 1991. Design of a restoration evaluation program. Rept. to the S. Fla. Water Manage. Dist. 46 pp. LANDERS, J.L., A.S. JOHNSON, P.H. MORGAN, AND W.P. BALDWIN. 1976. Duck foods in managed tidal impoundments in South Carolina. J. Wildl. Manage. 40: 721-728. LOFTIN, M.K. AND J. OBEYSEKERA. 1990. Kissimmee restoration engineering considerations in a multiobjective framework. Pp. 187-96. In: LOFTIN, M.K., L.A. TOTH, AND J. OBEYSEKERA.(eds.) Proc. Kissimmee River Restoration Symp., October 1988, Orlando, Florida. S. Fla. Water Manage. Dist., West Palm Beach, FL. , J. OBEYSEKERA, C. NEIDRAUER, AND S. SCULLEY. 1990a. Hydraulic performance of the Phase I Demonstration Project. Pp. 197-209. In: LOFTIN, M.K., L.A) TOnEayANDOR|: OBEYSEKERA.(eds.) Proc. Kissimmee River Restoration Symp., October 1988, Orlando, Florida. S. Fla. Water Manage. Dist., West Palm Beach, FL. , L.A. TOTH, AND J.T.B. OBEYSEKERA. 1990b. Kissimmee River restoration: Alternative plan evaluation and preliminary design report. S. Fla. Water Manage. Dist., West Palm Beach, FL. 147 dp. Mersin A.R., J.H. HARTWELL, D.S. ANTHONY, J.V. BETZ, A.E. LUGO, A.R. VERI, AND S.U. WILSON. 1972. The Kissimmee-Okeechobee basin. A report to the Cabinet of Florida. Univ. Miami, Div. Appl. Ecol., Ctr. Urban and Regional Stud., Miami, FL. 52 pp. MCCAFFREY, P.M., W.H. HINKLEY, J.M. RUDDELL, AND S.E. GATEWOOD. 1977. First annual report to the Florida legislature. Coordinating Council on the Restoration of the Kissimmee River Valley and Taylor Creek-Nubbin Slough Basin. Tallahassee, FL. 232 pp. MILLER, S.J. 1990. Kissimmee River fisheries: A historical perspective. Pp. 31-42. In: LOFTIN, M.K., L.A. TOTH, AND J. OBEYSEKERA.(eds.) Proc. Kissimmee River Restoration Symp., October 1988, Orlando, Florida. S. Fla. Water Manage. Dist., West Palm Beach, FL. 1992. St. John’s River Water Manage. Dist., Palatka, FL. Pers. Commun. J. WOOD, AND L. PERRIN. 1990. Vegetation community responses to restoration. Pp. 97-110. In: LOFTIN, M.K., L.A. TOTH, AND J. OBEYSEKERA.(eds.) Proc. Kissimmee River Restoration Symp., October 1988, Orlando, Florida. S. Fla. Water Manage. Dist., West Palm Beach, FL. MILLESON, J.F. 1976. Environmental responses to marshland reflooding in the Kissimmee River basin. S. Fla. Water Manage. Dist., Tech. Publ. #76-3. 39 pp. , R.L. GOODRICK, AND J.A. VAN ARMAN. 1980. Plant communities of the Kissimmee River valley. S. Fla. Water Manage. Dist., Tech. Publ. #80-7. 42 pp. OBEYSEKERA, J. AND M.K. LOFTIN. 1990. Hydrology of the Kissimmee River basin - influence of man- made and natural changes. Pp. 211-222. In: LOFTIN, M.K., L.A. TOTH, AND J. OBEYSEKERA.(eds.) Proc. Kissimmee River Restoration Symp., October 1988, Orlando, Florida. S. Fla. Water Manage. Dist., West Palm Beach, FL. PERRIN, L.S., M.J. ALLEN, L.A. ROWSE, F. MONTALBANO III, K.J. FOOTE, AND M.W. OLINDE. 1982. A report of fish and wildlife studies in the Kissimmee River Basin and recommendations for restoration. Fla. Game and Fresh Water Fish Comm., Office of Environ. Serv., Okeechobee, FL. 260 pp. PIERCE, GI. A.B. AMERSON, AND L.R. BECKER JR. 1982. Pre-1960 floodplain vegetation of the lower Kissimmee River valley, Florida. Final Rept. Environ. Consult., Inc., Dallas, Texas. Biol. Serv. Rept. 82-3. 24 pp. PRUITT, B.C. AND S.E. GATEWOOD. 1976. Kissimmee River floodplain vegetation and cattle carrying capacity before and after canalization. Fla. Div. State Planning, Tallahassee, FL. 57 pp. RUTTER, R.P., D.E. SESSIONS, G.M. SLOANE, AND D.A. WINKLER. 1986. Kissimmee River restoration project: Pre-construction monitoring. Fla. Dept. Environ. Reg., S. Fla. Dist., Punta Gorda, FL. , D.E. SESSIONS, AND D.A. WINKLER. 1989. Kissimmee River restoration project: Post- construction monitoring. Fla. Dept. Environ. Reg., S. Fla. Dist., Punta Gorda, FL. SHAPIRO, A.E., F. MONTALBANO II, AND D. MAGER. 1982. Implications of a construction of a flood control project upon bald eagle nesting activity. Wilson Bull. 94: 55-63. STIEGLITZ, W.O. 1972. Food habits of the Florida duck. J. Wildl. Manage. 36: 422-428. TOLAND, B.R. 1990. Effects of the Kissimmee River Pool B restoration Demonstration Project on Ciconiiformes and Anseriformes. Pp. 83-92. In: LOFTIN, M.K., L.A. TOTH, AND J. OBEYSEKERA.(eds.) Proc. Kissimmee River Restoration Symp., October 1988, Orlando, Florida. S. Fla. Water Manage. Dist., West Palm Beach, FL. TOTH, L.A. 1990a. Impacts of channelization on the Kissimmee River ecosystem. Pp. 47-56. In: LOFTIN, M.K., L.A. TOTH, AND J. OBEYSEKERA.(eds.) Proc. Kissimmee River Restoration Symp., October 1988, Orlando, Florida. S. Fla. Water Manage. Dist., West Palm Beach, FL. 1990b. An ecosystem approach to Kissimmee River restoration. Pp. 125-133. In: LOFTIN, M.K., No. 1, 1993] TOTH—THE ECOLOGICAL BASIS OF THE KISSIMMEE RIVER RESTORATION PLAN 5] 1990b. An ecosystem approach to Kissimmee River restoration. Pp. 125-133. In: LOFTIN, M.K., L.A. TOTH, AND J. OBEYSEKERA.(eds.) Proc. Kissimmee River Restoration Symp., October 1988, Orlando, Florida. S. Fla. Water Manage. Dist., West Palm Beach, FL. 1991. Environmental responses to the Kissimmee River Demonstration Project. S. Fla. Water Manage. Dist., Tech. Publ. #91-2, West Palm Beach, FL. 96 pp. , |. T.B. OBEYSEKERA, W.A. PERKINS, AND M.K. LOFTIN. 1992. Flow regulation and restoration of Florida’s Kissimmee River. Regulated Rivers, in press. U.S. ARMY CORPS OF ENGINEERS. 1956. Central and southern Florida, Kissimmee River basin and related areas. Supplement 5 - General Design Memorandum, Kissimmee River Basin. U.S. Army Corps of Eng., Jacksonville Dist., Jacksonville, FL. , 1985. Central and southern Florida, Kissimmee River, Florida. Final feasibility report and environmental impact statement. U.S. Army Corps of Eng., Jacksonville Dist., Jacksonville, FL. , 1991. Central and southern Florida. Final integrated feasibility report and environmental impact statement: Environmental restoration of the Kissimmee River, Florida. U.S. Army Corps of Engineers, Jacksonville District, Jacksonville, Florida. 264 pp. U.S. FISH AND WILDLIFE SERVICE. 1959. A detailed report of the fish and wildlife resources in relation to the Corps of Engineers’ plan of development, Kissimmee River basin, Florida. Appendix A in Central and Southern Florida Project for Flood Control and Other Purposes. Part II. Kissimmee River Basin and Related Areas. Supplement 5. U.S. Army Corps of Eng., Jacksonville Dist., Jacksonville, FL. WILLIAMS, V.P. 1990. Management and mis-management of the upper Kissimmee River basin chain of lakes. Pp. 9-30. In: LOFTIN, M.K., L.A. TOTH, AND J. OBEYSEKERA. (eds.) Proc. Kissimmee River Restoration Symp., October 1988, Orlando, Florida. S. Fla. Water Manage. Dist., West Palm Beach, FL. WULLSCHLEGER, J.G.,S.J. MILLER, AND L.J. DAVIS. 1990a. An evaluation of the effects of the restoration Demonstration Project on the Kissimmee River fishes. Pp. 67-81. In: LOFTIN, M.K., L.A. TOTH, AND J. OBEYSEKERA.(eds.) Proc. Kissimmee River Restoration Symp., October 1988, Orlando, Florida. S. Fla. Water Manage. Dist., West Palm Beach, FL. , S.J. MILLER, AND L.J. DAVIS. 1990b. A survey of fish communities in Kissimmee River oxbows scheduled for phase II restoration. Pp. 143-48. In LOFTIN, M.K., TOTH, L.A., AND OBEYSEKERA, J. (eds.) Proc. Kissimmee River Restoration Symp., October 1988, Orlando, Florida. S. Fla. Water Manage. Dist., West Palm Beach, FL. Florida Scient. 56(1):25-51.1993 Accepted: June 5, 1992. 59 FLORIDA SCIENTIST [VOL 56 Conservation LOW CLUTCH VIABILITY OF AMERICAN ALLIGATORS ON LAKE APOPKA ALLAN R. WooDWaRD"”, H. FRANKLIN PERCIVAL”), MICHAEL L. JENNINGS®”, AND CLINTON T. Moore" (Florida Game and Fresh Water Fish Commission, 4005 South Main Street, Gainesville, FL 32601 U.S. Fish and Wildlife Service, Florida Cooperative Fish and Wildlife Research Unit, Department of Wildlife and Range Sciences, University of Florida, Gainesville, FL 32611 ABSTRACT—Clutch viability of American alligators was evaluated on lakes Apopka, Griffin, Jessup, and Okeechobee, Florida, during 1983-86 to examine its association with alligator population trends. Clutch viability was lower (P < 0.05) on Lake Apopka and higher (P < 0.05) on Lake Griffin than on any other lake. Annual clutch viability rates declined (P < 0.05) on Lake Apopka during the study, but no trends in viability rate were detected on other lakes. Juvenile alligator density was relatively stable during 1980-87 on lakes Griffin and Jessup, but plunged to 10% of the 1980 level on Lake Apopka (P = 0.002), coincident with falling clutch viability. Viability rates were not related to clutch size but did increase with clutch weight (P = 0.013). Egg banding rates declined on lakes Apopka and Jessup and increased on lakes Griffin and Okeechobee. Unexplained mortality of large alligators was commonly observed on Lake Apopka. We discuss several possible causes of low clutch viability including pesticide contamination, shifts in the age structure of the breeding population, and density-related stress of the adult population, and we recommend a course for further investigation. THE Florida population of the American alligator (Alligator mississippiensis) increased after federal protection in the early 1970s (Hines, 1979; Wood et al., 1985) and continued to grow on most wetlands through the 1980s (Woodward and Moore, 1989). However, Jennings and co-workers (1988) reported a rapid, severe (approx. 90%) decline of the juvenile alligator population on Lake Apopka during 1981-86 and proposed reproductive failure as a primary cause. Prior to the 1940s, Lake Apopka was a clear, open lake (12,960 ha) with an adjacent marsh (8,000 ha). However, events in the 1940s drastically altered the water quality, turning Lake Apopka into one of Florida’s most polluted (nutrient and pesticide) lakes (U.S. Environ. Prot. Agency, 1979). During the mid-late 1940s, 7285 ha of marsh on the north side of Lake Apopka was converted to vegetable farming, and much of the uplands surrounding the lake was planted with citrus (Fig. 1). Concurrently, there was an increase in effluent discharges into the southeast part of the lake from a citrus processing plant and the sewage treatment facility of the city of Winter Garden. The ecological integrity of Lake Apopka was further damaged in 1947 when a hurricane uprooted most of the aquatic vegetation in the main lake, reducing the nutrient cycling capacity of the wetland (Conrow et al., 1989). *Present address: U.S. Fish and Wildlife Service, National Ecology Research Center, 4512 McMurry Avenue, Fort Collins, CO 80525 **U_.S. Fish and Wildlife Service, Office of Migratory Bird Management, Patuxent Wildlife Research Center, Laurel, MD 20708. No. 1, 1993] WOODWARD ET. AL.—LOW CLUTCH VIABILITY OF AMERICAN ALLIGATORS Lake Beauclair f a Citrus Groves Vegetable Farms Municipalities Wooded Swamp Fic. 1. Lake Apopka, Florida and surrounding agricultural operations during the early 1980s. 54 FLORIDA SCIENTIST [VOL 56 Agricultural operations have introduced substantial pesticide loads into Lake Apopka since the early 1940s through direct pumping or through seepage into the Apopka drainage (Huffstutler et al., 1965; Florida Dept. Environ. Reg., 1979). Vegetable farms must continually remove water from fields during the wet season, and most operations flood and drain fields every 2 years for nematode control. Pesticides commonly used during the late 1970s included toxaphene, parathion, and chlorobenzilate, all of which have been restricted by the U.S. Environmental Protection Agency (U.S. EPA) because of chronic or acute effects on wildlife or fish (U.S. EPA, 1990). In 1980, an extensive spill of Kelthane, an EPA-approved pesticide primarily composed of dicofol, was reported (U.S. EPA, unpubl. rep.) at the Tower Chemical Company located between State Road 50 and Gourd Neck of Lake Apopka (Fig. 1). Few accounts exist of fish and wildlife populations on Lake Apopka and how they may have been affected by the above events. Increased build-up of unconsolidated bottom sediment and subsequent accelerated eutrophication contributed to a large fish kill in 1963 (Clugston, 1963; Huffstutler et al., 1965). Shotts and co-workers (1972) reported a die-off of fish, alligators, and Florida softshell turtles (Apalone ferox) during 1971 and attributed it to stress-induced bacterial (Aeromonas spp.) infections exacerbated by low water level. Johnson and Jenkins (1984) reported reproductive failure of largemouth bass (Micropterus salmoides) on Lake Apopka in 1982. Heinz and co-workers (1991) found Lake Apopka alligator eggs collected during 1984-85 to have elevated levels of several organochloride compounds relative to eggs from lakes Griffin and Okeechobee, but they found no direct association with clutch viability. Although alligator populations normally respond positively to eu- trophication (Wood et al., 1985), the alligator population crash reported by Jennings and co-workers (1988) indicated that conditions in Lake Apopka were unfavorable for alligator population growth during the early 1980s. Artificially incubated alligator clutches from the Jennings and co-workers (1988) study provided an opportunity for us to compare clutch viability rates among 4 Florida lakes, estimate clutch viability trends, and develop hypotheses about the possible relation of clutch viability to the alligator population status of Lake Apopka. MATERIALS AND METHODS—Fntire clutches of alligator eggs were systematically collected (see Woodward et al., 1989 for procedures) from accessible nests on lakes Apopka, Griffin (5742 ha), Jessup (4862 ha), and Okeechobee (191,223 ha) during 1983-86 as part of experimental egg and hatchling harvests (Jennings et al., 1988). Two subareas of Okeechobee were sampled; Indian Prairie Marsh on the northwest shore and Observation Shoal on the southwest portion of the lake. Clutches were transported to incubators at commercial alligator farms, incubated in either natural nest material or Paspalum notatum (see Woodward and co-workers [1989]), and maintained at 30-33 C and >95% relative humidity until hatching. Clutch allocations to the 1-3 participating farms in any year were determined primarily on logistical convenience. Prior to incubation, eggs were inspected for presence of an opaque band indicating fertility (Ferguson, 1985; Webb et al., 1987) and for early embryo viability (Woodward et al., 1989). Nonviable eggs were opened to verify status and stage of embryo mortality and then discarded. From each clutch, one egg that represented the most advanced stage of development was sacrificed to determine embryo age. Only clutches collected in 1986 were weighed (WT). Clutch size (C) was the total of all shelled and unshelled eggs found in a nest. Banding rate (B) was the number of banded eggs divided by C. Clutch viability rate (V) was the number of hatchling alligators surviving 21 day divided by C-1 (C less the sacrificed egg). Because we were mainly interested in the inherent viability of eggs, clutches from flooded or disturbed (by predators, turtles, humans, or other alligators) nests were excluded from viability analyses. However, only clutches from disturbed nests were excluded from analyses of C and B. No. 1, 1993] | WOODWARD ET. ALLOW CLUTCH VIABILITY OF AMERICAN ALLIGATORS 55 The clutch was the experimental unit in analyses of V, C, and B. None of the variables was distributed normally or with constant variance. We squared C to normalize the distribution then analyzed C? via ordinary least squares regression methods. The rate variables V and B included many values representing either complete viability failure (V = 0.0, 19% of clutches) or complete banding success (B = 1.0, 21%). No single transformation could normalize the truncated distributions of V and B. We considered the nonparametric alternative of ranking the data and performing regression analyses of the ranks (Conover and Iman, 1981), but we were concemed that the preponderance of tied values resulting from this approach would bias tests of significance. Instead, we divided the rate data into 2 components, each of which could be suitably transformed for analysis. Into the first component, we assigned all rate values not representing complete successes or failures, i.e.,00, we reassigned V the indicator value 1. Thus, each clutch was designated as either completely nonviable or partially viable. For B<1, we reassigned B the value 0 to distinguish clutches with some unbanded eggs from clutches with all banded eggs. We assumed the indicator values were binomially distributed and used logit analysis (Agresti, 1990) to estimate relative probabilities of complete/partial viability failure and banding success. Although data assigned to both components comprised transformed values of V and B, we present median and untransformed means for reference and comparison purposes. Means presented for C represent back-transformed means of C”. In all analyses, our study design was a 4-year by 5-area fixed effects factorial, but we tested for differences between Okeechobee subareas for possible pooling as 1 area. Woodward and co-workers (1989) found no differences in hatch rate among clutches incubated at 3 facilities in 1985, thus we assumed that facility effects were negligible and could be ignored. We assumed that accessible clutches constituted acompletely random sample of clutches. An effect for a rate variable was considered significant (P < 0.05) if analysis of data in either or both so indicated. If we detected no area-year interaction, we estimated linear, quadratic, and cubic contrasts for the year effect means and compared all possible pairs of area means. When areas and years did interact, we compared area means if the area effect was significant. Because analysis of trends among areas was a principal study objective, we tested whether a significant area-year interaction was due to differences in trend among areas. In this circumstance, we estimated linear contrasts of year within areas and compared all possible pairs of contrast estimates. Mean or trend pairs were declared different when P < 0.05/k (where k was the number of comparisons to be made), to guarantee that the probability of falsely declaring significant 1 of the k comparisons was no greater than 5% (Bonferroni adjustment). F-tests or G-tests were used to determine significance of effects in the factorial design for the regression and logit analyses. V was regressed on area, year, C, and WT to determine the relationship of selected clutch characteristics to clutch viability. Counts of juvenile (30-121 cm total length [TL]), 2122 cm TL, and adult ( 2183 cm TL) alligators were obtained from night-light surveys conducted during 1980-87 on Apopka, Griffin, and Jessup and provided estimates of observed densities (alligators/shoreline km). Density data were not available for Okeechobee. We regressed log-transformed density on year to estimate population trends. A water level covariate was included in the regression models to statistically remove its effect from alligator observability. Unlike Jennings and co-workers (1988), our trend analysis included alligators 30-60 cm in the juvenile size class and added 1987 data to reflect the population response to 1986 events. The observation rate of dead alligators was the number of dead alligators 2122 cm observed during night-light and tagging activities divided by the number of activity nights. REsuLTS—Clutch Viability—In Okeechobee, mean V did not vary by subarea or subarea-year combination. Complete viability failures did not occur in 2 years on Indian Prairie Marsh; thus, we could not investigate the subarea-year interaction in the logit analysis. The odds of sampling a partially nonviable clutch rather than a completely nonviable clutch was 7 times greater on Indian Prairie Marsh (28:1) than on Observation Shoal (4:1) (P = 0.002). Recognizing that complete viability failure probabilities differed within Okeechobee, we nevertheless ignored subarea distinc- 56 FLORIDA SCIENTIST [VOL 56 TABLE 1. Sample size (n), mean (x) and median (m) clutch viability, and percentage observations of zero alligator clutch viability rates (%0) for alligator clutches collected on 4 Florida lakes 1983-86. Data consist of only non-flooded and non-disturbed clutches. Year Study area 1983 1984 1985 1986 All years Lake Apopka n Wil 8 28 22 69 5 0.54 0.22 0.18 0.13 0.21 m 0.45 0.05 0.01 0.00 0.03 %O 18.2 Bihan) 50.0 59.1 46.4 Lake Griffin n 1 oF 67 61 167 X 0.53 0.65 0.59 0.60 0.60 m 0.74 0.75 0.73 O72 0.73 %O 95.0 Sal 19.4 Se 15.0 Lake Jessup n 18 32 37 40 127 c 0.36 0.54 0.34 0.63 0.48 m 0.26 0.65 0.30 0.77 0.54 %O 16.7 6.2 S24 LD 1537 Lake Okeechobee =n 15 35 42 94 116 % 0.47 0.40 0.40 0.59 0.45 m 0.46 0.44 0.44 O71 0.46 %O 6.7 14.3 16.7 235 13.8 All areas n 56 102 174 147 479 Xx 0.44 0.50 0.42 0.54 0.48 m 0.47 0.58 0.46 0.63 0.56 %O 16.1 10.8 26.4 18.4 19.4 tion and combined Indian Prairie Marsh and Observation Shoal for comparisons among main study areas. Mean and median V for 479 clutches from all 4 lakes, 1983-86, were 0.48 and 0.56 (Table 1). Mean viability varied (P < 0.001) among areas in both the regression and logit analyses. Griffin mean V’ was greater than for any other area, and Apopka mean V’ was less than for Jessup. No other mean V’ pair was different. Complete viability failures were more likely to be encountered on Apopka than on the other areas, but Griffin, Jessup, and Okeechobee were not different (Table 1). Across areas, no yearly effect was detected in the regression analysis, but a cubic trend (P = 0.025) was discovered in the logit analysis; mean probability of complete viability failure was smallest in 1984, greatest in 1985, and intermediate in 1983 and 1986. Interaction between area and year was nearly significant (P = 0.066) in the regression analysis of V’, particularly in the component that measured linear trend variability among areas (P = 0.051). However, we failed to find differences between any trend pairs even though V’ decreased (P = 0.026) with time on Apopka. No interaction was detected in the logit analysis. No. l, 1993] WOODWARD ET. AL.—LOW CLUTCH VIABILITY OF AMERICAN ALLIGATORS 57 o————0 Lake Apopka @ ---- @ Lake Griffin 4— — —A Lake Jessup 50-122 cm Alligators per km (Adjusted for Water Level) 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 Year FIG.2. Estimated trends of juvenile (30-122 cm) alligator populations on lakes Apopka, Griffin, and Jessup during 1980-87. Estimates were derived from log-transformed night-light counts, adjusted for water level, and presented on an untransformed scale. Clutch Size and Weight—Overall mean C for all areas was 45.5 eggs/clutch (Table 2). Although mean C was greater on Indian Prairie Marsh (45.4 eggs/clutch) than on Observation Shoal (41.6 eggs/clutch), we pooled data from the Okeechobee subareas as | area. C? differed (P < 0.001) among areas, and the pooled Okeechobee mean was less than either the Griffin or Jessup means (Table 2). Linear trends in C? means also varied (P = 0.026) by area; clutch size declined on Okeechobee and increased on Griffin. We detected no effect of clutch size on V’, but V increased (P = 0.001) with clutch weight. A weak (P = 0.067) negative quadratic relationship was detected, suggesting that clutch viability peaked at intermediate clutch weights, then declined for the heaviest clutch weights. TABLE 2. Back-transformed mean (x,) and untransformed mean (x) clutch sizes and mean (x) and median (m) banding rates for undisturbed alligator clutches collected from 4 Florida lakes, 1983-86. Clutch size Banding rate Study area n ae eG n x m Lake Apopka 81 45.8 45.2 80 0.70 0.89 Lake Griffin ZA: 46.1 45.8 209 0.86 0.94 Lake Jessup 161 AT, 46.5 159 0.89 0.94 Lake Okeechobee 229 43.6 42.4 229 0.85 0.95 Masa 682 155 44.8 677 0.84 0.94 58 FLORIDA SCIENTIST [VOL 56 Banding Rates—Overall median and mean banding rates were 0.94 and 0.84 (Table 2). We found no differences between Okeechobee subarea mean banding measures in either the regression or logit analysis; thus, we pooled the data. Mean B’ did not vary among areas or among years. However, we detected an area-year interaction (P = 0.014), especially in the cubic component (P = 0.005) which indicated that banding oscillated but was asynchronous among study areas. Mean probability of complete banding success varied (P = 0.021) among years and among levels of the area-year interaction (P = 0.028), but not among areas. Complete banding success probability declined on Apopka and Jessup but increased on Griffin and Okeechobee. The increasing Okeechobee trend was also greater than the increasing Griffin trend, but we could detect no difference between the Jessup and Apopka trends. Population Densities—Juvenile alligator densities were similar on Apopka and Griffin during 1980-82, but declined on Apopka during 1980-87 by 90% (28%/year, b = -0.323, 10 df, P = 0.002; Fig. 2). No trends were detected in densities of larger size classes on Apopka. Densities of juvenile alligators increased on Griffin (6.7%/ year, b = 0.065, 9 df, P = 0.029; Fig. 2) but not on Jessup. Densities of 2122 cm TL and adult alligators increased on both Griffin (28.3%/year, b = 0.080, 9 df, P < 0.01) and Jessup (211%/year, b = 0.100, 9 df, P < 0.05). Ranges of estimated adult alligator densities during 1982-86 were 0.1-3.2 alligators/km on Apopka, 1.4-3.2 on Griffin, and 1.5-3.1 on Jessup. Observed Mortality—Dead alligators were more frequently observed on Apopka (0.414 alligators/night, n = 29 nights) than on Griffin (0.032, n = 93) and Jessup (0.038, n = 26). Fish kills were common on Apopka, and, on one occasion in 1983, we observed 2 dead Florida softshell turtles, a dead, unidentified water snake (approx. 1.2 m TL), and 2 dead alligators (21.8 m). Superficial examinations of carcasses revealed no evidence of trauma caused by humans or other animals. DiscussioN—The alligator population decline on Lake Apopka reported by Jennings and co-workers (1988) and reconfirmed in our analysis coincided with rapidly declining clutch viability rates during 1983-86. Unfortunately, clutch viabil- ity rates prior to 1983 were not available for comparison with post-decline rates. However, juvenile alligator population densities on Lake Apopka during 1980-82 were similar to those observed on Lake Griffin, suggesting that clutch viability had been greater before 1983. Although harvests of eggs and hatchlings were concurrent during this study, we submit, as did Jennings and co-workers (1988), that a decline in juveniles would have been evident on all lakes had over-harvest been a factor. On the contrary, juvenile population densities were relatively stable on Lake Jessup and increased on Lake Griffin. The coincident juvenile alligator population decline and low clutch viability level on Lake Apopka suggest that poor reproductive success contributed to the alligator population decline. Median viability rates among lakes in our study were extremely variable and ranged from 0.73 on Lake Griffin to 0.03 on Lake Apopka. Clutch viability was higher on Lake Griffin and lower on Lake Apopka than on other lakes and did not differ No. 1, 1993] | WOODWARD ET. AL.—LOW CLUTCH VIABILITY OF AMERICAN ALLIGATORS 59 between lakes Jessup and Okeechobee. Clutch viability also differed within Lake Okeechobee, being lower on Observation Shoal than on Indian Prairie Marsh. We did not detect clutch viability trends on lakes Griffin, Jessup, and Okeechobee, but we found strong evidence of annual fluctuations. Viability rates of clutches in this study were much lower than viability rates reported for artificially incubated wild produced eggs from Rockefeller Refuge, Louisiana (V = 0.86; Joanen and McNease, 1987) or for eggs collected from Texas coastal marshes (V = 0.83; Johnson et al., 1989). No other published viability data are available for artificially incubated wild alligator eggs. We note that our rates were likely biased low, but by no more than 0.03, because we sacrificed the best-developed egg in each clutch for age determination and excluded it from analyses. The causes of low and decreasing clutch viability on Lake Apopka are unclear. Woodward and co-workers (1989) found no embryo age-specific differences in hatch rates of alligator clutches when collected and handled under careful procedures. They also presented evidence that substantial mortality of alligator embryos from lakes Griffin, Jessup, and Okeechobee occurred prior to collections and continued through incubation, independent of collection time. We found a decreasing trend in banding rate on Lake Apopka, and an average of 30% of each clutch was unbanded (Table 2), indicating substantial infertility or pre-oviposition embryonic death. Thus, we believe that major factors contributing to low egg viability are manifest prior to embryo attachment to the shell membrane. Some factors that may influence egg viability are female age (Ferguson, 1985), density-related stress (Joanen and McNease, 1989; Elsey et al., 1990); adult nutri- tional status (Joanen and McNease, 1989), toxicosis through exposure to environ- mental contaminants (Clark, 1990), nest flooding (Joanen et al., 1977), and extremes in pre-collection clutch cavity temperatures (Joanen and McNease, 1987; Webb and Cooper-Preston, 1989). We were able to eliminate flooding and disturbance as sources of variation by selecting unflooded and undisturbed nests. However, we could not sample and control early nest temperature nor obtain dietary data for assessment of adult nutritional status. Population densities of adult alligators were not greater on Lake Apopka than on lakes Griffin and Jessup during 1982-86. Superficially, this does not support the hypothesis that density-related stress caused low clutch viability on Lake Apopka. Ferguson (1985) and Joanen and McNease (1989) reported that the youngest and oldest females produce the least viable clutches, and Wilkinson (1983) and Ferguson (1985) found an association between number of eggs per clutch and female body size. Clutch size means did not differ among lakes Apopka, Jessup, and Griffin, but were noticeably greater than means reported for most other alligator populations (Metzen, 1977; Goodwin and Marion, 1978; Deitz and Hines, 1980: Ruckel and Steele, 1984; Carbonneau, 1987; Joanen and McNease, 1989; Kushlan and Jacobsen, 1990). Mean clutch size for South Carolina coastal impoundments was comparable to our findings (Wilkinson, 1983). These comparisons imply that nesting females in our study areas were larger and possibly older than those of other alligator popula- tions, although site-specific factors (e.g., nutrition and genetics) may affect the 60 FLORIDA SCIENTIST [VOL 56 female size/clutch size relationship. Clutch weight rather than clutch size may be more closely associated with female size (Hall, 1990). Clutch weight was less on Lake Apopka than the other lakes. The positive linear and negative quadratic relationship between viability and clutch weight implies that greater clutch weight enhances clutch viability, but that the rate of enhancement decreases for the greatest clutch weights (oldest females). There- fore, lower clutch viability on Lake Apopka may have resulted from a predominance of both younger and older, less productive females. Our clutch weight sample size was small and limited to 1 year (1986), but these findings identify an area that deserves more investigation. Although we detected no decline in population densities of alligators 2122 cm on Lake Apopka, population growth apparently stagnated while population densities on lakes Griffin and Jessup increased sharply. The frequent occurrence of dead alligators may further reflect chronic underlying problems in the lake. Mortality of alligators, softshell turtles, and fish on Lake Apopka during 1971 and 1972 was attributed to bacterial (Aeromonas spp.) infections induced by drought-related stress (Shotts et al., 1972). Aeromonas are common organisms in Florida wetlands and usually are non-pathogenic. However, when conditions sufficiently stress alligators, Aeromonas can proliferate and cause mortality (Shotts et al., 1972). Although Lake Apopka water levels were relatively stable during our study, fish kills and alligator mortality were common. Therefore, low water levels probably did not cause the decline. Surrounding agricultural operations have introduced considerable amounts of pesticides into Lake Apopka since the 1940s (Huffstutler et al., 1965; U.S. EPA, 1979). Heinz and co-workers (1991) found no association between selected pesticide levels and clutch viability, but emphasized that toxic chemicals could adversely affect adult alligator reproductive endocrinology in ways that may not be measured in the eggs. If such a phenomenon was at work, we would have expected chronic low reproductive success rather than an apparent acute problem. The strong association between the 1980 Kelthane spill and subsequent declines in the juvenile alligator population and egg viability provides the best clue to the cause of acute reproductive failure. Although the full impact of the 1980 Kelthane spill is not yet known, the maximum effect of the organochloride pesticide and its breakdown metabolites, DDD and DDE (Clark, 1990), would have likely occurred during the early 1980s. Fish kills resulting from the advanced eutrophic state of Lake Apopka may have been amplified by periodic release of toxicants from bottom sediments. Alligators commonly fed on dead fish and, as a consequence, may have been exposed to occasional high concentrations of toxicants. In high concentrations, dicofol can be lethal to vertebrates, and in lower doses it can cause reproductive impairment (Clark, 1990). Thus, the high incidence of dead alligators and low clutch viability could have been caused by the Kelthane spill. The implications of low alligator clutch viability are important from ecological, aesthetic, and economic perspectives. Alligators are top carnivores (Delany and Abercrombie, 1986; Delany, 1990) in the aquatic food web on Florida lakes and may No. l, 1993] WOODWAED ET. AL.—LOW CLUTCH VIABILITY OF AMERICAN ALLIGATORS 61 play an important role in maintaining a balance in prey species populations. Furthermore, most Floridians value alligators for aesthetic reasons (Hines and Scheaffer, 1977; Delany et al., 1986). Lower alligator densities on Lake Apopka will reduce opportunities for the public to observe alligators. Depressed populations and productivity also will have a negative effect on sustained-yield alligator hunting and egg harvests. We estimate that the harvest potential of alligators in Lake Apopka has been reduced by 150 large alligators (@ $400 each) and 1200 hatchlings per year (@ $15 each) for an estimated loss in wholesale value of $78,000 per year. The effects of lake degradation on fish and wildlife may be expressed through acute mortality or more subtly through depressed reproduction, poor growth, and low survival. Factors contributing to low alligator clutch viability on Lake Apopka may have contributed to mortality of fish and turtles (Shotts et al., 1972), poor reproductive success in largemouth bass (Johnson and Jenkins, 1984), and may have negatively affected other, less conspicuous vertebrates that depend on wetlands for food. Adverse effects of environmental degradation may not be limited to Lake Apopka. Low alligator clutch viability rates observed on lakes Griffin, Jessup, and Okeechobee relative to rates reported for Louisiana and Texas may indicate chronic reproductive problems on those wetlands as well. Our clutch viability comparisons among areas and trend estimations over years may have suffered from the variation sources of uncontrollable error, sampling bias, and inadequate sample size. Therefore, we may not have been able to detect differences that actually occurred among areas and years. We recommend continu- ing clutch viability investigations and making special efforts to reduce sources of variation and sampling bias by carefully controlling collection and handling proce- dures, incubating eggs under standardized conditions at a common facility, and increasing clutch sample sizes to increase statistical power. Further studies should address relationships of clutch viability to adult nutritional status, population density, nest characteristics, nest temperatures, female age, and toxic contaminant levels. We suggest broadening the sample of wetlands to gain a better idea of the range in clutch viability rates throughout Florida. ACKNOWLEDGMENTS—We thank the many employees and volunteers associated with the Florida Cooperative Fish and Wildlife Research Unit (FCFWRU), the Florida Game and Fresh Water Fish Commission (GFC), and the Florida Alligator Farmers Association (FAFA) for assistance in egg collections. In particular we thank J. D. Ashley (FAFA) and T. C. Hines for their contributions to the design and implementation of the study and A. T. Bush and J. T. DeFazio (FCFWRU) for technical assistance. We also thank F. Godwin (Gatorland Zoo), E. Froehlich (Froehlich’s Gator Farm), D. Morgan (C.S.T. Gator Farm), L. Wells (Hilltop Farms), G. O. Parrott (Flying P Ranch) and their staffs for incubating clutches. J. R. Brady, D. N. David, L. J. Guillette, Jr., W. E. Johnson, G. R. Masson, P. E. Moler, and K. G. Rice provided critical reviews of the manuscript. The GFC, U. S. Fish and Wildlife Service, FAFA, and the University of Florida provided funding. LITERATURE CITED AGRESTI, A. 1990. Categorical data analysis. John Wiley and Sons, New York, N.Y. 558 pp. BOX, G. E. P. AND D. R. Cox. 1964. An analysis of transformations. J. Royal Stat. Soc., Ser. B. 26:211- 62 FLORIDA SCIENTIST [VOL 56 243. CARBONNEAU, D. A. 1987. Nesting ecology of an American alligator population in a freshwater coastal marsh. M.S. Thesis, La. State Univ., Baton Rouge. 53 pp. CLARK, D. R. 1990. Dicofol (Kelthane) as an environmental contaminant: a review. U.S. Fish and Wildl. Serv. Tech. Rep. 29. 48 pp. CLUGSTON, J. P. 1963. Lake Apopka, Florida, a changing lake and its vegetation. Quart. J. Fla. Acad. Sci. 26:168-174. CONOVER, W. J., AND R. L. IMAN. 1981. Rank transformations as a bridge between parametric and nonparametric statistics. Am. Statistician 35:124-129. CONROW, R., W. GODWIN, M. F. COVENEY, AND L. E. BATTOE. 1989. SWIM plan for Lake Apopka. St. Johns Water Manage. Dist., Palatka, Florida. 119 pp. DEITZ, D. C. AND T. C. HINES. 1980. Alligator nesting in north-central Florida. Copeia 1980:249-258. DELANY, M. F. 1990. Late summer diet of juvenile American alligators. J. Herpetol. 24:418-421. ____ AND C. L. ABERCROMBIE. 1986. American alligator food habits in northcentral Florida. J. Wildl. Manage. 50:348-353. ____ T. C. HINES, AND C. L. ABERCROMBIE. 1986. Selected public’s reaction following harvest of American alligators. Proc. Ann. Conf. Southeast. Assoc. Fish and Wildl. Agencies 40:349-352. ELSEY, R. M., T. JOANEN, L. MCNEASE, AND V. LANCE. 1990. Stress and plasma corticosterone levels in the American alligator — relationships, stocking density, and nesting success. Comp. Biochem. Physiol. 95a:55-63. FERGUSON, M. W. J. 1985. Reproductive biology and embryology of the crocodilians. Pp 329-491 In: GANS, C., F. BILLET, AND P. MADERSON, (eds.) Biology of the Reptilia. Vol. 14. John Wiley and Sons, New York, N.Y. FLORIDA DEPT. ENVIRON. REG. 1979. Agricultural nonpoint source element: state water quality management plan — upper Oklawaha River Basin. Vol. II — Tech. Appendix VI. Tallahassee, Fla. 76 pp. Cent M. AND W. R. MARION. 1978. Aspects of the nesting ecology of American alligators (Alligator mississippiensis) in north-central Florida. Herpetologica 34:43-47. HALL, P. M. 1990. Demographics of selected crocodilian populations based upon the patterns of harvest, reproduction, and skull morphometry: Case studies for Crocodylus novaeguineae, C.porosus, and Alligator mississippiensis. Ph.D. Thesis, Univ. of Fla., Gainesville. 224 pp. HEINZ, G. H., H. F. PERCIVAL, AND M. L. JENNINGS. 1991. Contaminants in American alligator eggs from lakes Apopka, Griffin, and Okeechobee, Florida. Environ. Monit. Assess. 16:277-285. HINES, T. C. 1979. The past and present status of the American alligator in Florida. Proc. Ann. Conf. Southeast. Assoc. Fish and Wildl. Agencies 33:224-232. _____ AND R. SCHEAFFER. 1977. Public opinion about alligators in Florida. Proc. Ann. Conf. Southeast. Assoc. Fish and Wildl. Agencies 31:84-89. HUFFSTUTLER, K. K., J. E. BURGESS, AND B. B. GLENN. 1965. Biological, physical, and chemical study of Lake Apopka, 1962-1964. Fla. St. Board Health. Jacksonville. 56 pp. JENNINGS, M. L., H. F. PERCIVAL, AND A. R. WOODWARD. 1988. Evaluation of alligator hatchling and egg removal from three Florida lakes. Proc. Ann. Conf. Southeast. Assoc. Fish and Wildl. Agencies 42:283-294. JOANEN, T. AND L. MCNEASE. 1987. Alligator farming research in Louisiana, U.S.A. Pp. 329-340 In: WEBB, G. J. W., S.C. MANOLIS, AND P. J. WHITEHEAD,(eds.) Wildlife management: crocodiles and alligators. Surrey Beatty and Sons, Chipping Norton, N.S.W., Aust. ____ AND L. MCNEASE. 1989. Ecology and physiology of nesting and early development of the American alligator. Amer. Zool. 29:987-998. ____L. MCNEASE, AND G. PERRY. 1977. Effects of simulated flooding on alligator eggs. Proc. Ann. Conf. Southeast. Assoc. Fish and Wildl. Agencies 31:33-35. JOHNSON, L. A., A. COOPER, B. THOMPSON, AND R. WICKWIRE. 1989. Texas alligator survey, harvest, and nuisance summary, 1988. Ann. Rep., Texas Parks and Wildl., Austin. 19 pp. JOHNSON, W. E. AND L. J. JENKINS. 1984. Lake Apopka predator sportfish investigations. Final Rep., Florida Game and Fresh Water Fish Comm., Eustis. 41 pp. KUSHLAN, J. A. AND T. JACOBSEN. 1990. Environmental variability and the reproductive success of Everglades alligators. J. Herpetol. 24:176-184. METZEN, W. 1977. Nesting ecology of alligators on the Okefenokee National Wildlife Refuge. Proc. Ann. Conf. Southeast. Assoc. Fish and Wildl. Agencies 31:29-32. RUCKEL, S. W., AND G. W. STEELE. 1984. Alligator nesting ecology in two habitats in southern Georgia. Proc. Ann. Conf. Southeast. Assoc. Fish and Wildl. Agencies 38:212-221. SHOTTS, E. B., JR., J. L. GAINES, JR., L. MARTIN, AND A. K. PRESTWOOD. 1972. Aeromonas-induced No. 1, 1993] 63 deaths among fish and reptiles in an eutrophic inland lake. J. Amer. Vet. Med. Assoc. 161:603-607. U. S. ENVIRONMENTAL PROTECTION AGENCY. 1979. Environmental impact statement. Lake Apopka restoration project, Lake and Orange counties, Florida. U.S. EPA, Region 4, Atlanta, Ga. 445 pp. ____. 1990. Suspended, canceled, and restricted pesticides. U.S. EPA Tech. Booklet. Washington D. C. WEBB, G. J. W. AND H. COOPER-PRESTON. 1989. Effects of incubation temperature on crocodiles and the evolution of reptilian oviparity. Amer. Zool. 29:953-971. ___, S.C. MANOLIS, P. J. WHITEHEAD, AND K. DEMPSEY. 1987. The possible relationship between embryo orientation, opaque banding and the dehydration of albumen in crocodile eggs. Copeia 1987:252-257. WILKINSON, P. M. 1983. Nesting ecology of the American alligator in coastal South Carolina. Study Completion Rep., S.C. Wildl. and Marine Resource. Dept., Columbia, S.C. 113 pp. WOOD, J. M., A. R. WOODWARD, S. R. HUMPHREY, AND T. C. HINES. 1985. Night counts as an index of American alligator population trends. Wildl. Soc. Bull. 13:262-272. WoopwakrbD, A. R.ANDC.T. MOORE. 1989. Statewide alligator surveys. Final Rep., Fla. Game and Fresh Water Fish Comm., Tallahassee. 24 pp. ___,M.L. JENNINGS, AND H. F. PERCIVAL. 1989. Egg collecting and hatch rates of American alligator eggs in Florida. Wildl. Soc. Bull. 17:124-130. Florida Scient. 56(1):52-63. 1993 Accepted: September 1, 1992 FIRST RECORD OF THE EASTERN BIG-EARED BAT (PLECOTUS RAFINESQUID) IN SOUTHERN FLORIDA — Larry N. Brown’? anp Curtis K. BRown”’, ‘Environmental Studies, Inc., 2410 Old Monticello Road, Thomasville, Georgia 31792 and ” Florida Game and Fresh Water Fish Commission, LaBelle, Florida 33935 ABSTRACT: The first specimen of the eastern big-eared bat, Plecotus rafinesquii, taken in southern Florida, was recorded roosting in an old cabin in northeastern Collier County, a short distance south of the Seminole Indian Reservation. It was an adult female, and this occurrence extends the known range of the eastern big-eared bat significantly southward from central Florida to the Big Cypress Swamp region of south-western Florida. BROWN (1974) first reported the presence of Plecotus rafinesquii as far south as central Florida. The overall status of the species was later summarized by Brown (1978). Asubsequent update on rare and endangered biota in Florida, by Humphrey (1992) failed to include the known records for Plecotus rafinesquii from central Florida. Brown (1993) corrects this oversight, and summarizes current life history and distribution information for the species in Florida. This bat tends to live in small colonies often associated with pine flatwood forests and abandoned cabins or hollow trees in the forest. 64 FLORIDA SCIENTIST [VOL 56 On October 30, 1992, an adult female Plecotus rafinesquii was located roosting in an old abandoned cabin situated in the Big Cypress Swamp region of northeastern Collier County. The cabin, made entirely of cypress logs, had an open from door which allowed access by bats and other wildlife. The big-eared bat was captured by hand after being aroused to activity from its roosting site in the rafters, by the activity of camping deer hunters. The ecological habitat surrounding the cabin is dominated by cypress trees interspersed with a few scattered oaks and other hardwoods. The entire cypress strand is situated immediately south of the Big Cypress Seminole Indian Reservation and adjacent to the L-28 Canal in northeastern Collier County. This bat record extends the known geographical range of the species approxi- mately 130 miles southward into the Big Cypress Swamp region of southwestern Florida. The external measurements of the adult female were as follows: Total length-99 mm, Tail length-48 mm, Hindfoot length-10 mm, Ear length-34.5 mm, and Forearm length-42 mm. This specimen was deposited in the U.S. National Museum, catalog number USNM-56242. LITERATURE CITED BROWN, L. N. 1974. Southern extension of the range of Plecotus rafinesquii. Bat Research News. 15:7-8. BROWN, L. N. 1978. Southeastern big-eared bat, Plecotus rafinesquii. Pp. 35. In J. N. Layne, ed., Rare and endangered biota of Florida, Vol I. Mammals. University Press of Florida, Gainesville. 52 Pp. BROWN, L. N. 1993. Mammals of Florida. Windard Publ., Inc., Miami, FL, 230 pp. (In Press). HUMPHREY, S. R., (ed.) 1992. Rare and endangered biota of Florida, Vol I, Mammals. University Press of Florida, Gainesville. 392 pp. Florida Scient. 56(1):63-64. 1993. Accepted: Febraury 9, 1993. INSTRUCTIONS TO AUTHORS Individuals who publish in the Florida Scientist must be active members in the Florida Academy of Sciences. Submit a typewritten original and two copies of the text, illustrations, and tables. All typewrit- ten material—including the abstract, literature citations, footnotes, tables, and figure legends— shall be double-spaced. Use one side of 81/2 x 11 inch (21!/2 cm X 28 cm) good quality bond paper for the original; the copy may be xeroxed. Margins should be at least 3 cm all around. Number the pages through the Literature Cited section. Avoid footnotes and do not use mimeo, slick, erasable, or ruled paper. Use metric units for all measurements. Assistance with production costs will be negotiated directly with authors of pages which exceed 10 printed pages. Excess over 10 pages by an author in a single volume will result in charges for additional pages. AppRESS follows the author’s address. ABSTRACT—AIl manuscripts shall have a short, concise, single-paragraphed abstract. The abstract follows immediately the author’s address. ACKNOWLEDGMENTS are given in the body of the text preceding immediately the Literature Cited section. LITERATURE CITED section follows the text. Double-space every line and follow the format in the current issue. MANUuscRIPTS of 5 or less, double-spaced typewritten pages should conform to the short arti- cle protocol. See recent issue for proper format of both long and short articles. TABLES shall be typed on separate sheets of paper, double-spaced throughout. Each table shall contain a short heading. Do not use vertical rulings. Maximum character width, including spaces, of tables is 98.0. This includes a minimum space of 5 characters between columns. Tables are charged to authors at $45.00 per page or fraction. ILLUSTRATIONS—AIl drawings shall be done in good quality India ink, on good board or draft- ing paper. Letter by using a lettering guide or equivalent. Typewritten letters on illustrations are unacceptable. Drawings and photographs should be large enough to allow 1/3 to 1/2 reduction in size. Photographs shall be glossy prints of good contrast. Whenever possible, mount photographs in lots size. The author’s name and figure number should be penciled lightly on the back of each figure. Figure legends must be listed on a separate page and not on the drawing or photograph. The leg- end must be double-spaced. Illustrations are charged to authors at $25.00 per page or fraction. We ask authors to prepare accepted manuscripts in personal computer formats. Please do not submit a manuscript on disk. Once reviewers have recommended a manuscript be accept- ed, you will be given directions for submitting the final manuscript on disk. PROOF must be returned promptly. Notification of address change and proofreading are the author’s responsibility. Alterations after the type has been set will be charged to the author. REPRINTS may be ordered from the printer on forms provided when the corrected proofs are returned to the Editor. FLORIDA ACADEMY OF SCIENCES INSTITUTIONAL MEMBERS FOR 1993 Archibold Biological Station St. Petersburg (Florida) Public Library Brevard Community College, Florida Advanced U.S. Department of Commerce, National Technology Center Library, Palm Bay Marine Fisheries (Panama City) Florida Institute of Technology U.S. Environmental Protection Agency Library, Harbor Branch Oceanographic Institution, Inc. Gulf Breeze, FL Newfound Harbor Marine Institute at Sea University of Central Florida, Division of Camp Sponsored Research St. John’s River Community College, Palatka University of South Florida Corporate Members Lockheed Space Operations, Titusville Membership applications, subscriptions, renewals, changes of address, and orders for back numbers should be addressed to the Executive Secretary, Florida Academy of Sciences, P.O. Box 033012, Indialantic, FL 32903-0012 PUBLICATIONS FOR SALE by the Florida Academy of Sciences Complete sets. Broken sets. Individual numbers. Immediate delivery. A few numbers reprinted by photo-offset. All prices strictly net. Special bargain: $500 for nearly complete sets. Prices quoted include domestic postage. PROCEEDINGS OF THE FLORIDA ACADEMY OF SCIENCES (1936-1944) Volumes 1-7—$10.00 per volume; single numbers $3.50 QUARTERLY JOURNAL OF THE FLORIDA ACADEMIC OF SCIENCE (1945-1972) Volumes 8-35—$10.00 per volume; single numbers $3.50 FLORIDA SCIENTIST (1973-75) Volumes 36-38—$10.00 per volume; single issues $3.50 except for symposium numbers priced separately. Volumes 39-47 —$13.00 per volume; single numbers $4.25 except for symposium numbers priced separately. Volumes 48 onward—$20.00/volume; single numbers $5.00 except for symposium numbers priced separately. Florida’s Estuaries—Management or Mismanagement?—Academy Symposium FLORIDA SCIENTIST 37(4)—$5.00 Land Spreading of Secondary Effluent—Academy Symposium FLORIDA SCIENTIST 38(4)—$5.00 Solar Energy—Academy Symposium FLORIDA SCIENTIST 39(3)—$5.00 (includes do-it-yourself instructions) Anthropology—Academy Symposium FLORIDA SCIENTIST 43(3)—$7.50 Shark Biology—Academy Symposium FLORIDA SCIENTIST 45(1)—$8.00 Future of the Indian River System—Academy Symposium FLORIDA SCIENTIST 46(3/4)—$15.00 Second Indian River Research Symposium FLORIDA SCIENTIST 53(3)—$15.00 Individual orders should be sent with payment. A statement will be sent in re- sponse to a bona fide purchase order over $10.00 from a recognized institution. Address all orders to: The Florida Academy of Sciences, Inc. P.O. Box 033012 Indialantic, Florida 32903-0012 Telephone: (407) 723-6835 Florida clentist — om WY Volume 56 Spring, 1993 CONTENTS Cattle Fatalities From Prolonged Exposure to Aedes Taenior STUD TETRIS DENIS TYG B VALS see RPE David S. Addison and Scott A. Ritchie Habitat Structure and the Dispersion of Gopher Tortoises on a Nature PPS TE IS wedla lle Re ae Re en OE RET ORE Oe RY Ae M.C. Stewart, D. F. Austin and G.R. Bourne Natural History of a Small Population of Leiocephalus Schreibersii (Sauria: Tropiduridae) from Altered Habitat in the Dominican Republic ........... Michael C. Schreiber, Robert Powell, John S. Parmerlee Jr., | Amy Lathrop and Donald D. Smith LE ES ce ne ee ene James N. Layne Trends in Numbers of Loggerhead Shrikes on Roadside Censuses in Pemintanarseoracla. U9 (AR 1G ccs ke ctsebell ooh neha Bl ocdavsoseetvenecudeteas Reuven Yosef, James N. Layne and Fred E. Lohrer Adsorption of Several Atmospheric Polluting Gases Upon Dehydrated SS SNERIMP re eae cre ee hse ds ade sated dna atom sd tine aeesvnisnallt iy sce auagleaties Robert F. Benson and George D. Blyholder Scaled Chrysophyceae and Synurophyceae From Florida: IV. The Flora of Memer take Myakka and Wake Tarpon ...0..2.:.-.cscesesscseesvesressendesessecsesnevess Peter A. Siver and Daniel E. Wujek Predation on Artificial Ground Nests in Southwest Florida ..............cccccceee OE eB FE I ely ora asic vac hse ia winhelsaibiawSbsadwhunevarbenes mmmeeeetrd Of |Pi(dien)X|Y¥ Complexes |...........:...-.s:-asccscesetesstssvensecetenecseronee Jay W. Palmer, William E. Swartz Jr., David King and Joseph A. Stanko Book Review SSPE TH THT SHEET ETTEHESEESEEEEH EET ES HEHE TEETH HEE THESE SH HEHEHE EEHHESHH EHS EH EOE EH HET EHH HEHEHE HOHE ETE HOES ISSN: 0098-4590 65 70 82 91 92 98 109 118 122 123 128 QUARTERLY JOURNAL OF THE FLORIDA ACADEMY OF SCIENCES FLORIDA SCIENTIST QUARTERLY JOURNAL OF THE FLORIDA ACADEMY OF SCIENCES Copyright© by the Florida Academy of Sciences, Inc. 1993 Editor Dr. DEAN MARTIN Co-Editor: Mrs. BARBARA B. MARTIN Institute for Environment Studies Department of Chemistry University of South Florida 4202 East Fowler Avenue Tampa, Florida 33620-5250 THE FLORIDA SCIENTIST is published quarterly by the Florida Academy of Sciences, Inc., a non-profit scientific and educational association. Membership is open to indi- viduals or institutions interested in supporting science in its broadest sense. Applica- tions may be obtained from the Executive Secretary. Both individual and institutional members receive a subscription to the FLORIDA SCIENTIST. Direct subscription is avail- able at $40.00 per calendar year. Original articles containing new knowledge, or new interpretation of knowledge, are welcomed in any field of Science as represented by the sections of the Academy, viz., Biological Sciences, Conservation, Earth and Planetary Sciences, Medical Sciences, Physical Sciences, Science Teaching, and Social Sciences. Also, contributions will be considered which present new applications of scientific knowledge to practical problems within fields of interest to the Academy. Articles must not duplicate in any substantial way material that is published elsewhere. Contributions are accepted only from members of the Academy and so papers submitted by non-members will be accepted only after the authors join the Academy. Instructions for preparation of manuscripts are inside the back cover. Officers for 1992-93 FLORIDA ACADEMY OF SCIENCES Founded 1936 President: Dr. PATRICK J. GLEASON Treasurer: Dr. MARVIN Ivey 1131 North Parkway 14452 Hillview Drive Lake Worth, Florida 33460 Largo, Florida 34644 President-Elect: Dr. Patricia M. Dooris Executive Secretary: Mrs. BETTY PREECE Institute for Environmental Studies P.O. Box 033012 Department of Chemistry Indialantic, Florida 32903-0012 University of South Florida Tel: (407)723-6835 Tampa, FL 33620-5250 Program Chair: Dr. DEL DELUMYEA Secretary: Ms. MarcELA GUITERREZ-MAYKA Millar Wilson Laboratory for Chemical Research 701 E. River Dr. Jacksonville University Temple Terrace, FL 33617 Jacksonville, Florida 32211 Published by the Florida Academy of Sciences, Inc. P.O. Box 033012 Indialantic, Florida 32903-0012 Printing by C & D Printing Company, St. Petersburg, FL Florida Scientist QUARTERLY JOURNAL OF THE FLORIDA ACADEMY OF SCIENCES DEAN F. Martin, Editor BARBARA B. MartTIN, Co-Editor Volume 56 Spring, 1993 Number 2 Biological Sciences CATTLE FATALITIES FROM PROLONGED EXPOSURE TO AEDES TAENIORHYNCHUS IN SOUTHWEST FLORIDA Davip S. ADDISON AND Scott A. RITCHIE, Environmental Protection Division, The Conservancy, Inc., 1450 Merrihue Drive, Naples. FL 33942. USA ABSTRACT: On 25 July 1988 a veterinary autopsy attributed the deaths of 3 cattle to prolonged exposure to mass feeding by mosquitoes. During the winter and spring of 1988 an unseasonal delay in the rainy season in coastal mangrove areas resulted in high populations of Aedes taeniorhynchus (Wiedemann). The presence of large numbers of Ae. taeniorhynchus was reflected in data collected on larval and adult mosquito populations within 5 km of the pasture where the cows died. This report suggests that prolonged exposure to large numbers of Ae. taeniorhynchus caused these deaths. THE scientific literature indicates that livestock deaths (Hearle, 1938; Sanders et al., 1968; Gillett, 1972) and weight loss (Steelman et al., 1972) can result from massive mosquito attacks. Abbitt and Abbitt (1981) recorded the deaths of 34 cattle pastured in low areas adjacent to a salt marsh in Brazoria County, Tex. This occurrence was attributed to exsanguination by Aedes sollicitans (Walker). The area was experiencing a severe drought when unusually high tides resulting from the close passage of hurricane Allen caused extensive flooding, triggering an explosive out- break of Ae. sollicitans. During May, June and July 1962 the deaths of 1,550 cattle along the gulf coast of Texas and Louisiana were attributed to Ae. sollicitans by local ranchers (Hoffman and McDuffie, 1962). In 1932 near Miami feeding by Psorophora columbiae (Dyar and Knab) was reported to have caused the deaths of 137 animals, of which 80 were cattle (Bishopp, 1933). The cause of death was considered to be exsanguination rather than suffocation. Although Aedes taeniorhynchus (Wiedemann) population cycles can result in enormous numbers of adult mosquitoes in coastal areas, reports implicating this species in livestock deaths could not be found in the literature. end Institute of Medical Research, The Bancroft Center, 300 Herston Rd., Brisbane, Queensland 4029, Australia 66 FLORIDA SCIENTIST [VOL 56 In each of the above cases, conclusions were based upon retrospective data. Similarly, the etiology of the cattle deaths reported in this paper is supported by data which retrospectively indicates that the pathology of the cattle was similar to deaths caused by mosquitoes in other cases, that the mosquito populations were high during the weeks prior to the cattle deaths, and that the vast majority were Ae. taeniorhynchus. MATERIALS AND METHODS—During the summer of 1988 a herd of approximately 60 mixed angus beef cattle (Langford, 1988) were pastured in a low-lying area adjacent to the eastern boundary of Rookery Bay National Estuarine Research Reserve, a 3,400 ha preserve dominated by mangrove forested wetlands. By chance, the pasture where the cows died is located about 5 km from mosquito surveillance stations maintained by the Collier Mosquito Control District. On 25 July 1988 the acute deaths of 3 cattle were reported. The autopsies were routine procedures performed by large animal veterinarians. The data compiled for this paper were not collected as part of any specific study, but were available as a result of routine summer mosquito surveillance activities by the Collier Mosquito Control District. At one station a single CDC light trap was baited with CO, and operated 1 night per week with the exception of the night of 7 July when it malfunctioned. Mosquitoes captured were counted by species. At another site, a single New Jersey light trap was operated 4 nights a week with the exception of the first week in July. It had a 25 watt bulb and was not baited. The total number of mosquitoes collected were counted. Landing rate counts were determined according to Ritchie (1984). Larval counts were conducted several times each week. As Ae. taeniorhynchus is the only floodwater mosquito recovered in Collier County mangrove forests (Ritchie and Johnson, 1991), all larvae were assumed to be Ae. taeniorhynchus. Counts were determined by taking a minimum of 10 dips at each site with a 350 ml dipper. The range of the larvae counted in each series of dips was recorded. Rainfall data were recorded from 3 rain gauges maintained by the South Florida Water Management District. They were located 3.2, 8.4 and 12.9 km from the pasture. RESULTS—As was the case in previously documented cattle deaths due to mass mosquito attack, the dead cattle were on the outer perimeter of a circle where they had slept the night before (Abbitt and Abbitt, 1981). They ranged in age from 8 months to 3 years, were not emaciated, and showed no outward signs of trauma. The veterinary autopsy reported that the quadraceptes and semitendinosus muscles appeared to be void of blood. The lungs were nearly void of blood, but were otherwise normal. The nares and trachea of each cow had 5 to 10 mosquitoes (not identified) in the lumen. The attending veterinarian concluded that the acute death of these cattle was due to exsanguination and suffocation resulting from prolonged exposure to biting mosquitoes (Randall, 1988). Weather conditions in 1988 were conducive to explosive outbreaks of Ae. taeniorhynchus. In coastal areas, mean rainfall for May was 4.6 cm, 57.4% below normal. These conditions, coupled with high spring tides, resulted in alternating cycles of flooding and drying in mangrove forests and a buildup of Ae. taeniorhynchus populations. This was compounded by an unseasonal delay in the rainy season; the same 3 gauges had a mean of 6.1 cm of rain in June, 71.3% below normal. Excepting tidal flooding of some sites in early June, the mean number of the sites inspected in mangrove forests in June which were dry and available for oviposition was 85.0%. Regular heavy rains in the first 2 weeks of July (mean of 12.0 cm for 3 gauges) flooded 100% of the inspected mangrove forests by July 15. From July 1 to 15, significant larval broods were evident in mangrove forests less than 5 km from the pasture (Table 1); on July 14 the authors recorded 19.9 larvae/dip (n=50) at the Rookery Bay Reserve. No. 2, 1993] ADDISON AND RITCHIE—CATTLE FATALITIES 67 TABLE 1. Larval and adult mosquito counts within 5 km of pasture where cattle died. Date Larval counts Landing rate counts (LARC) % Sites with larvae? Range (no/dip) %Sites with LRC>10/min.* Range June 22 0.0 25.0 12-50 June 23 0.0 25.0 25-50 June 27 0.0 6.2 29 June 29 0.0 31.2 25 July 1 6.2 0-2 1.2 12-25 July 5 25.0 0-10 Ba.0 13-25 July 7 25.0 0-10 July 8 25.0 0-50 18.8 16-25 July 9 SBBP 14 July 11 6.2 17-25 6.2 20 July 13 43.8 0-25 62.5 18-50 July 15 Sie 0-10 50.0 17-50 July 17 33.3° 25 July 18 6.2 1-3 31.2 50-500 July 19 100.0° 11-25 July 20 SIE? 0-5 56.2 19-50 July 21 100.0 25 July 22 18.8 1-3 50.0 100-200 July 23 100.0° 22-25 *16 total sites. >Data from 3 sites only. Landing rate counts recorded adjacent to Rookery Bay on July 17 to 23 showed extremely high numbers of biting mosquitoes (Table 1). The New Jersey light trap located next to Rookery Bay documented significant numbers of mosquitoes from July 12 to 20 (Table 2), most of which were Ae. taeniorhynchus (Bonvechio, 1988). In June and July, the CDC light trap adjacent to Rookery Bay collected primarily Ae. taeniorhynchus (Table 2). Although significant numbers of Culex nigripalpus 68 FLORIDA SCIENTIST [VOL 56 TABLE 2. Mosquito trap data within 5 km of pasture where cows died. Date NJ light trap* CDC light trap Aedes taeniorhnychus Culex nigripalpus Psorophora columbiae June 21 6912 June 22 4608 June 23 2176 752 104 8 June 24 3744 June 28 4672 June 29 4800 June 30 3250 32 0 0 July 1 3984 July 5 1088 July 12 3232 July 13 2600 July 14 1144 1280 392 16 July 15 1392 July 19 S706 July 20 2016 July 21 1728 912 160 *Total mosquitoes. Theobald were collected in late July, this species has a strong host preference for birds rather than large mammals. While Psorophora columbiae do have a strong host preference for large mammals, the small number collected was not considered significant. During the weeks immediately preceding the cattle deaths, unusually large numbers of biting Ae. taeniorhynchus (landing rate counts well over 100) were encountered by the authors while they conducted field work in mangrove forests less than 4 km from the pasture. Discusst1oN—The available information suggests that the deaths of these cattle can be attributed to prolonged exposure to swarms of feeding Ae. taeniorhynchus. We postulate that the cattle were slowly weakened by the cumulative effects of No. 2, 1993] ADDISON AND RITCHIE—CATTILE FATALITIES 69 lengthly exposure to feeding mosquitoes during June and July. Eventually they became submissive and no longer resisted attack. At this point, over several nights, the weakened cattle were exsanguinated as evidenced by the autopsy. This is similar to the chain of events in Abbitt and Abbitt (1981) except that, in this case, they occurred over a longer period of time. The dry conditions conducive to explosive increases in mosquito populations were similar to those reported in Abbitt and Abbitt (1981) and Hoffman and McDuffie (1962). Unfortunately, in the 1962 case the cause of death was not adequately determined. Although the 1932 report did not detail meteorological events, the condition of the dead cattle was similar to that described by Abbitt and Abbitt (1981) and in this report except that mosquitoes were present in the lumen of the nares and trachea. ACKNOWLEDGMENTS—We thank David C. Randall for providing his autopsy report and Bill Langford for information regarding his cattle. Additional assistance was also provided by Craig Bonvechio of the Collier Mosquito Control District and Christine Ramsey of The Conservancy, Inc. LITERATURE CITED ABBITT, B. AND L. G. ABBITT 1981. Fatal exsanguination of cattle attributed to an attack of salt marsh mosquitoes (Aedes sollicitans). J. Am.Vet. Med. Assoc. 179:1397-1400. BISHOPP, F.. C. 1933. Mosquitoes kill livestock. Science 77:115-116. BONVECHIO, C. 1988. Collier Mosquito Control District, P. O. Box 7069, Naples, FL, Pers. Commun. GILLETT, J. D. 1972. The Mosquito: Its Life, Activities, and Impact on Human Affairs. Doubleday and Co., Inc., Garden City N.Y. HEARLE, E. 1938. Insects and Allied Parasites Injurious to Livestock and Poultry in Canada. Canadian Dept. of Agriculture, Publ. 604. Ottawa, Can. HOFFMAN, R. A. AND W.C. MCDUFFIE. 1963. The 1962 gulf coast mosquito problem and the associated losses in livestock. NJ Mosq.Exterm. Assoc. 50: 421-424. LANGFORD, B. 1988. Myrtle Cove, Naples, FL, Pers. Commun. RANDALL, D. C. 1988. Big Cypress Animal Clinic, 11363 Tamiami Trail South, Naples, FL, Pers. Commun. RITCHIE, S. A. 1984. Record winter rains and the minimal opulations of Aedes taeniorhynchus (Wiedemann): cause and effect? J. Fla. Anti-Mosq. Assoc. 55:14-21. RITCHIE, S.A. ANDE.S. JOHNSON. 1991. Distribution and sampling of Aedes taeniorhynchus (Wiedemann) (Diptera:Culicidae) eggs in a Florida mangrove forest. J. Med. Entomol. 28:270-274. SANDERS, D. P., M. E. RIEWE AND J.C. MCNEILL. 1968. Salt marsh mosquito control in relation to beef cattle production: a preliminary report. Mosq. News 28:311-313. STEELMAN, C. D., T. W. WHITE AND P. E. SCHILLING. 1972. Effects of mosquitoes on the average daily gain of feedlot steers in southern Louisiana. J. Econ. Entomol. 65:462-468. Florida Scient. 56(2):65-69.1993. Accepted: September 22nd, 1992. 70 FLORIDA SCIENTIST [VOL 56 Biological Sciences HABITAT STRUCTURE AND THE DISPERSION OF GOPHER TORTOISES ON A NATURE PRESERVE M. C. Stewart"), D. F. Austin ® anp G. R. Bourne “College of Law, Florida State University, Tallahassee, FL 32306; Department of Biological Sciences, Florida Atlantic University, Boca Raton, Fl. 33431 Pp g ty ABSTRACT: Environmental parameters were quantified to develop a physiognomically based system for describing and predicting gopher tortoise (Gopherus polyphemus) habitat on the Florida Atlantic University Nature Preserve. There were significant correlations among gopher tortoise densities and all physiognomic features except saw palmettos. Some correlations were positive, others were negative. However, the only feature with predictive usefulness was bare ground. Burrows were regularly dispersed and the highest densities occurred in historic wet prairie associations. Tortoises located burrows in areas with less canopy and shrub cover but greater herbaceous cover and more extensive bare ground. If we are to maintain current densities of tortoises on the Preserve, management of exotic vegetation will have to be employed Ir Is KNOWN that terrestrial animals do not disperse randomly to occupy landscapes, but actively select their habitats. Animals apparently select on the basis of characteristic recognition features (Wiens, 1969), or “sign stimuli” of the habitat (Hildén, 1965; Lack, 1933, 1937). These cues presumably serve as guides to indicate habitats which can potentially provide the correct ultimate factors, habitats to which a species is adapted (Wiens, 1969). Such cues or proximate factors may include the distribution of vegetation in space, soil type, features of the general landscape, feeding and/or nesting sites, conspecific animals, etc. (Wiens, 1969; Partridge, 1978). Furthermore, dispersion and spatial arrangement of individuals tend to be impor- tant determinants of population density, gene flow within and between groups, and thus, persistence of a species (Partridge, 1978; Brown and Brown, 1984). Dispersal and dispersion are usually the consequences of individual habitat selection (Par- tridge, 1978) and behavioral interactions among individuals (Brown and Brown, 1984; Stamps, 1988). Thus, analyses of the correlates of habitat selection and of animal dispersion patterns are important facets in comprehending a species’ behav- ioral ecology. Throughout the world, competition with humans for space has reduced the availability of suitable environments to free-ranging terrestrial animals. In Florida, the gopher tortoise (Gopherus polyphemus) is considered a species of Special Concern by the Florida Game and Fresh Water Fish Commission [FGFWFC]}), piimarily because of habitat loss (FGFWFC, 1988; Myers, 1990). This status is the direct result of appropriation of tortoise habitat by humans, by the rapid march of introduced plant species across the landscape, and the absence of fire (Diemer, 1987; Abrahamson and Hartnett, 1990; Myers, 1990). Habitat occupancy by the gopher tortoise is an important political and biological issue in southern Florida as No. 2, 1993] STEWART ET AL.—HABITAT STRUCTURE 71 natural areas continue to be developed at increasing rates (Cox et al., 1987). Under some circumstances, the FGFWFC requires developers to relocate tortoises when their habitats are converted to real estate uses. Many ecological and political issues must be addressed regarding the relocation of displaced tortoises (Diemer, 1989). Among these issues is the “proper” selection of relocation sites; this requires an ecological evaluation of habitat suitability, security and carrying capacity (Diemer, 1989). Some ecological parameters that appear to be associated with habitat occupancy by the gopher tortoise are relative amounts of vegetative cover types, soil type, and successional sere (Garner and Landers, 1981; Auffenberg and Franz, 1982; Cox et al., 1987; Breininger and co-workers, 1988). Breininger and co-workers, (1988) demonstrated correlations and linear relationships of tortoise density with some of these parameters on Merritt Island in scrub and slash pine flatwoods plant communities. Gopher tortoises apparently use physical characteristics of the habitat rather than particular plant associations as cues for selecting appropriate habitats (Campbell and Christman, 1982). The general physical and biotic features thought to charac- terize suitable tortoise habitat are: (1) presence of well drained, sandy soils, which facilitate burrowing; (2) an abundance of herbaceous ground cover for nutrition; and (3) an open canopy, which allows sunlight to reach the ground while providing shrub- free areas for tortoise thermoregulation and growth of herbs (Ernst and Barbour, 1972; Auffenberg and Iverson, 1979; Landers, 1980; Landers and Speake, 1980; Garner and Landers, 1981; Auffenberg and Franz, 1982; Cox et al., 1987). The goal of this study was to improve our ecological predictive capability for making biologically sound decisions regarding habitat maintenance for gopher tortoises and habitat suitability for relocating displaced individuals. Therefore, we quantified environmental parameters thought to be important in gopher tortoise habitat selection on the Florida Atlantic University Nature Preserve by evaluating the following questions: (1) What is the physiognomic composition of the plant associations? (2) What are the gopher tortoise densities and burrow dispersion patterns in each plant association? (3) What is the nature of the relationship between soil permeability and tortoise densities? and (4) What are the relationships between physiognomic parameters and gopher tortoise densities and burrow dispersion patterns? This study, while similar to that of Breininger et al. (1988), differs in two major ways: (1) we studied different plant communities in which the gopher tortoises lived; and (2) more importantly, we analyzed habitat occupancy by tortoises using physiognomic features rather than species composition relationships. MATERIALS AND METHODS—Study site—The Florida Atlantic University Nature Preserve (FAUNP; 26°22' N, 80°7' W) in Boca Raton, Palm Beach County, Florida, is a triangular plot encompassing about 37 ha. It is located on the northern edge of the main campus. Detailed descriptions are available for the region’s climate (Chen and Gerber, 1990), geology (Brown et al., 1990; Petuch, 1990), and flora and fauna (Austin, 1990). Austin (1990) described the land-use history of the site. A portion of the preserve was cleared of trees, stumps and saw palmettos (Serenoa repens) between 1936 and 1937 when the city airport was built. A larger area was bulldozed in 1942 to accommodate an Air Force Base. Later the base was turned over to the state, which established FAU at that location in 1961. About 1964, the FAU Grounds Department began systematically mowing the parcel at least once yearly. Large trees of several species surviving these 79 FLORIDA SCIENTIST [VOL 56 mowings included entire oak hammocks (Quercus spp.) and saw palmetto clumps, a few pine trees (Pinus spp.), and isolated cabbage palms (Sabal palmetto). By 1973, the Biological Sciences Department convinced FAU to stop mowing the triangle. The site has been undergoing unmanaged secondary succession since 1973 (Austin, 1990), and there are four distinct plant associations on the Preserve. Historically, the habitats were wet prairie, dry prairie, low hammock, and sand pine scrub. The historical pattern has changed in recent years (Fig. 1). Austin (1990) described these communities and showed how they have changed in response to a number of factors. Vegetation—We gathered data from May through August 1989 and April through July 1990 by subdividing the Preserve into 46 unequally sized plots (x=8700 + 9700 s.d. m7); this facilitated measuring and mapping the physiognomic features (Fig. 2). The unequal plot sizes resulted from delineations that mirrored the distributions of plant associations. We delineated plot boundaries by installing PVC pipe “corner-posts.” To assure randomness of location of sample sites within plots, we established single transects through the longest axis of each plot. A sample unit was located along the transect by using three random numbers—the first determined the distance measured, in meters, along the transect; the second determined on which side of the transect the sample unit was to be located (even number on the right; odd number on the left); and the third number indicated the distance to be measured perpendicular to the transect (Wiens, 1969). We sampled the physiognomic features of the habitat using a 1 m? quadrat. The six parameters included bare ground (i.e., areas devoid of living plant cover), graminoids (i.e., narrow- leaf herbaceous plants), forbs (i.e., broad-leaf herbaceous plants), woody plants (i.e., perennial plants having a secondarily thickened lignified stem), saw palmettos (Serenoa repens), and cacti (Opuntia spp.). We developed species area curves (Brower and Zar, 1977) to determine when adequate sample sizes were attained for each plot. The number of m?* quadrats/plot ranged from three to 16 because higher plant species diversity in a plot resulted in greater sampling effort to produce the leveling off of the species area curve. Percentage cover of each vegetative parameter (percentage of ground canopied by each vegetative parameter) in each quadrat was recorded, and mean percentage cover from quadrat data for each plot and each plant association was determined. twee WET PRAIRIE DRY PRAIRIE LOW HAMMOCK SAND PINE SCRUB e4, © 4 > 22. me were . ., Ors <> Si 0; &2 re o> &2 x2 1) ¢ $5555 x es ey S55 7 ye. > > AN Scale se SLA Icm=134m A 3 SORA ON ° oO CRA oA CRA (ARP oe a ts CON AS VEGETATION MAP EF VP OCCO SAS 7. ELL> et) NS. \ e EEL MY? 4 SAQA EARS ERK <% ERAN RR CESS L AS SOARS ee GL ec OG 4 eee © AR LONG, oe Nee ey, oe. ALL LAE ry (W229 .2.4 eis Or GN ODS CALNE L) OTN FOROOCKA RERARARE RS? PAAR SY. AAA OS®: ‘AAAS ee: PR AAA SSSR: VO NL SID OVV OOS: LSA, PIP as —— ri LEBEL Vian SaR CC OOOCOCOOOOOOO aor Fic. 1. FAU Nature Preserve indicating present distributions of plant associations. Dark lines are trails that traverse the Preserve and white areas are dry swales cut as ditches by the military during World War II. No. 2, 1993] STEWART ET AL—HABITAT STRUCTURE We Scale lcm=134m SAMPLE PLOTS Fic. 2. FAU Nature Preserve showing the distribution of plot locations used for random sampling of the physiognomic features of the habitat and plot identification numbers. Gopher tortoise densities —Gopher tortoise densities in each of the plots and plant associations were determined by counting the number of active and inactive burrows. First, we decided whether burrows were active or inactive by applying Auffenberg and Franz’s (1982) criteria. Second, we applied a modification of the stick method (Coxet al., 1987; Breininger et al., 1988) to obtain a site-specific estimate of tortoise population size. See also a recent paper by Witz et al. (1992) for additional comments on correction factors and estimates of population sizes in tortoises. Entrances to 25 active and 25 inactive burrows were blocked by Y-shaped sticks and monitored for seven days (4-10 July 1990). These 50 burrows were located in plots 1,3 and6 not less than 15 meters apart. An assumption was made that no tortoises used more than one of these burrows. For the first two days, we considered burrows showing evidence of entrance or egress tracks to be occupied. However, only burrows showing exit tracks were considered occupied over the next five days. We then applied the two correction factors, one for active and the other for inactive burrows, to the number of active and inactive burrows in each plot. These values were then summed to estimate the numbers of tortoises in each plot. Then, we divided the number of tortoises in a plot by the number of active and inactive burrows in that plot to derive an overall correction factor for each plot. The overall correction factors for each plot were then averaged to obtain the overall correction factor for the entire Preserve (active plus inactive burrows). Gopher tortoise population sizes by plant association and by mean density/ha were estimated. Burrow dispersion—We measured nearest neighbor interburrow distances in ten randomly chosen plots (1, 2, 3,6, 11, 12, 13, 17, 18, 36 [Fig. 2]). Indices of aggregation were calculated to determine burrow 74 FLORIDA SCIENTIST [VOL 56 dispersion using the Clark and Evens (1954) method to categorize dispersion patterns (Krebs, 1989). Soil permeability—We developed an index of soil permeability for each plant association by taking a 30 cm x 3.5 cm soil core in a PVC pipe when the soil appeared to be completely dry to sight and touch. Twenty random samples were taken from each of the four plant associations, and 20 more from about 60 cm from active burrows. The index was the time required for the first drop of 300 ml of water to run through this soil packed pipe which had removable nylon window screening covering one end. Statistical analyses—Data were analyzed using the computer software SYSTAT (Wilkinson, 1989). We transformed all variables by square root to correct for positive correlations between means and variances (Sokal and Rohlf, 1981). The Kruskal-Wallis (KW) 1-way ANOVA model was used to determine whether soil permeability differed in the four plant associations and near burrows (Sokal and Rohlf, 1981). Moreover, we compared distributions of soil permeability indices for the four plant associations by employing the nonparametric multiple comparison by simultaneous test procedure (STP) (Sokal and Rohlf, 1981). The relationship between tortoise density and soil permeability using the stepwise linear regression model for each of the plant associations was also examined. Soil permeabilities were omitted from the multiple regression analysis because of the small sample size (n=4 vegetation types). We executed a 7 x 7 Pearson correlation matrix to determine the relationship between plant associatio parameters and gopher tortoise densities. A Bartlett X? test that indicates whether correlations are robust was also employed (Wilkinson, 1989). The variables in the matrix were bare ground, graminoids, forbs, woody vegetation, saw palmettos, cacti, and tortoise density. However, because so Scale 134m GOPHERS DENSITY May-July 1989 Fic. 3. FAU Nature Preserve showing the gopher density in the sample plots. The numbers represent the estimated number of tortoises per m*x10* in each plot. Plots with no numbers contained no tortoises. No. 2, 1993] STEWART ET AL.—HABITAT STRUCTURE 75 many variables correlated with each other, we used a multiple regression to elucidate habitat variables with the strongest linear relationships to tortoise density (Draper and Smith, 1981). To determine whether there was a linear relationship between dispersion patterns of burrows and vegetation parameters we ran a multiple regression. The index of burrow aggregation was designated the dependent variable, and bare ground, graminoids, forbs, woody vegetation, saw palmetto, and cacti were the independent variables. RESULTS—Vegetation—The mean percentages of vegetative parameters in each plant association are presented (Table 1). All bare ground was canopied by woody vegetation in the sand pine scrub and hammock; graminoids were absent from dry prairies, sand pine scrub and hammock; while prickly pear cactus did not grow in sand pine scrub or hammock plant communities. Former wet prairies were highest in bare ground, and dry prairies, sand pine scrub and hammocks were dominated by woody vegetation. Gopher tortoise densities—The correction factor was 0.84 for active burrows and 0.24 for inactive burrows, while the overall correction factor for the entire Preserve was x = 0.67+0.10s.d. There were 681 active and 280 inactive burrows for a total of 961. Table 2 indicates tortoise population sizes, and mean density by plant association after applying the correction factor to active and inactive burrows. We estimated 639 tortoises in the Preserve. The drained wet prairies supported almost five times the numbers and eight times the density of tortoises in the dry prairies (Table 2). Eighty-one percent of the tortoises were found on the historic wet prairies TABLE 1. Relative cover (x + s.d.%) of vegetation types by plant associations on the FAU Nature Preserve. GROUNDCOVER ASSOC BARE GRAM FORB WOOD SAWP CACT WP* A oa) eee) Iie taaled 14.4+7.4 19.0£20.4 0.140.2 1.9+3.6 Pr 1.344.0 0 1.9+4.8 96.4+6.2 0.4+1.0 1.9+3.6 SPS* 0 0 0.641.2 83.2£17.7 16.3418.2 0 HAM! 0 0 0.540.7 98.341.4 1.3+1.5 0 Groundcover: BARE, Bare ground; GRAM, Graminoids; FORB, Forbs; WOOD, Woody vegetation; SAWP, Saw palmettos; CACT, Cacti. lant association: ‘Wet prairie; ‘Dry Prairie; ‘Sand pine scrub; ‘Hammock. TABLE 2. Gopher tortoise population estimates and mean densities/ha (x + s.d.) by plant association on the FAU Nature Preserve. Plant Association No. Tortoises Tortoise Density Wet Prairie 516 45+18 s.d. Dry Prairie 111 647 s.d. Sand Pine Scrub 12 4+2 s.d. Hammock 0 0+0 s.d. 76 FLORIDA SCIENTIST [VOL 56 even though that association comprised only 25% of the area of the Preserve. The mean number of tortoises was 6 + 7 s.d. in the dry prairies, indicating high variability among plots in that plant association. Burrow dispersion—tThe indices of aggregation are listed (Table 3). Since all z values were greater than 1.96, the null hypothesis of randomness was rejected. Moreover, because all indices of aggregation were > 1 (Table 3), gopher tortoise burrows on the Preserve were regularly dispersed. Soil permeability—The time for water to percolate through soil core samples varied considerably, ranging from 73 to 298 seconds. Soil permeability indices from the four plant associations and active burrows differed significantly (KW-ANOVA = 18.6, n = 100, P < 0.001). The means are as follows: burrows, 149 + 50 s.d.; historic wet prairie, 124+ 26s.d_.; dry prairie, 176 + 56 s.d.; hammock, 140 + 22s.d.; and sand pine scrub, 127 + 17 s.d. There were, however, differences in the distributions of permeability indices only for dry and historic wet prairie (STP U = 325, n = 100, P < 0.05), and for dry prairie and sand pine scrub (STP U = 341, n = 100, P < 0.05). Furthermore, there was no linear relationship between permeability indices and tortoise densities (r?= 0.12; P > 0.05). Correlates of tortoise habitat selection—The significant Bartlett X? = 288.5 (d.f. = 21, P < 0.0001) indicates that the correlations between tortoise densities and vegetation parameters were robust (Table 4). Although all the variables except saw palmetto were significantly correlated with tortoise densities (Table 4), the multiple regression indicated that the only significant linear relationship with tortoise density was bare ground (t = 4.58, n = 46, P < 0.0001). Thus, plant associations with a TABLE 3. Nearest neighbor burrow distance indices of dispersion indicating regular dispersion (R > 1) of active gopher tortoise burrows. Plot PA? n Index of Dispersion (R) Ta ik WP 50 1.89 6.63 2 WP 13 1.73 5.05 3 DP 11 3.06 13.06 6 WP 50 145 2.09 1 WP 13 1.61 4.20 12 DP 9 3.76 15.86 13 DP 8 1.63 3.42 17 DP 11 2.25 7.96 18 DP 3 4.81 12.64 36 SPS 12 2.64 10.85 *Plant association, WP, Wet prairie, DR, Dry prairie, SPS, Sand pine scrub. n = number of tortoise burrows. * Significant at P < 0.05. No. 2, 1993] STEWART ET AL.—HABITAT STRUCTURE TG relatively high percentage of unvegetated areas had the highest concentrations of gopher tortoises. Additionally, there was a significant negative correlation between tortoise density and woody vegetation (Table 4), indicating an absence of burrows in plots dominated by woody vegetation. Bare ground correlated with all other physiognomic parameters except saw palmettos and was negatively correlated with woody vegetation (Table 4). Graminoids showed the same correlations as bare ground. Forbs correlated with all vegetation parameters except saw palmettos and cacti. Woody vegetation correlated negatively with all vegetation parameters except that it did not correlate at all with saw palmettos. Saw palmettos did not correlate with any of the vegetation parameters. Cacti correlated with all vegetation parameters except forbs and saw palmettos. There was, however, no linear relationship between dispersion patterns of burrows and the cover composition (r?= 0.39; P > 0.05). DiIscussION—Our results clearly indicate that the relative percent cover of the various physiognomic types varied by plant associations (Table 1), and that historic wet prairies had the highest density of active tortoise burrows (Table 2). Physiognomic composition and gopher tortoise densities — Our mean correc- tion factor of 0.67 was higher than those of Auffenberg and Franz (0.61; 1982) and Breininger and co-workers (0.17 to 0.28 for fall and spring; 1988). This may be because the tract has reached its carrying capacity, and the lack of other suitable habitat is forcing the tortoises to remain on the site. Likewise, our estimates of tortoise density (Table 2) are higher than those of Auffenberg and Franz (0.9 tortoises/ha in sandhill habitats; 1982) and Breininger and co-workers (1.1 and 2.4 tortoises/ha in scrub, flatwoods and disturbed habitats on the Kennedy Space Center; 1988). No other study has reported gopher tortoises in such high concentra- tions as those on the FAUNP. McCoy and Mushinsky (1992), in comparing mainland and island gopher tortoise habitats, found that tortoises on islands cannot disperse and consequently may be forced to occupy marginal habitats. FAUNP can be considered a “habitat TABLE 4. Pearson correlation matrix with 21 degrees of freedom for gopher tortoises and vegetation physiognomy on the FAU Nature Preserve. Variable BARE GRAM FORB WOOD SAWP CACT TORT BARE 1.00 GRAM 0.88° 1.00 FORB OMiSe 0.74° 1.00 WOOD -0.92° -0.89° -0.81° 1.00 SAND -0.22 -0.19 -0.27 -0.07 1.00 CACT 0.45°° 0.49° 0.27 -0.45°° —_ -0.08 TORT 0.88° OMe 0.75° -0.79° -0.20 0.44°° 1.00 Abbreviations: Bare ground; Graminoids; Forbs; Woody vegetation; Saw palmetto; Cacti; Tortoise density. *Significant at P<0.0001; °°Significant at P<0.05 78 FLORIDA SCIENTIST [VOL 56 island” to which the tortoises are confined. There is no adjoining suitable habitat to which additional individuals may retreat, and the nearest suitable habitat is separated from the Preserve by highways, buildings and other human-populated areas. Thus, the tortoises are effectively confined to the preserve and exist at high density. Additionally, exotic plants such as Brazilian pepper (Schinus terebinthifolius), have been encroaching on suitable tortoise habitat areas within the Preserve at a rapid rate (Austin, unpubl. data), further crowding the population. Tortoises are seasonally found by students or faculty wandering across the road headed away from the Preserve. Such departures may indicate a search for more suitable habitat. These animals are usually returned to the Preserve, but their fates have not been studied. Additionally, it is known that tortoises have migrated under the fence between the Preserve and the airport and have built burrows in the grassy areas there. These grasslands were only temporary refugia, as airport officials recently have had the tortoises relocated “for safety reasons” (Austin, unpubl. data). Occasionally, tortoises have been found burrowing in the grassy areas on the FAU campus surrounding the Preserve. Proximity to student parking lots prohibits long- term tortoise occupancy in this planted grass. Burrow dispersion patterns—Tortoises are dispersed regularly at FAUNP (Table 3). Since gopher tortoises are not highly social except during courtship, and since they do not provide parental care of young, one would not expect to find clumping patterns in the species. An anonymous reviewer informs us that clumping of young of some species, apparently not G. polyphemus, has been reported near the maternal burrow, but we did not measure this parameter. Co-occupation of a single burrow by one or more males and a female, by two males, and by two immature tortoises has been observed during the months of May through November (Diemer, 1991). Recently, burrows have been found to be occupied simultaneously by males and females on the Preserve (Hicklin, 1991). The male-female co-occupancy appears to be related to courtship behavior, which occurs during those months. Soil permeability—Though there were differences in the distributions of permeability indices for dry and historic wet prairies and for dry prairies and sand pine scrub, the differences may be misleading. There may be differences within the FAUNP, but they are not large enough to affect discrimination by the gopher tortoises. There was no linear relationship between the soil permeability indices and tortoise densities, indicating that no soil type was preferred for burrowing over any other. Since the water table has been dropped in southern Florida, the soils on the FAUNP have become dry enough to support a larger gopher tortoise population. On the FAUNP, it is not the differences in soil permeability but the lack of water due to the drop in the water table that has affected the habitat selection of gopher tortoises. Therefore, soil permeability within the FAUNP has a minor effect on tortoise habitat selection in comparison to vegetation parameters. Correlates of tortoise habitat selection—This study of the FAUNP is consistent with previous findings that gopher tortoise density is related to herbaceous biomass and is greatest in grassy, open-canopy associations. Historic wet prairies on the FAUNP had only 19% woody vegetation, only a small percentage of which were canopy species, and supported high tortoise densities (x = 45+ 18 s.d.). Hammocks No. 2, 1993] STEWART ET AL.—HABITAT STRUCTURE 79 had practically 100% canopy cover and supported no gopher tortoises. These results are consistent with Cox and co-workers (1987) and Breininger co-workers (1988), who reported a canopy cover of less than 60% or 80% is necessary for tortoise habitation. Correlations existed between tortoise densities and percentages of all vegeta- tion parameters studied except saw palmettos. There was a negative correlation between woody vegetation and tortoise density. Additionally, there was a significant linear relationship between tortoise density and bare ground. Diemer (1986) and Breininger and co-workers (1988) likewise showed that tortoises selected burrow locations with less canopy and shrub cover but greater herb cover and bare ground. CONCLUSIONS—Auffenberg and Franz (1982) found that the successional stage of a habitat is one of the most important factors regulating population densities of gopher tortoises. As succession proceeds, lower light levels inhibit growth of forbs and grasses, and the site becomes less suitable for tortoise habitation. However, disturbance may open up the canopy and allow ground cover vegetation to grow and thereby support tortoise reoccupation (Auffenberg and Franz, 1982). To promote the growth of grasses and legumes favored by the tortoises, fire has been recom- mended as a management tool (Landers and Speake, 1980). Carrying capacity of tortoises in appropriate habitat can be increased by periodic burning or by removal of shrubs, particularly in thickly overgrown areas. Stout and co-workers (1989) reported a marked increase in ground-level vegetation following a burn and specu- lated that the increase in plant growth would increase the growth, survival, and reproductive output of resident gopher tortoises. The high density tortoise population on the FAUNP is probably due to past disturbance and clearing of the area. Three factors have probably led to the current tortoise densities. Past disturbances have produced an abundance of the proper ecological requirements for gopher tortoises such as open-canopy areas with high percentages of herbaceous growth. The drop in the water table has resulted in drying of the sandy soils on the site and made them suitable for gopher tortoises. Addition- ally, the grassy areas surrounding the Preserve provide a predictable source of food. Gopher tortoise densities at FAUNP may begin to decline as native woody plants and exotic plants invade otherwise suitable tortoise habitat, thus increasing the percentages of shrub and canopy cover; exotic plants such as Brazilian pepper may be especially detrimental in this regard. Diemer (1987) reports that benefits accrued by creating openings in tortoise habitats may be short-lived due to subsequent invasion by exotic species such as Brazilian pepper. Further research is now being conducted to determine the rate of exotic plant invasion and canopy closure (Hicklin and Ferriter, in prep.). A comparison of tortoise densities as the percentage of exotic cover increases or decreases is expected to reveal the extent to which exotic species may be affecting tortoise densities, and this may be critical to formulation of a management plan for the Preserve. ACKNOWLEDGMENTS—This paper is based on a thesis by the first author. We thank R. B. Grimm and M. L. Boss for their assistance and suggestions. Special appreciation is extended to Eileen L. Garcia 80 FLORIDA SCIENTIST [VOL 56 for her advice and help with the statistical program, SYSTAT, and to A. Chris Collins who assisted in the use of the Macintosh graphics program Pixel Paint. Financial support was provided by the Department of Biological Sciences of Florida Atlantic University and by the Boca Raton Garden Club. We gratefully acknowledge those who assisted in collecting field data, especially James T. Stewart, and Samuel A. Spiegel, who both volunteered many hours of labor. LITERATURE CITED ABRAHAMSON, W. G. AND D. C. HartneETT. 1990. Pine flatwoods and dry prairies. Pp. 103-149. In: R. L. Myers AND J. J. Ewet (eds.), Ecosystems of Florida. Univ. of Central Florida Press, Orlando. AUFFENBERG, W. AND J. B. Iverson. 1979. Demography of terrestrial turtles. Pp. 541-569. In: M. HarLess AND H. Mor.ockx (eds.), Turtles: Research and Perspectives. Wiley-Interscience, New York. AND R. Franz. 1982. The status and distribution of the gopher tortoise (Gopherus polyphemus). Pp. 95-126 In: Bury, R.B. (ed.), North American Tortoises: Conservation and Ecology. Wildl. Res. Rep. 12, U.S. Dept. Interior, Washington, D.C. AustTIN, D. F. 1990. Vegetation on the Florida Atlantic University ecological site. Florida Scient. 53(1):11- 27. BREININGER, D. R., P. A. SCHMALZER, D. A. RYDENE, AND C. R. HINKLE. 1988. Burrow and habitat relationships of the gopher tortoise in coastal scrub and slash pine flatwoods on Merritt Island, Florida. Florida Game and Fresh Water Fish Comm. Nongame Wildl. Prog. Final Rep. BroweR, J. E. AND J. H. Zar. 1977. Field and Laboratory Methods for General Ecology. 2nd ed. Wm. C. Brown, Dubuque. Brown, L. AND J. Brown. 1984. Dispersal and dispersion in a population of soldier beetles Chauliognathus pennsylvanicus. Environ. Entomol. 13:175-178. Brown, R. B., E. L. SToNnE, AND V. W. Car isLe. 1990. Soils. Pp. 35-69. In: Ecosystems of Florida. Univ. of Central Florida Press, Orlando. CamPBELL, H. W., NDS. P. CuRISTMAN. 1982. The herpetological components of Florida sandhill and sand pine scrub associations. Pp. 163-171. In: Scott, N. J., JR. (ed.), Herpetological Communities. U.S. Fish and Wildl. Serv., Wildl. Res. Rep. 13. CHEN, E. AND J. F. GerBer. 1990. Climate. Pp. 11-34 In: R. L. Myers and J. J. Ewel (eds.), Ecosystems of Florida. Univ. of Central Florida Press, Orlando. Car, P. J. AND F. C. Evens. 1954. Distance to nearest neighbor as a measure of spatial relationships in populations. Ecology 35: 445-453. Cox, J., D. INKLEY, AND R. Kautz. 1987. Ecology and habitat protection needs of gopher tortoise (Gopherus polyphemus) populations found on lands slated for large-scale development in Florida. Florida Game and Fresh Water Fish Comm. Nongame Wildl. Prog. Tech. Rep. No. 4. Tallahassee. Diemer, J. E. 1986. The ecology and management of the gopher tortoise in southeastern United States. Herpetologica 42:125-133. . 1987. Tortoise relocation in Florida—solution or problem? Proc. Desert Tortoise Counsel Symposium. 1987:131-135. . 1989. An overview of tortoise relocation. Pp. 1-6. In: Diemer, J. E., D. R. Jackson, J. L. Lanbers, J. N. LayNE AND D. A. Woop (eds.), Gopher Tortoise Relocation Symposium Proceed- ings. Florida Game and Fresh Water Fish Comm. Nongame Wildl. Prog. Tech. Rep. No. 5. Tallahassee. . 1991. Game & Freshwater Fish Commission, Gainesville, Fl. Pers. comm. DraPeER, N. AND H. Smitu. 1981. Applied Regression Analysis, 2nd ed. Wiley and Sons, New York. Ernst, C. H. anp R. W. Barsour. 1972. Turtles of the United States. The Univ. Press of Kentucky, Lexington. FLORIDA GAME AND FRESH WATER FIsH Commission [FGFWFC]. 1988. Official List of Endangered and Potentially Endangered Fauna and Flora of Florida. 1 July. Tallahassee. GARNER, J. A. AND J. L. Lanpers. 1981. Foods and habitat of the gopher tortoise in southwestern Georgia. Proc. Ann. Conf. S. E. Assoc. Fish and Wildl. Agen. 35:120-134. HIckun, J. M. 1991. Department of Biological Sciences, Florida Atlantic University, Boca Raton, FL 33431. Pers. comm. HILDEN, O. 1965. Habitat selection in birds. Ann. Zool. Fenn. 2:53-75. Kress, C. J. 1989. Ecological Methodology. Harper and Row, New York. Lack, D. 1933. Habitat selection in birds, with special reference to the effects of afforestation on the Breckland avifauna. J. Anim. Ecol. 2:239-262. No. 2, 1993] STEWART ET AL.—HABITAT STRUCTURE 8] . 1937. The psychological factor in bird distribution. British Birds 31:130-136. LANDERS, J. L. 1980. Recent research on the gopher tortoise and its implications. Pp. 8-14. In: Franz, R. AND R. J. Bryant (eds.), The Dilemma of the Gopher Tortoise—Is There a Solution? Proc. Ist Ann. Mtg., Gopher Tortoise Counc. AND D. W. SpEAKE. 1980. Management needs of sandhill reptiles in southern Georgia. Proc. Ann. Conf. S. E. Assoc. Fish and Wildl. Agen. 34:515-529. McCoy, E, AND Musuinsky, H. 1992. Studying a species in decline: changes in populations of the Gopher Tortoise on federal land in Florida Florida Scientist 57(2):116-125. Myers, R. L. 1990. Scrub and high pine. Pp. 150-193. In: Myers, R. L. AND J. J. Ewe (eds.), Ecosystems of Florida. Univ. of Central Florida Press, Orlando. ARTRIDGE, L. 1978. Habitat selection. Pp. 351-376. In: Kress, J. R. AND N. B. Davies (eds.), Behavioral Ecology: An Evolutionary Approach. Blackwell Scientific, Oxford. Etucu, E. J. 1990. The Pliocene pseudoatoll of southern Florida: a template for the evolution of the Everglades. Pp. 183-194. In: ALLMAN, W. D. anv T. M. Scorr (eds.), Plio-pleistocene Stratigraphy and Paleontology of South Florida. Southeastern Geological Society Annual Field Trip Guide- book. Guidebook No. 31. SoxaL, R. R. AND F. J. Rouir. 1981. Biometry. 2nd ed. Freeman. New York. Stamps, J. A. 1988. The effect of body size on habitat and territory choice in juvenile lizards. Herpetologica 44:369-376. Stout, J. I., D. R. RICHARDSON, AND R. E. Roserts. 1989. Response of resident and relocated gopher tortoises to a prescribed burn in a sand pine scrub community. Pp. 84-85. In: DieM_r, J. E. , D. R. Jackson, J. L. Lanpers, J. N. LayNe anp D. A. Woop (eds.), Gopher Tortoise Relocation Symposium Proceedings. Game and Fresh Water Fish Comm. Nongame Wildl. Prog. Tech. Rep. No. 5. Tallahassee. Wiens, J. A. 1969. An approach to the study of ecological relationships among grassland birds. Omith. Mono. No. 8. Wixkinson, L. 1989. SYSTAT: The System for Statistics. Evanston, IL: SYSTAT, Inc. Wirz, B., D. S. WiLson AND M. D. Pater. 1992. Estimating population size and hatchling mortality of Gopherus polyphemus. Florida Scientist 57:14-19. Florida Scient. 56(2):70-81.1993. Accepted: November 6, 1992. 89 FLORIDA SCIENTIST [VOL 56 Biological Sciences NATURAL HISTORY OF A SMALL POPULATION OF LEIOCEPHALUS SCHREIBERSII (SAURIA: TROPIDURIDAE) FROM ALTERED HABITAT IN THE DOMINICAN REPUBLIC 1.5) MICHAEL C. SCHREIBER, ® ® ROBERT POWELL, ® JOHN S. PARMERLEE, JR., (3) Amy LATHROP, AND 4) DONALD D. SMITH Department of Biology, Rockhurst College, Kansas City, Missouri 64110; ®) Department of Natural Sciences, Avila College, Kansas City, Missouri 64145; ‘) Museum of Natural History, University of Kansas, Lawrence, Kansas 66045; “ Division of Allergy & Rheumatology, University of Kansas Medical Center, Kansas City, Kansas 66103; ©) Current address: Department of Zoology, Miami University, Oxford, Ohio 45056; ) Corresponding author Apstract: We describe aspects of the natural history of Leiocephalus schreibersii from altered habitat in Barahona, Dominican Republic. We marked and observed 18 lizards in June 1991. Estimates of population size and density were 20 + 4 and 1/70 m2, respectively. Examination of ingesta from stomachs of 17 additional animals suggested that this species is a generalist-opportunist. The sex ratio (M:F, 20:15) did not differ significantly from 1:1. Snout-vent lengths (SVL) of adult (> 50 mm) males (53-86 mm) were larger than those of adult females (55-68 mm). Adult home range (HR) areas did not differ significantly by sex (M, 22-130 m?; F, 16-53 m?), nor were HR areas correlated with SVL. Activity periods extended from 0900-1600 h. Males were aggressive and demonstrated territorial defense. We did not observe female aggression. Lizards most commonly used a sit-and-wait feeding strategy, but some individuals foraged actively for prey. LEIOCEPHALUS schreibersii is a ground-dwelling lizard inhabiting semi-open, xeric areas with moderately sized rocks for basking (Jenssen et al., 1989). A Hispaniolan endemic, the species ranges widely throughout dry regions of both Haiti and the western Dominican Republic (Schwartz and Henderson, 1991), and has been introduced into Dade and Broward counties, Florida (Conant and Collins, 1991). Only a few studies have addressed the natural history of these lizards. Regal (1978, 1983) examined aspects of thermal relations and foraging behavior. Marcellini and Jenssen (1989, 1991) discussed thermal ecology and learning behavior, respec- tively. Jenssen et al. (1989) described differential infanticide and territoriality. Here, we examine aspects of the natural history of a small population of L. schreibersii occupying altered habitat in the courtyard of the Hotel Guarocuya in Barahona, Provincia de Barahona, Reptiblica Dominicana. Barahona is located on the southwestern shore of the Bahia de Neiba. The mean annual temperature is 26.1°C and the mean annual rainfall is 1047.1 mm. In June, the month during which this study took place, the mean temperature is 28.0°C and the mean rainfall is 200.0 mm (Salcedo et al., 1983). The Hotel Guarocuya is on the southern edge of the city at sea level. The study area was a walled courtyard of ca. No. 2, 1993] SCHREIBER ET AL.-LEIOCEPHALUS SCHREIBERSII 83 1400 m?immediately west of the hotel. The area was enclosed by a 1 m high concrete wall andthe hotel itself. The wall was broken by two iron gates through which lizards could enter or leave. Large sections of the courtyard were being planted with flower beds of spider lilies. Other vegetation included oyster plants, large crotons, hibiscus, grasses, and both palm and broad-leaf trees. Between patches of vegetation were large expanses of packed sand with piles of debris, which lizards used as basking sites. Other lizards sharing the courtyard included the locally rare and apparently syntopic Leiocephalus barahonensis, terrestrial Ameiva chrysolaema and Celestus costatus, largely arboreal Anolis brevirostris, A. chlorocyanus, A. coelestinus, A. cybotes, and the uncommon grass anole, A. olssoni. We observed no interspecific interactions involving L. schreibersii. Lizards had become accustomed to human activity, allowing observations withovt substantially altering their behavior (diminished observer’s effect of Jenssen et al., 1989). Lewis and Saliva (1987) discussed advantages of conducting studies in such areas. Still, as noted by Marcellini and Jenssen (1991), L. schreibersii are proficient at avoiding capture. During initial attempts, nooses could be brought to within several centimeters of individuals before they would flee. As early as the second day of the study, an approach by acollector to within 2 m caused many animals to retreat into refuges. Once animals were marked, possible modifications of lizard behavior were minimized by making observations from hotel balconies or distances 25m. MetHops—We captured lizards by noosing and immediately measured cloacal temperatures with a quick-reading thermometer (Miller and Weber, Inc., Queens, New York). We measured temperatures of the substrate and air at 2 cm and 1 m. Environmental temperatures were taken in the shade and sheltered from any wind. Sexes of individual lizards were determined, then each animal was measured, weighed, uniquely marked with acrylic paint to allow individual recognition without recapture, toe- clipped for permanent identification, and released at the exact site of capture. We assumed adults to have snout-vent lengths (SVL) > 50 mm, at which size males began to acquire reddish coloration and females were determined by palpation to have either large follicles or ova. We estimated population size using both the Lincoln-Peterson index and Schnabel method. Home range (HR) area estimates were based on a minimum of three non-sequential observations on three different days (prior to any introductions) and were calculated using the convex polygon method of Rose (1982). Behavior was recorded with a Chinon CV-T65 video camera. We introduced a large adult male and two juveniles, collected outside the study area, into home ranges of resident males to evaluate territorial and/or aggressive behavior. Stomach contents of L. schreibersii collected from or near the same site in March 1986 (n=1), August 1987 (n=6), March 1988 (n=8), and May 1989 (n=2) were examined and arthropodan food items were counted and identified to order. We determined volumes of food items using methods of Milstead (1957) and calculated importance values as in Powell and co-workers. (1990). These specimens are in the Bobby Witcher Memorial Collection, Avila College, Kansas City, Missouri (BWMC 03017, 03190-03195, 03295- 03302, 04035-04036). We conducted statistical analyses using StatView II (Abacus Concepts, Inc. Berkeley, California). All means are presented plus or minus one standard deviation, and for all tests, a < 0.05. RESULTS AND DISCUSSION—Size and Sex Ratio—Leiocephalus schreibersii is sexually dimorphic in size. Contrary to data presented by Schwartz (1968), lizards of both sexes in the southern areas of the Dominican Republic are smaller than individuals from the northwest (Burns et al., 1992; Jenssen, 1992). Adult male SVLs (n=17, 53-84 mm, X =74.4+10.0 mm) were significantly larger than those of adult females (n=11, 55-68 mm, X =60.7+4.7 mm) (DF=26, t=4.2, P=0.0003). Juvenile 84 FLORIDA SCIENTIST [VOL 56 SVLs ranged from 29-45 mm ( X =35.6£6.9 mm). The sex ratio (M:F) was 20:15, and did not differ significantly from a 1:1 ratio (DF=3, x°=0.95). Population Size and Density—We estimated population size using both the Peterson-Lincoln index (20+4) and Schnabel method (20). The latter indicated that we approached actual population size. These figures produce a density estimate of 1/70 m?. None of these estimates address the possibility of migration into or out of the study area. We observed such movements on three occasions (one lizard avoided capture by running through a gate and two lizards entered the courtyard). Home Ranges—Home ranges of adult males (n=4, 22-130 m?, X =74.8+10.9 m?) were not significantly larger than those of adult females (n=3, 16-53 m?, X =30.0+20.1 m’) (DF=5, t=-0.93, P=0.40). Two juvenile male HRs were 3 and 7 m’, and both overlapped with adult male HRs, one with that of an adult female, but neither was completely within an adult HR. None of the observed consexual adult HRs over- lapped, although several were in peripheral contact with those of other individuals. However, we observed neither male nor female lizards in adjacent regions of their HRs at the same time. Individuals of both sexes maintained a minimum distance of > 2m. Of the three female HRs, one overlapped with that of one male, one with two males, and one was entirely within the HR of a single male. Jenssen and co-workers (1989) reported larger HRs (except for juveniles) in northern Hispaniola with overlap among males and complete inclusion of juvenile ranges within those of adults. These differences may be due to the limited sample size of our study and the confined nature of this site, thus limiting the areas of home ranges and encouraging active defense thereof; or alternately, the larger ranges reported by Jenssen and co- workers (1989) were a reflection of larger SVLs of northern L. schreibersii or of cannibalistic tendencies among some of the adult lizards in that area. As we never observed saurophagy (nor did we find any evidence of such in stomach samples), smaller ranges may be characteristic of populations in which cannibalism does not occur. We found no correlation between SVL and HR areas among all males (n=6, r=0.34, P=0.52), adult males (n=4, r=0.57, P=0.43), or females (n=3, r=0.13, P=0.92), indicating that HR quality (possibly based on availability of refuges and shaded versus exposed areas) was more important than size. Each HR contained at least one favored refuge and one or more secondary refuges. These retreats consisted of burrows under logs or concrete (apparently dug by and occasionally shared with crabs) or drain pipes, concrete blocks, or other human debris. Lizards on nearby beaches used similar crab-constructed refuges under vegetation, although many also had burrows in the open sandy regions, apparently of their own construction. Each HR also included some form of cover capable of providing shade without necessitating retreat into refuges. Activity Period and Thermal Regime—Lizards were active from 0900-1600, with occasional observations as early as 0826 and as late as 1630. Animals in a nearby population, on the beach just north of the hotel, emerged from and retreated into refuges approximately one hour earlier. The shadow cast by the hotel over the courtyard before 0900 was probably responsible for the later activity period of the courtyard population. This relatively short activity period of ~7 h was longer than that No. 2, 1993] SCHREIBER ET ALLEIOCEPHALUS SCHREIBERSII 85 of sympatric Ameiva chrysolaema (~5 h) (Schell et al., 1993), but substantially shorter than the dawn to dusk cycle observed in anoles at the study site (Fobes et al., 1992: Moster et al., 1992; Smith, 1991). Shuttling in and out of the sun occupied much of the activity period. Marcellini and Jenssen (1989) described Leiocephalus schreibersii as a precise thermoregulator that maintained relatively constant optimal temperatures by moving from sun to shade in an alternating fashion. Lizards spent much of the morning (0900-1200) on elevated perches such as piles of debris, exposed tree roots, or rocks. Postures involved orientation at right angles to prevalent sunlight and extension of the front legs to elevate the head and anterior portion of the body while the posterior portion of the body rested on the substrate. We observed lateral compression only during agonistic encounters. Lizards also may have used this posture, although not consis- tently at right angles to prevailing sunlight, to monitor their immediate surroundings for food or competitors. From 0900-1200, time spent in the sun ranged from 2-70 min ( X =24.5421.4 min, n=20), in the shade from 2-58 min ( X =22.7+19.4 min, n=18). The amount of time in the sun decreased significantly after 1200 (1-26 min, X =9.646.9 min, n=14; Mann-Whitney U, z=-2.08, P=0.04). Time in the shade from 1200-1400 ranged from 2-77 min ( X =23.2+26.0 min, n=13). During this time, movements in and out of the sun or attempts to catch prey were frequent and were characterized subjectively as highly energetic, generally consisting of brief spurts of 1-2 m with tails held high. This flurry of activity may be due to the stenothermic nature of L. schreibersii (Marcellini and Jenssen, 1989). From 1400-1600 activity decreased, lizards moved infrequently, and more time (17/21 observations) was spent sitting in shaded areas such as the flower beds or openings of refuges. After 1600 we rarely saw lizards. Marcellini and Jenssen (1989) found that cloacal temperatures of L. schreibersii did not vary with those of the substrate and were maintained at much higher levels. In contrast, we found cloacal temperatures (n= 13, 28.5-38.8°C, xX =34.90+2.75°C) were usually, but not significantly, lower than those of the substrate (n=13, 26.8- 43.0°C, X =35.9544.60°C)(DF=12, x°*5.51, P=0.94) (Table 1). Our data suggest thermal conformity, although observed behavior, presumed to be thermoregulatory, indicated otherwise. The presence of a large building in the immediate vicinity may have altered local conditions and influenced the thermal regimes of lizards in this population. Unpublished temperature data we have collected from L. schreibersii in the Valle de Neiba and the Llanos de Azua are similar to those reported by Marcellini and Jenssen (1989). Behavior—Males actively defended specific territories by aggressive displays and, occasionally, combat. Aggressive displays involved two types of head bobbing. Lizards performed the first type, hereafter referred to as a “large bob” (Fig. 1), at the beginning of each display. This movement consisted of large, slow, alternating downward and upward motions of the head. The second type, or “small bob,” consisted of very rapid, low amplitude oscillations of the head. Each display began with the head held high. Lizards repeated large bobs 3-6 times, followed by 3-8 (n=7, X =4.9) small bobs in rapid succession, after which two additional large bobs were followed by a variable number of small bobs (n=14, 3-21, X =7.0). We observed this 86 FLORIDA SCIENTIST [VOL 56 TaBLE 1. Cloacal and environmental temperatures of Leiocephalus schreibersii from Barahona, Domini- can Republic. Substrate temperatures were taken with a shaded bulb in contact with the surface and sheltered from the wind. Air temperatures were taken with shaded bulbs sheltered from the wind. Temperatures in °C Time Cloacal Substrate Air at 2 cm Air at 1m 28.5 28.8 28.2 28.0 0826 36.0 38.0 33.5 32.1 1020 32.0 30.0 30.0 30.0 1041 36.2 42.0 36.0 33.0 1050 34.0 35.2 30.4 29.0 1120 34.4 42.0 38.6 32.0 1150 38.0 43.0 38.0 32.0 1315 36.5 32.4 32.6 31.0 1407 38.8 33.4 40.0 32.8 1410 32.4 38.2 Oi.2 33.4 1445 36.4 36.8 34.8 34.0 1449 36.0 35.0 32.0 31.4 1455 34.5 32.2 32.2 32.2 1527 general pattern, repeated for the duration of the display, seven times between male lizards and > 20 times when directed toward persons who approached males too closely (~5 m). Noticeable lateral compression accompanied head bobbing in seven instances. We observed actual combat four times, always among males. Three additional incidents of territorial aggression resulted from intentional introductions of lizards into occupied territories, one of which resulted in displace- ment. We introduced a large male (SVL 76 mm) from outside the area into the territory of a resident male (SVL 71 mm). These individuals spent an entire day in territorial displays and fighting. The next day we found the resident in an adjacent territory, which had previously been the domain ofa smaller male (SVL61 mm), who had shifted his territory by ca. 15 m. The largest male then forced the original resident out of this newly acquired territory, resulting in an additional territorial shift of some 10 m. Each of the three male-male encounters recorded began with a “face off” (sensu Bels, 1986). Lizards oriented themselves in a parallel fashion 0.5-1.0 m apart, facing either in the same or opposite directions. Alternating displays preceded a charge and No. 2, 1993] SCHREIBER ET AL.-LEIOCEPHALUS SCHREIBERSII 87 Fic. 1. Display action patterns of male Leiocephalus schreibersii. drawn from videotape by A. Lathrop. The drawing illustrates “large bobs.” Diagrams illustrate variations observed; some short sequences were observed more than once. Large symbols indicate “large bobs,” small symbols “small bobs.” Each horizontal sequence represents a single display by one individual. Line equals 2 sec. collision, invariably initiated by the resident. During one fight the resident retreated after the initial face off, then charged the introduced male, who was apparently unaware of the approach, hit it with his head and tail, causing the intruder to go rolling across the ground, whereupon it was chased entirely out of the study area. We also observed three aggressive encounters between adults and juveniles, two of them staged. We released a juvenile male (SVL41 mm) within | m of an adult male (SVL71 mm). After no initial reaction by either lizard and without warning, the larger male charged the juvenile, chased him around a tree, and then out of the territory. Another juvenile male (SVL 29 mm) fled immediately upon release near a large male (SVL 61 mm). In another instance, a large male (SVL 83 mm) retreated into a hole in the middle of its territory after disturbance by the gardener. A juvenile had previously sought refuge in the same hole. The male chased the juvenile out, bobbed briefly, then returned to the hole. We recorded only one male-female interaction. An adult female (SVL 51 mm) approached a male (SVL 71 mm) basking on a tree stump from her usual perch on a cinder block 4 m away. When she was within 2 m, the male left his perch and followed her onto the stump. After nudging her vent, he moved to the opposite side of the stump and bobbed once. Following a 4 min pause, the female jumped off the log, moved about 1 m, and executed a head bob directed away from the male. In response the male arched his back and executed a series of large bobs, high in both amplitude and frequency. We observed this pattern only this one time. The male 88 FLORIDA SCIENTIST [VOL 56 subsequently approached the female, nudged her again, upon which she retreated to her cinder block. The sequence was repeated three times, one approach involving a display-action-pattern similar to that of aggressive encounters. The entire incident lasted an hour, with no attempt at copulation. Feeding Behavior and Food Habits—Leiocephalus schreibersii spent very little time (usually < 5 min, rarely as long as 15 min) actively searching for food. Instead, these lizards usually employ a sit-and-wait strategy (Regal, 1983). The most com- monly observed version involved an elevated observation point from which individu- als scanned the surrounding area for suitable prey. When prey was identified, lizards actively pursued and consumed it. Though distance precluded precise evaluations of success, the ratio of success: failure was apparently high. An essentially similar strategy was used in shaded flower beds, especially later in the day, and differed only in the lack of an elevated perch. We observed four instances of active foraging, three of them late in the day (after 1400). After sitting in the shade for 3-5 min, lizards foraged in open sun for 1-2 min before returning to the shade. These searches involved a great deal of activity, lizards moving about quickly with occasional pauses during which they raised their heads and scanned the immediate area. During these forays, lizards were twice observed pausing to root in sand or loose soil. Five of 11 observed incidents of prey capture involved the ingestion of ants; the other six prey items were unidentified larger insects. On three of these occasions, ants had concentrated around a dead beetle or roach. Each time lizards ate the dead insect after consuming some ants. Examination of ingesta from stomachs of 17 previously collected specimens (Fig. 2) revealed that ants (Hymenoptera, Formicidae) were present in greatest numbers and in the most individuals. Orthoptera and Hemiptera constituted the greatest volume. However, the presence of ten orders of arthropods suggests that L. schreibersii is a generalist-opportunist. Larger lizards consumed larger prey, but were apparently not size selective. Mean prey item size was not significantly correlated with SVL (DF=15, r=0.37, P=0.15), nor were differences in food importance values between males and females statistically significant (DF=21, t=0.07, P=0.95). Many Leiocephalus frequently consume plant material (Schoener et al., 1982; Jenssen et al., 1989). Although never observed in this study, plant material was present in the stomachs of five males. This material consisted largely of twigs and bits of leaves, items of limited nutritional value, and may have been ingested adventi- tiously. Consumption of vegetation may be a seasonal or regional phenomenon, neither of which we were able to evaluate effectively. Our unpublished data on stomach contents of L. schreibersii collected at various times from throughout the Valle de Neiba indicate that the data presented here are representative of southern populations. We found no evidence of cannibalism, apparently common in a northern Dominican population (Jenssen et al., 1989). ACKNOWLEDGEMENTS—Thomas Jenssen made a number of helpful comments on an earlier draft of this manuscript. Janet Lynxwiler, Jane Moster, Pete Zani, Julia Smith, Lisa White, Tim Fobes, Paul No. 2, 1993] SCHREIBER ET AL._LEIOCEPHALUS SCHREIBERSII 89 ARACHNIDA ARANEAE ARACHNIDA OTHER COLEOPTERA DERMAPTERA DICTYOPTERA DIPTERA HEMIPTERA HOMOPTERA HYMENOPTERA FORMICIDAE LEPIDOPTERA ORTHOPTERA SHED SKIN 0 100 200 300 400 ARACHNIDA ARANEAE ARACHNIDA OTHER COLEOPTERA DERMAPTERA DICTYOPTERA DIPTERA HEMIPTERA HOMOPTERA HYMENOPTERA FORMICIDAE LEPIDOPTERA ORTHOPTERA SHED SKIN PLANT ARACHNIDA ARANEAE ARACHNIDA OTHER COLEOPTERA DERMAPTERA DICTYOPTERA- DIPTERA HEMIPTERA HOMOPTERA HYMENOPTERA FORMICIDAE LEPIDOPTERA ORTHOPTERA SHED SKIN PLANT 0 2 4 6 8 10 12 Fic. 2. Stomach contents of Leiocephalus schreibersii from Barahona, Dominican Republic. A. numbers of individual food items; B. volumes (in cm’) of types of stomach contents; C. frequencies of occurrence (numbers of stomachs in which the type of item was found). Schell, and Hoa Bui helped in the field. Sixto J. and Yvonne A. Inchdustegui provided assistance in the Dominican Republic. José A. Ottenwalder of the Parque Zoolégico Nacional facilitated opportunities for field work. Permits were provided by Emilio A. Bautista M., Director, Departamento de Vida Silvestre, Republica Dominicana. This investigation was supported in part by Grant No. BBS-9100410 from the National Science Foundation awarded to R. Powell. 90 FLORIDA SCIENTIST [VOL 56 LITERATURE CITED BeLs, V. L. 1986. Analysis of the display-action-pattern of Anolis chlorocyanus (Sauria: Iguanidae). Copeia 1986(4):963-970. Burns, J. K., C. A. CUNNINGHAM, R. A. Dupuis, M. N. Trask, J. S. TULLocu, R. POWELL, J. S. PARMERLEE, Jr., K. Kopecky, anD M. L. Jouuey. 1992. Lizards of the Cayos Siete Hermanos, Dominican Republic, Hispaniola. Bull. Chicago Herpetol. Soc. 27(11):225-232. Conant, R., AND J. T. Coins. 1991. A Field Guide to Reptiles and Amphibians: Eastern and Central North America. 3rd ed. Houghton Mifflin Co., Boston, MA. Foses, T. M., R. Powe, J. S. PARMERLEE, JR., A. LaTHRoP, AND D. D. Situ. 1992. Natural history of Anolis cybotes (Sauria: Polychridae) from an altered habitat in Barahona, Dominican Republic. Carib. J. Sci. 28(3-4):200-207. Frost, D. R. AND R. ETHERIDGE. 1989. A phylogenetic analysis and taxonomy of iguanian lizards (Reptilia: Squamata). Univ. Kansas Mus. Nat. Hist. Misc. Publ. No. 81. JENSSEN, T. A. 1992. Virginia Polytechnic Institute and State University, Blacksburg, Pers. Commun. ,D. L. MarcE Lint, K. A. BUHLMAN, AND P. A. GororTu. 1989. Differential infanticide by adult curly-tailed lizards, Leiocephalus schreibersi. Anim. Behav. 38(6):1054-1061. Lewis, A. R. AND J. E. Sativa. 1987. Effects of sex and size on home range, dominance, and activity budgets in Ameiva exsul (Lacertilia: Teiidae). Herpetologica 43(3):374-383. MakcELLINI, D. L. AND T. A. JENSSEN. 1989. Thermal ecology of the tropical iguanid lizard, Leiocephalus schreibersi. Amer. Midl. Nat. 122(1):44-50. AND T. A. JENSSEN. 1991. Avoidance learning by the curly-tailed lizard, Leiocephalus schreibersi, implications for anti-predator behavior. J. Herpetol. 25(2):238-241. MixsteabD, W. W. 1957. Some aspects of competition in natural populations of whiptail lizards (Genus Cnemidophorus). Texas J. Sci. 9(4):410-447. MosteR, J. A., R. POWELL, J. S. PARMERLEE, JR., D. D. Smit, AND A. Laturop. 1992. Natural history notes on a small population of Anolis brevirostris (Sauria: Polychridae) from altered habitat in the Dominican Republic. Bull. Maryland Herpetol. Soc. 28(4):150-161. OWELL, R., J. S. PARMERLEE, JR., M. A. Rice, AND D. D. Situ. 1990. Ecological observations of Hemidactylus brookii haitianus Meerwarth (Sauria: Gekkonidae) from Hispaniola. Carib. J. Sci. 26(1-2)-67-70: REGAL, P. J. 1978. Behavioral differences between reptiles and mammals: an analysis of activity and mental capabilities. Pp. 183-202. In: GREENBERG, N., AND P. D. MacLean (eds.), Behavior and Neurology of Lizards: An Interdisciplinary Colloquium. Natl. Inst. Mental Health, Rockville, MD. . 1983. The adaptive zone and behavior of lizards. Pp. 105-118. In: Huey, R. B., E. R. PIANKA, AND T. W. SCHOENER (eds.), Lizard ecology: Studies of a Model Organism. Harvard University Press, Cambridge, MA. Rose, B. 1982. Lizard home ranges: methodology and functions. J. Herpetol. 16(3):253-269. SALCEDO, R. L., J. CZERWENKa, AND E. Bo ay. 1983. Atlas de diagramas climaticos de la Republica Dominicana. Secretaria de Estado de Agricultura. Santo Domingo, R. D. SCHELL, P. T., R. POWELL, J. S. PARMERLEE, JR., A. LATHROP, AND D. D. Situ. 1993. Natural history of Ameiva chrysolaema (Sauria: Teiidae) from Barahona, Dominican Republic. Copeia 1993(3):859- 862. SCHOENER, T. W., J. B. SLADE, AND C. H. Stinson. 1982. Diet and sexual dimorphism in the very catholic lizard genus, Leiocephalus, of the Bahamas. Oecologia 53(1):160-169. ScHwarTz, A. 1968. The Leiocephalus (Lacertilia, Iguanidae) of Hispaniola. III. Leiocephalus schreibersi, L. semilineatus, and L. pratensis. J. Herpetol. 1:39-63. AND R. W. HENDERSON. 1991. Amphibians and Reptiles of the West Indies: Descriptions, Distributions, and Natural History. Univ. Florida Press, Gainesville. Smitu, J. W. 1991. Saint Ambrose Univ., Davenport, Iowa, Pers. Commun. Florida Scient. 56(2):82-90.1993. Accepted: November 6, 1992. No. 2, 1993] LAYNE—REVIEW QO] REVIEW Enge, Kevin M. and C. Kenneth Dodd, Jr. An Indexed Bibliography of the Herpetofauna of Florida. Florida Game and Fresh Water Fish Commission Nongame Wildlife Program Technical Report 11. Pp: 231. Price: Available free from Fla. Game and Fresh Water Fish Comm., Tallahassee, FL. ($2.00 donation to nongame program appreciated) BECAUSE of their unusual abundance and variety, the amphibians and reptiles of Florida have attracted the attention of amateur and professional naturalists and scientists in many fields of biology. The result is a voluminous literature dating back well over 200 years and widely scattered in books, scientific journals, and numerous other kinds of publications. Thus, this comprehensive bibliography of the Florida herpetofauna compiled by Kevin M. Enge and C. Kenneth Dod4, Jr., is a particularly welcome volume. It contains 2,820 references, with a cutoff of January 1992. Every citation is indexed by taxon and subject. Citations pertaining to a given species (subspecies are lumped under species) are listed by subject under the species name; whereas, references in which species are not indentified or were too numerous to list separately are listed under suborder, order, or class, depending upon coverage. The 23 subject categories employed cover all important aspects of ecology and life history, as well as general subjects such as Conservation and Management, Human Exploitation, Husbandry, and Area Inventory. The bibliography is not limited only to references from standard scientific publications. It also covers popular articles, newsletter or newspaper articles, and theses deemed by the authors to contain significant scientific information. Selected unpublished reports of in-house or contracted research of governmental agencies - the so-called “gray” literature - are also included. Such reports, which often contain data of basic scientific value or of use in conservation programs, are usually not widely circulated and thus may be missed by potential users. Even if eventually published, a report may be so condensed that it is still necessary to go to the original document for the raw data. The organization of the bibliography makes it easy to use. Nevertheless, looking up references for a particular species still involves a lot of page turning and one cannot help but wish that a computerized version were available. Hopefully, the authors and the Florida Game and Fresh Water Fish Commission will explore this possibility. This volume will not only be a boon to researchers, students, and amateur naturalists with an interest in Florida herpetology, but should also be helpful to personnel of governmental agencies, consultants, and others involved in such fields as environmental assessment, land-use planning, natural areas protection and management, and environmental education. — James N. Layne, Archbold Biological Station P.O. Box 2057, Lake Placid, FL 33852 99 FLORIDA SCIENTIST [VOL 56 Biological Sciences TRENDS IN NUMBERS OF LOGGERHEAD SHRIKES ON ROADSIDE CENSUSES IN PENINSU LAR FLORIDA, 1974-1992. REUVEN YOSEF’, JAMES N. LAYNE AND FRED E. LOHRER Archbold Biological Station, P.O. Box 2057, Lake Placid, FL 33852. ABSTRACT: The Loggerhead Shrike (Lanius ludovicianus) has been declining in numbers for most of the 20th century and is currently diminishing at about 5% per year. We present data on trends in numbers of shrikes in southcentral Florida based on roadside counts conducted along 505 km of roads, twice a year in summer and winter, from 1974 to 1981, in January 1989, July 1991, and January and July 1992. The annual mean count method indicates that from 1976 to 1992 the winter population declined at the rate of 37%, and the summer counts by 41%. The magnitude of the decline documented by us is greater than that recorded by Breeding Bird Surveys for the species nationwide. LOGGERHEAD Shrikes (Lanius ludovicianus) are prominent birds of open habitats and are important as an indicator species of environmental degradation because they are predatory and closely associated with agricultural areas (Hands et al., 1989). Once relatively common throughout much of North America, the Loggerhead Shrike has been declining in numbers for most of the 20th century and is currently diminishing at about 5% per year (Hess, 1910; Graber et al., 1973; Bystrak and Robbins, 1977; Geissler and Noon, 1981; Morrison, 1981; Burnside and Sheperd, 1985; Hands et al., 1989). It is one of the few species to exhibit significant declines in Breeding Bird Surveys (BBS) in all continental regions (Robbins et al., 1986). During 1966 - 1989, 37 of 43 states and provinces of the U. S. and Canada showed negative trends, 25 of which were statistically significant (Droege and Sauer, 1990). Regions most affected appear to be those with breeding populations of the migratory subspecies (Bystrak, 1983). The Loggerhead Shrike has been included in the National Audubon Society’s Blue List since 1972 (Tate, 1986) and is under consideration for listing as Threatened or Endangered by the FWS”. The species is considered Extirpated in 4 states, Endangered in 5; Threatened in 2; and of Special Concern in 3 (Hands et al., 1989). In a recent Status Review of the southeastern states by the FWS, 71% of the respondents opposed Federal listing at present, 17% favored listing as Threatened, and 12% did not comment (Flemming, 1991). Several factors have been suggested as causes for the decline of the Loggerhead Shrike (Porter et al., 1975; Busbee, 1977; Anderson and Duzan, 1978; Craig, 1978; Kridelbaugh, 1982; Bystrak, 1983; Cadman, 1985; Hands et al., 1989), while loss of foraging habitat and hunting perches to modern agricultural practices in the last 'Department of Zoology, The Ohio State University, 1735 Neil Avenue, Columbus, OH 43210; present address: Archbold Biological Station. 2FWS: United States Fish and Wildlife Service No. 2, 1993] YOSEF ET AL.—TRENDS IN NUMBERS OF LOGGERHEAD SHRIKES 93 decade may be the most likely explanation (Cadman, 1985; Hands et al., 1989). Flemming (1991) cautioned that shrike population trends should be considered in a historical perspective. Much of the eastern U. S. was forested in pre-settlement times, so the species may have increased its range as the result of widespread clearing of forests for agriculture. Thus, the current decline may reflect loss of suitable agricultural and other open habitats due to changing agricultural practices (Novak, 1989). Three major strongholds of Loggerhead Shrike populations are considered to be peninsular Florida, Oklahoma, and New Mexico (Robbins et al., 1986). Cox (1987) showed that BBS indicate a statewide decline of approximately 3.7 % annually. Here, we present additional data on trends in numbers of shrikes in southcentral Florida based on roadside counts conducted twice a year in summer and winter from 1974 to 1981, in January 1989, July 1991, and January and July 1992. METHODS—Censuses were conducted in January and July along 505 km (314 miles) of roads in Highlands, DeSoto, Charlotte, Glades, Hardee and Okeechobee counties. The total route was divided into four transects (Appendix 1) of 158 km (98 miles), 104 km (65 miles), 116 km (72 miles), and 127 km (79 miles) respectively, and surveyed on consecutive days. Maps of the census routes are on permanent file at the Archbold Biological Station. he predominant habitat sampled by the census route was open, improved pasture with widely spaced trees. Other habitat types included bottomland and upland forests, citrus groves, marshes and wet prairies, pine flatwoods, scrub and urban areas. All roads were bordered by utility lines, grassy road shoulders and, for most portions of the route, roadside ditches. Loggerhead Shrikes observed within a distance of a quarter (1/4) mile from the road were recorded from a vehicle travelling at 50-58 km/h (30 - 35 mph), and the odometer reading was recorded to allow plotting the location of each sighting on maps. Notes were taken on habitats, weather, perch selection and other details. Surveys started at 0800 hr and usually ended before 1200 hrs. All transects were driven in the same direction in all the surveys. Four observers participated in each survey. Regression analysis was performed with a Statview program; all data presented are mean +SE, unless otherwise specified. We chose P = 0.05 as the minimum acceptable level of significance. RESULTS—Numbers of shrikes on the 505 km census route from 1974 to 1992 varied from 111 to 336 (mean of 233.4 + 77.0) in winter and from 95 to 234 (mean of 158.5 + 41.3) in summer (Table 1). The difference between the January and July censuses, for all counts combined, was significant (Chi-square = 442.6, 7 DF, P = 0.0001), as was the difference for each year (Chi-square = 9.03 - 42.34, P < 0.05). A pronounced increase in numbers occurred between 1974 and 1976, but populations in July and January steadily declined during the period 1976 to 1992 (Table 1). Each transect route was analyzed separately from 1974 to 1992 to determine if the trend was due to particular routes. Significant declines occurred on three of the four routes (route 1: r? = 0.4, F-test = 5.6, P = 0.05; route 2: r? = 0.5, F- test = 8.3, P = 0.02, route 3: r? = 0.2, F-test = 1.5, P = 0.3; route 4: r? = 0.5, F-test = 9.7, P = 0.01). From 1976 to 1992 the winter counts decreased by 37%, the summer counts by 41% (i.e., a decline of 1.4% and 1.6% per year, respectively, over 16 years). In January 1976, 0.67 individuals were observed per kilometer, but this number declined to 0.25 in 1992 (Table 1). Numbers on the July census decreased from 0.46 individual per mile in 1976 to 0.18 in 1981. The rates of decline in summer and winter were strongly correlated; r? value was 0.94 between each July and the following January. 94 FLORIDA SCIENTIST [VOL 56 TABLE 1. Number of Loggerhead Shrikes observed per km in southcentral Florida on a 505 km roadside census route, 1976-1989, and percentage change from one year to the next. Year January July Ind/km % change Ind/km % change 1974 0.36 0.25 1975 0.49 +34 0.35 +36 1976 0.66 OT 0.46 +34 1977 0.65 -2 0.39 -16 1978 0.54 -18 0.34 -14 1979 0.52 -3 0.37 +12 1980 0.51 =) 0.26 -3] 1981 0.41 -20 0.27 “5 5) 1989 0.22 =6 199] 0.25 1992 0.25 0.18 Difference between: January and July censuses Chi-square = 442.6, DF = 7; P = .0001 Each year Chi-square = 9.03 - 42.34, P = 0.05 DiscusslION—The results of our roadside censuses and those of Bohall-Wood (1987) indicate that shrike populations in Florida are larger in winter than in summer. The increased abundance in winter may be due to wintering migrants, residents and young of the year, or acombination of both. Burnside (1987) suggested that Loggerhead Shrike populations east of the Rocky Mountains migrate partly or wholly to the southeastern states for the winter, but presently there is no evidence to indicate that northern individuals winter in peninsular Florida. Banding data for Loggerhead Shrikes in the USFWS Office of Migratory Bird Management contain records of only 11 banded shrikes recovered in Florida, all of which were banded there. Additional evidence indicative of no significant movement of northern birds into southcentral Florida in winter was obtained in a study of Loggerhead Shrikes by the senior author on a 4,200 ha ranch located within the area of the roadside census routes. All resident adults and fledglings were color-banded during the summers of 1990 and 1991, and in neither year did any unmarked shrikes appear on the study area during winter. The boundary of the ranch study area is adjacent to the nearest segment of the roadside census route, so migrants might have settled along roads rather than in areas remote from roads. Further support for residents rather than migrants as the source of increased numbers of shrikes along roads in winter is the close similarity in the trends of summer and winter counts (1? = 0.94 between counts in July and the following January). Such a strong correlation would not be predicted if the winter increase was due largely to migrants because yearly differences in breeding success and winter temperatures in the north would be expected to result in year-to-year variation in the migrating population, independent of resident numbers. No. 2, 1993] YOSEF ET AL—TRENDS IN NUMBERS OF LOGGERHEAD SHRIKES 95 If the increase in roadside counts in winter largely or entirely reflects a change in the resident population, at least part of the build-up probably represents juveniles produced the previous breeding season. Adults whose territories are located along roads also might tend to shift their activity to roadsides in winter in response to higher prey abundance and/or vulnerability in the short grass of road shoulders (Hill, 1976; Carlson, 1985). In cool weather insects may be more active in short-grass areas than in habitats with taller ground cover because of greater insolation. The presence of utility poles and wires and fences for perching presumably is an additional factor contributing to the value of roadsides as winter shrike habitat. The positive trend from 1974 to 1976 was not associated with any significant change in habitats of the census routes and thus may have reflected a period of favorable environmental conditions leading to higher reproduction and/or survival unless habitat changes in a wider peripheral geographic area influenced the popu- lation in our census area. The 1981 numbers approximate those of 1974, so one could argue that the curvilinear, downward trend from 1976 to 1981, may not reflect along- term decline but rather is part of a “normal” fluctuation, perhaps resulting from years of low productivity and/or survival. However, the very low three groups of counts in 1989, 1991, and 1992 further support the contention of a long-term decline. If the positive relationship between winter and summer counts in 1974-1981 was real, the decline to the July 1992 level seems ominous. The annual mean counts method, which represents a species average density (Robbins et al., 1986), indicates that the population of Loggerhead Shrikes on our census routes in southcentral Florida declined 8.9% per year between 1976 and 1981, anda further 5.5% per year between 1981 and 1992. The magnitude of the decline is greater than that recorded by BBS nationwide (Robbins et al., 1986; Droege and Sauer, 1990), and for southcentral Florida in particular (1986-1989, -3.5%). Ifthe current trend continues, southcentral Florida may lose Loggerhead Shrikes from its breeding fauna. The problem of distinguishing between population fluctuations of long periodicity and a true long- term decline clearly indicates the need for continuous, long-term monitoring of Florida’s Loggerhead Shrike populations. Peninsular Florida presently is experiencing significant land-use changes that affect shrikes negatively. The increasing human population is resulting in increasing urbanization and development, and in the southern one-third of the peninsula extensive conversion of pasture-lands to citrus groves and row crops is occurring. Southcentral Florida is one of the three major strongholds of Loggerhead Shrikes (Robbins et al., 1986), hence the present evidence of decline is of particular concern. The Loggerhead Shrike is but one of a group of open-area/prairie birds that is declining. Northern Bobwhites (Colinus virginianus) have also undergone extensive declines (Brennan, 1991), as have Barn Owls (Tyto alba), and Henslow’s Sparrows (Ammodramus henslowii) (Dunning, 1989). However, none of these other open- area species of special concern exhibits continent-wide declines to the extent recorded for the Loggerhead Shrikes. ACKNOWLEDGMENTS—We thank C. E. Winegarner, D. Amadon, M. McCauley, L. C. Layne, D. Yosef, O. Amadon, J. Stallcup, M. MacMillian, P. Martin, T. Bancroft, S. Halkin, W. Sheehan, E. J. Fisk, 96 FLORIDA SCIENTIST [VOL 56 N. Olds, L. Saul, R. B. Root, D. Carter, J. M. Joseph, S. Joseph, B. Millsap, B. Pranty, and R. Titus for their help as observers in the surveys. We thank D. Bystrak, Bird Banding Laboratory, Office of Migratory Bird Management, USFWS, for supplying band-recovery data for Florida, and T. A. Bookhout, A. S. Gaunt, T. C. Grubb, Jr., and an anonymous reviewer for commenting on earlier drafts of this manuscript. LITERATURE CITED ANDERSON, W. L. AND R. E. DUZAN. 1978. DDE residues and eggshell hinning in Loggerhead Shrikes. Wilson Bull. 90:215-220. BOHALL-WOOD, P. 1987. Abundance, habitat use, and perch use of Loggerhead Shrikes in north-central Florida. Wilson Bull. 99:82-86. BURNSIDE, F. L., AND W. M. SHEPHERD. 1985. Population trends of the Loggerhead Shrike in Arkansas. Arkansas Acad. of Sci. Proc. 39:25-28. BURNSIDE, F. L. 1987. Long distance movements by Loggerhead Shrikes. J. Field Omithol. 58:62-65. BUSBEE, E. L.1977. The effects of dieldrin on the behavior of young Loggerhead Shrikes. Auk 94:28-35. BysTRAK, D. 1983. Loggerhead Shrike (Lanius ludovicianus). Pp. 301-310 In: ARMBUSTER, J.S. (ed.). Impacts of coal surface mining on 25 migratory bird species of high federal interest. U.S. Fish and Wildlife Service FWS/OBS-83/35. AND C. S. ROBBINS. 1977. Bird population trends detected by he North American Breeding Bird Survey. Pol. Ecol. Stud. 3:131-143. CADMAN, M. D. 1985. Status report of the Loggerhead Shrike in Canada. Unpublished report to the Committee on the Status of Endangered Wildlife in Canada. 97 pp. CARLSON, A. 1985. Prey detection in the Red-backed Shrike: an experimental study. Anim. Behav. 33:1243-1249. Cox, J. 1987. The Breeding Bird Survey in Florida: 1969-1983. Fla. Field Nat. 15:29-56. CRAIG, R. B. 1978. An analysis of the predatory behavior of the Loggerhead Shrike. Auk 95:221-234. DROEGE, S., AND J. R. SAUER. 1990. North American Breeding Bird Survey annual summary, 1989. US Fish Wildl. Serv. Biol. Rep. 90(8):1-22. DUNNING, J. B., Jr. 1989. Management of nongame migratory birds in farmland, suburban and urban habitats. PP 153-163 In: Proc. nongame migratory bird workshop. USFWS, Atlanta, Georgia. FLEMMING, D. P. 1991. Loggerhead Shrike status survey in the southeast region. USFWS, Atlanta, Georgia. Unpubl. ms. GEISSLER, P. H., AND B. R. NOON. 1981. Estimates of avian population trends from North American Breeding Bird Survey. Stud. Avian Biol. 6:42-51. GRABER, R. R., J. W. GRABERAND E. L. KIRK. 1973. Illinois birds. Laniidae. Illinois Nat. Hist. Surv. Biol. Notes 83. 18 pp. HANDS, H. M., R. D. DROBNEY, AND M. R. RYAN. 1989. Status of the Loggerhead Shrike in the northcentral United States. Unpublished report 0 the U.S. Fish and Wildlife Service, Missouri Cooperative Fish Wildl. Res. Unit, Univ. of Missouri, Columbia, Missouri. HEss, I. E. 1910. One hundred breeding birds of an Illinois ten-mile radius. Auk 27:19-32. HILL, H. R. 1976. Feeding habits of the Ring-necked Pheasent chick, and the evaluation of available foods. Ph.D. diss. Michigan State Univ., East Lansing. 84 pp. KRIDELBAUGH, A. L. 1982. An ecological study of Loggerhead Shrikes in central Missouri. M.S. Thesis. Univ. of Missouri, Columbia, Missouri. 114 pp. MoRRISON, M. L. 1981. Population trends of the Loggerhead Shrike in the United States. Am. Birds 55:754-757,. Novak, P. G. 1989. Breeding ecology and status of the Loggerhead Shrike in New York state. M.S. Thesis. Cornell Univ., Ithaca, New York. 156 pp. PORTER, D. K.,M. A. STRONG, J. B. GIEZENTANNER, AND R. A. RYDER. 1975. Nest ecology, productivity, and growth of the Loggerhead Shrike on the shortgrass prairie. Southwest Nat. 19:429-436. ROBBINS, C. S., D. BYSTRAK, AND P. H. GEISSLER. 1986. The Breeding Bird Survey: its first fifteen years, 1965-1979. US Fish Wildl. Serv. Resour. Publ. 157. 196 pp. TATE, J., Jr. 1986. The Blue List for 1986. Am. Birds 40:227-236. Florida Scient. 56(2):92-97.1993. Accepted: December 8, 1992. No. 2, 1993] YOSEF ET AL—TRENDS IN NUMBERS OF LOGGERHEAD SHRIKES 97 APPENDIX 1. Census route along 505 km (314 miles) of roads through Highlands, DeSoto, Charlotte, Glades, Hardee and Okeechobee counties. The total route was divided into four transects. 1) Commence at junction US 27 and SR 70; W on SR 70 to SR 31; S on SR 31 to SR 74; E on SR 74 to SR 29; E on SR 29 to US 27; N on US 27 to SR 70. Included are portions of Highlands, DeSoto, Charlotte and Glades counties. 158 Km (98 miles). 2) Commence at junction US 27 and SR 70; N on US 27 to SR 66; W on SR 66 to SR 17; S on SR 17 to SR 70; E on SR 70 to SR 31. Included are parts of Highlands, Hardee and DeSoto counties. 104 km (65 miles). 3) Commence at junction US 27 and SR 70; E on SR 70 to CR 721 (at Brighton); S on SR 721 to SR 78; E on SR 78 to US 441; N on US 441 to SR 70; W on SR 70 to CR 721. Included are portions of Highlands, Glades and Okeechobee counties. 116 km (72 miles). 4) Commence 3 km east of US 27 on SR 70 at intersection with SR 29; N on SR 29 to CR 619; N on CR 619 to CR 621; E on CR 621 to US 98; E on US 98 to CR 68; CR 68 to US 441; S on US 441 to SR 70; W on SR 70 to US 98; N on US 98 to CR 68; (continue W on US 98 to CR 721 - this section is not counted); S on CR 721 to SR 70. Included are portions of Highlands and Okeechobee counties. 127 km (79 miles). 98 FLORIDA SCIENTIST [VOL 56 Environmental Chemistry ADSORPTION OF SEVERAL ATMOSPHERIC POLLUTING GASES UPON DEHYDRATED GYPSUM ROBERT F.. BENSON! AND GEORGE D. BLYHOLDER Department of Chemistry, University of Arkansas, Fayetteville, Arkansas 72701 ABSTRACT: This work explores surface interactions between gypsum and selected prominent gaseous emissions in to the atmosphere. Gypsum is a common air-borne mineral particulate that has a potential two-fold relationship to the air pollution problem: as a particulate pollutant and as a catalyst or adsorbent for pollutant gases. The Florida atmosphere represents a situation where sources of gypsum particulates are especially abundant from both natural sources and anthropogenic activities. This work explored the adsorption of eight gases (nitric oxide, nitrous oxide, ammonia, hydrogen sulfide, carbon dioxide, carbon monoxide, oxygen, and propane) at 24°C as representative of gases present in the atmosphere. Propane, hydrogen sulfide, and oxygen were not adsorbed at pressures less than one atmosphere. Coverages were related to a monolayer based upon a surface area of 17.0 m?/gm as determined from a nitrogen adsorption isotherm and the B.E.T. method. Multilayer coverages were observed for ammonia, nitric oxide, carbon dioxide, and sulfur dioxide. Both reversible and irreversible adsorption were observed. Irreversible coverages of 0.52, 0.37, 0.41, and 0.33 monolayer were observed for SO,, CO,, NO, and NH,, respectively. Nitrous oxide and carbon monoxide did not have an irreversible component to their adsorption. GypsuM (CaSO,°2H,O) is acommon air-borne particulate that appears to have a two-fold relationship with the air-pollution problem. First, particulate matter has been identified by many investigators as an important component of air pollution (Cadle, 1966; Corn, 1976; Kellog et al., 1972). The chemical composition of atmo- spheric aerosol particulates shows CaSO, to be an abundant salt among the variety of salts possible from the Ca**, Mg**, Na*, NH,*, Fe***, SO,-, Cl, NO, and CO,— ions most commonly found (Cadle, 1966; Corn, 1976). Sulfate salts, such as gypsum, have been shown to have adsorption and catalytic activity (Tanabe and Takeshita, 1967). Second, gypsum and other sulfate salts have been known to have catalytic and adsorption properties which could be important as air pollution sinks. Potential surface interactions between gaseous air pollutants and particulates could influence air pollution characteristics. Florida represents an environmental situation where sources of gypsum par- ticulates are especially abundant (Davis, 1989). Sulfate salts are emitted into the atmosphere as aerosols from sea spray, dusts from industrial activity, and are formed from the oxidation of H,S and SO, emissions (Kellog et al., 1972; Robinson and Robbins, 1968). Gypsum particles can be distributed directly into the Florida atmosphere from industrial activity in fertilizer manufacture, construction and electrical power generation as well as from natural sources such as sea spray. Oxidation of the atmospheric sulfurous gases usually produces sulfuric acid aerosols 1 Present Address: Institute for Environmental Studies and Department of Chemistry, University of South Florida, Tampa , Florida 33620. No. 2, 1993] BENSON ET AL.—ABSORBTION OF SEVERAL POLLUTING GASES 99 that react to form sulfate salts. These sulfate aerosols account for much of the sink mechanism and for the bulk of the residence time (20-30 days for SO,— as compared with 1 day for H,S and 4-6 days for SO,) of the atmospheric sulfur compounds (Robinson and Robbins, 1968). Direct combination of SO, at ambient temperatures with oxides, metals, and other salts are potential routes of the sulfur sink mechanism leading to sulfates. Sulfates have been observed as a result of the reaction of SO, with transitional metal films at ambient temperatures (Blyholder and Cagle, 1971). Sulfur compounds, having direct bonding of metal to sulfur available, will react directly with transition metals over a wide temperature range (Maxted, 1956). At 25°C, SO, has been observed to adsorb on CaO (Low et al, 1971) and to react with CaO or CaCO, at elevated temperatures (350°C) to form CaSO, (Sprung, 1965; Pechkovski, 1964). It would seem evident that sulfates could be present to interact with other gaseous pollutant emissions. This work explores surface interactions between gypsum and selected prominent gaseous emissions in to the atmosphere. Catalysis by gypsum and other sulfates has been relatively unexplored in comparison with oxide and metal catalysts, but the literature indicates several interesting examples of sulfate catalysis (Benson, 1978). A wide variety of reactions have been observed to be catalyzed by sulfate salts (Tanabe and Takeshita, 1967). As a dehydration catalyst, CaSO, has been observed to catalyze dehydration and isomerization of 2,2 dimethyl-3-butanol to give 2,3-dimethyl-1-butene and 2,3 dimethyl-2-butene (Razouk et al, 1965). A mixed catalyst containing MgSO, and CaCl, was used to dehydrate cyclohexanol at 380°C (Balandin et al., 1961). Alkali and alkaline earth sulfates were used to catalyze the esterification of phthalic and adipic acids (Furukawa and Naruchi, 1962). Bisulfate salts were comparable to sulfuric acid in catalytic activity toward esterification with the best yields obtained between 145- 175°C. Polymerization of ethylene oxide was carried out at 70-120°C over sulfate catalysts (Kylov and Sinyak, 1962; Bobinova and Morozova, 1963). With a gypsum catalyst, a yield of 79% of polymer with a molecular weight range of 30-40,000 was obtained. Ethylenimine was prepared from ammonia and dichloroethane by treat- ment over CaSO, or BaSO, at 25-110°C (Dix, 1965). Elimination of HCl from cis or trans-chlorostilbene was catalyzed by CaSO, and other oxide catalysts (Andreu et al, 1964). Tanabe and Takeshita (1967) reviewed catalysis by sulfates and summarized other reactions as well as those just described. In their review, they correlated the catalytic activity of sulfates with Bronsted or Lewis acid sites. It should be noted that some of the catalytic activity observed for CaSO, occurred at temperatures near atmospheric conditions. Emissions into the atmosphere are exposed to reaction conditions that, through a variety of pathways, can lead to either air pollution or salts recycled back to the soil. Because reaction rates are proportional to temperature and reactant concentration, the relatively low atmospheric temperatures and emission concentrations of a few parts per million permit unusually long lifetimes (2-3 days) for some of the intermediates. Oscillatory patterns,.such as between daylight and darkness or high and low daily temperatures have been shown to be important in air pollution phenomena (Haagen-Smit, 1972; Haagen-Smit and Wayne 1976). 100 FLORIDA SCIENTIST [VOL 56 MATERIALS AND METHODS—Gases used for this adsorption study (hydrogen sulfide, carbon dioxide, carbon monoxide, oxygen, propane and sulfur dioxide) were used as received. The nitrogenous gases (nitrous oxide, nitric oxide, nitrogen dioxide and ammonia) were purified by vacuum sublimation. Purity was confirmed by mass spectral analysis. Gypsum Preparation— Analytical reagent grade calcium sulfate hydrate was the initial material for the adsorbates used in this work. The gypsum was prepared by dehydration in order to maximize the surface area. The sample was evacuated at room temperature and then was heated while under vacuum to approximately 110°C over a period of 90 minutes. This heating schedule resulted in a sample weight that remained stable at 10° torr pressure and gave a relatively high surface area for the adsorption study. Adsorption of Gases—A vacuum system containing an electronic recording microbalance was used to observe the adsorption on calcium sulfate. Mass changes were detected by a Cahn RG Electrobalance and recorded by a Sargent Model SRG Recorder. Pressure was monitored by means of a General Electric thermistor vacuum gauge in the micron pressure region and by means of a Wallace and Tiernan diaphram manometer in the range of 1-760 torr. Sample temperature was monitored with a copper-constant thermocouple referenced to an ice water bath and recorded by means of a millivolt recorder. Provisions were made in the vacuum system for introducing adsorbate gases and for isolating and removing gas samples. The signals from the instrumentation components were calibrated against standards. Pressure was calibrated against readings from a MacLeod gauge. The microbalance signal was calibrated using N. B. S. class M weights. Thermocouple output was correlated with literature values (Hodgman, 1960). Adsorption isotherms were determined for each gas as follows. A small volume of gas from a reservoir was admitted into the microbalance/ sample gas chamber. Pressure, time, temperature and mass were recorded and observed until the mass became constant. Successive volume increments were added until the isotherm was defined. Then the gases were evacuated from the sample and the irreversible adsorbed fraction determined. RESULTS AND DiscussioN—These results show that partially dehydrated cal- cium sulfate particulates are capable of interaction with gaseous emissions. They can act as a sink for emissions of sulfur dioxide, carbon dioxide, nitric oxide and ammonia by adsorption and subsequent particulate precipitation from the atmosphere. Gypsum probably is involved in the Florida atmospheric sinks because of the abundant sources. The phosphate industry in Florida produces 33 million tons of gypsum per year as waste from phosphoric acid production. In addition there are over 400 million tons stockpiled in phosphogypsum stacks (Chang, 1987). It would seem likely that wind dispersion would generate airborne gypsum particulates from this source. Florida has to rely on coal-fired electrical power generating plants many of which yield gypsum as a waste product from removal of sulfur emmisions. Seaspray is a natural source of gypsum (Robinson and Robbins, 1968) and the ample coastline provides a contribution to the air over Florida Gypsum’s adsorption properties are dependent on the surface area available after dehydration. Gypsum has a significant water vapor pressure at room tempera- ture. Water vapor pressures of 9.12 torr and 14.29 torr at 25°C have been reported (Washburn,1930) for the equilibrium between the dihydrate-hemihydrate and dihydrate-anhydrous forms of calcium sulfate, respectively. These data show, that under experimental conditions for measuring adsorption, the water vapor pressure would interfere with adsorption. Surface Preparation—Thermal activation of calcium sulfate used a technique that would produce the largest surface area and the most stable weight. Treatment of gypsum under vacuum for several days or thermal treatment with either a heating schedule of 25 -150°C over a period of 60 minutes or 25 - 110°C over a period of 90 minutes yielded weight losses at 18.0 +/- 0.2 % w/w. This weight loss was both No. 2, 1993] BENSON ET AL.—ABSORBTION OF SEVERAL POLLUTING GASES 101 reproducible and constant at one micron pressure. The sample weight as a function of temperature is shown (Fig. 1). This relationship shows that the dehydration process proceeds directly to a limiting weight without forming the hemihydrate as an intermediate. Surface areas were measured relative to the amount of water lost. An average surface area of 17.0 m’/gm was determined for the thermally activated 18.0% weight loss sample. Surface areas were determined from nitrogen isotherms using the B. E. T. method (Thomas and Thomas, 1967). Surface areas corresponding to weight changes are shown (Fig. 2). Two high surface area regions were observed. The first region occurred at a weight loss slightly less than the hemihydrate conversion, and the second region occured at a weight change slightly less than conversion to the anhydrous calcium sulfate. Apparently the transition into the relatively stable forms resulted in a loss of surface area. Prolonged simple evacuation to one micron pressure required several days to reach a constant weight (82%), but the resultant surface area was not large enough to measure. Adsorption was carried out on high area samples that had not undergone the prolonged evacuation route to water loss. Adsorption work was carried out under conditions closely approximating those expected in the atmosphere. A temperature of 24°C and gas pressure between 10* and 10? torr were the conditions chosen to survey the adsorption on partially dehydrated CaSO,,. The wide range of pressure permits some degree of comparison between physical and chemical adsorption. In practice, chemical adsorption was defined in this work as the irreversibly adsorbed material remaining on the surface after the surface has been evacuated to pressures of 10“ torr or less for a period necessary to attain a constant weight on the microbalance. 100 90 Weight Remaining (%) 80 20 40 60 80 100 120 Temperature (C) Fig. 1. Sample weight profile during thermal dehydration of gypsum. 102 FLORIDA SCIENTIST [VOL 56 Surface Area (m2/g) 0 We) 20 Weight Loss (%) Fig. 2. Surface area changes relative to weight loss during thermal dehydration of gypsum. In addition to adsorption, several other concepts need to be defined as used in this work. Since the microbalance provides a quantitative mass profile of the adsorption, the monolayer has been described in units of mass rather than the more conventional volume units. The fraction of the surface involved in chemisorption is determined from the ratio of irreversible adsorption to the nitrogen monolayer calculated from physical adsorption. An equilibrium between hydration and dehy- dration of the adsorbent adds the possible variation of the surface with changes in water vapor pressure in the system. Therefore the physical surface area must be referenced to the description of a stable partially dehydrated CaSO, rather than gypsum. Adsorption parameters from physical adsorption were related to the individual adsorbates by means of the constants in the van der Waals equation of state. Eqn. 1 was used to calculate the molecular radius and hence the cross sectional area, A, from the van der Waals parameter, b. The mass of the monolayer was calculated. (Eqn. 2). 4nr?N, mn () 3 (SA) (Msample) My (2) Mmono >. ee ko Se (No) (Amolecule) In these equations, r is the radius, b is the van der Waals parameter, SA is the surface area of the adsorbate per gram of adsorbent, Maple is the mass of adsorbent, No. 2, 1993] BENSON ET AL.—ABSORBTION OF SEVERAL POLLUTING GASES 103 M_ is the molecular weight of the adsorbate gas, N, is Avagadro’s number, and A is the molecular cross sectional area. Adsorption Survey—Adsorption of the gases SO,, NO, NH,, N,O, CO,, CO, C,H,, H,S and O, were surveyed on partially dehydrated CaSO, at 24°C. Of these gases, propane, H,S and oxygen were not adsorbed within the limits of detection for coverage, @ > 0.01. The remaining gases were adsorbed with the equilibrium containing both reversible and irreversible fractions. Adsorption survey results are summarized (Table 1) for the purpose of making comparisons between gases and between equilibrium and irreversible adsorption. Several gases—CO,, NO, NH,, and possibly SO,—show multilayer equilibrium adsorption. Of this sangre ore adsorption, only a itcletion of the surface is covered by irreversible adsorption . This pattern is not expected for solid-gas reactions and the experimental pressures are too small for capillary condensation. Absorption is a possible explanation for what appears to be multilayer coverage but then there should be some correlations with time, pressure and temperature that would support this possibility. The order of coverage according to molecular volumes should be NO, NH,, CO, H,S, CO,, N,O and SO, and this is not observed. A property common fe the gases shone ee adsorption is that they are with the exception of NO whichis a free radical, either Bronsted or Lewis acids and bases. From infra-red spectra and the stoichiom- etry of the dehydration there is known to be water and hydroxyl groups present which can be active sites. Irreversible adsorption coverage is similar for CO,, NO, NH,, and possibly SO,,. Since the irreversible coverage may relate to stereospecific attachment of a gas molecule to the surface, the use of van der Waals cross sectional areas may not be completely applicable. However, there is a trend in the irreversible coverage in the order of increasing van der Waals cross sectional area. Since the van der Waals radius Table 1. Summary of adsorption on dehydrated gypsum at 24°C. Gas Monolayer Adsorbate (mg) Pressure Bmax irr (mg) Total Irreversible (torr) SO, 0.576 0.61 0.30 49. 1.06 0.52 CO 0.158 0.10 -- 385. 0.63 -- CO, 0.475 0.843 0.178 147. 1.78 0.37 NO 0.423 0.86 0.125 20. 2.04 0.30 N,O 0.477 0.092 0.016 m9 0.20 0.03 NH, 0.261 2.46 0.086 Zs 9.42 0.33 0.180 0.69 104 FLORIDA SCIENTIST [VOL 56 arises from a statistical averaging of the orientation of the molecule, the average orientation must still be important on the adsorbent surface for these gases except N,O. The trend of the irreversible coverages could be inferred to arise from either a random orientation of the adsorbed molecule— a condition not expected for chemisorbed molecules-or the adsorbed molecules possess enough degrees of freedom to give a variable surface orientation. Averaged over time the area of the surface covered by the adsorbate would follow the same trends as those suggested by van der Waals considerations. In this work, N,O was exceptional,in that both the reversible and irreversible adsorption covered a small fraction of the physical surface, perhaps because the surface interaction is stereochemical and gives rise to a preferred orientation that permits a closer packing or simply that there were not as many sites active to N,O adsorption as for the other irreversibly adsorbed gases. Some more information about these atmospherically prominent gases adsorbed on partially dehydrated CaSO, can be gained directly from the isotherms at 24°C. The extent and shape of the equilibrium adsorption-desorption curve can be related to the type of surface texture, the monolayer and surface stoichiometry. The equilibrium adsorption curve of ammonia is presented (Fig. 3.). The outward appearance of the curve resembles a type I isotherm of the Brunaer classification. This type of isotherm can be described by a Langmuir model for adsorption but the Langmuir model applies only to coverages less than or equal to a monolayer. The adsorption of ammonia approaches 10 layers of coverage based upon a monolayer determined by nitrogen adsorption. Either the ammonia molecule has access to 10 times more surface area than nitrogen or the process observed is not adsorption but rather absorption. Since most of the equilibrium adsorption is weakly bound and 10 Ammonia Coverage 2 —t— + £=Ammonia ——e—-__ Desorb Ammonia 0 20 40 60 80 Pressure (torr) Fig. 3. Adsorption/desorption isotherm of ammonia on dehydrated gypsum at 24°C. No. 2, 1993} BENSON ET AL.—ABSORBTION OF SEVERAL POLLUTING GASES 105 reversible to reduced pressure, stereo-chemical considerations of the surface struc- ture are not important. Relative to the saturation pressure, the pressures leading to multilayers are small enough not to be considered capillary condensation. Therefore it seems unlikely that the interaction of ammonia is totally adsorption. Ammonia is chemically similar to water. It can form hydrogen bonds and donate lone pair electrons in the same conditions to those favorable to water. In addition to adsorption, two types of absorption phenomena can be used to explain the interac- tion of ammonia with partially dehydrated CaSO,,. Adsorption of ammonia accounts for the irreversible layer. Even the multiple of the irreversible adsorption observed in the repeated adsorption curve probably arises from an initial adsorption layer followed by a penetration of this layer into the lattice. Sites on the surface and in the interior of the lattice giving rise to the irreversible behavior are possibly HSO, groups. Absorption can arise from the replacement of lost water of hydration with ammonia or by hydrogen bonding of the ammonia to the remaining water of hydration in the partially dehydrated CaSO. In the dehydration process, gypsum lost 1.13 millimoles of water, while the ammonia absorption equivalent to 10 layers of coverage required 0.15 millimoles of ammonia. Approximately 0.38 millimoles of water still remained with the dehydrated CaSO,. The amount of ammonia absorbed is not enough to replace the water of hydration lost in the dehydration of the gypsum but there is an approximate correspondence of one ammonia molecule hydrogen bonding to two molecules of water remaining in the dehydrated CaSO,. The above stoichiometry is necessarily approximate because the exact amount of moisture in the original gypsum is not known. The extent of adsorption of carbon dioxide on dehydrated CaSO, was not as large as that of ammonia. At 24°C carbon dioxide adsorption approaches two monolayers containing reversible and irreversible fractions as shown (Fig. 4). Irreversible coverage accounted for 37% of the monolayer—a fraction comparable to the irreversible fraction of ammonia. A check for a stoichiometric correlation between the residual water of hydration remaining in the dehydrated CaSO, and the amount of carbon dioxide irreversibly adsorbed did not reveal any correspondence that would relate to a gas-solid reaction. The irreversible fraction of the surface ac- counted for 4.045 x 10° millimoles of CO, compared with 0.02 millimoles of residual water in the bulk. The irreversible fraction may be adsorbed at surface bound residual water sites or at CaO sites located at the surface. Adsorption of sulfur dioxide on dehydrated CaSO , at 24°C produced a type I isotherm of the Brunaer classification (Thomas and Thomas, 1967). This type of isotherm contains both physical and chemical adsorption in agreement with the equilibrium and irreversible coverages actually observed, and an example of the isotherm is shown (Fig. 5). Coverage of a monolayer adsorption for certain sites was complete at 1.0 torr and occupied 49% of the physical surface. This corresponds to the irreversible coverage of 52% found upon desorption. When the adsorption was carried out at higher pressures, the physical monolayer was completely filled at 50 torr. During the desorption a hysteresis loop was developed—an observation usually associated with pore structure. This hysteresis loop is not believed to arise from pore structure but rather from some weak surface interaction. It is probable that the SO,- 106 FLORIDA SCIENTIST [VOL 56 ——tt— Carbon Dioxide —e—-__Desorb CO2 Carbon Dioxide Coverage 0 10 20 30 40 50 60 Pressure (torr) Fic. 4. Adsorption/desorption isotherm of carbon dioxide on dehydrated gypsum at 24°C. —t— _ Sulfur Dioxide —-e—-_ Desorb SO2 Sulfur Dioxide Coverage 0 10 20 30 40 50 Pressure (torr) Fig. 5. Adsorption/desorption isotherm of sulfur dioxide on dehydrated gypsum at 24°C. No. 2, 1993] BENSON ET AL.—ABSORBTION OF SEVERAL POLLUTING GASES 107 partially dehydrated CaSO, interaction is strictly limited to surface phenomena because the amount adsorbed is within the limits of the physically adsorbed monolayer. Nitric oxide was adsorbed onto dehydrated gypsum to give acomplex adsorption isotherm. Multiple layers of coverage and a predominantly irreversible coverage was observed at room temperature as can be seen (Fig. 6). The initial adsorption response to a new pressure was rapid and reversible, but the reversible adsorption changed to irreversible over the duration of the isotherm. The shape of the isotherm at 24°C resembles a Langmuir type, which suggests sites having uniform energy and single occupancy of adsorbate. Around 1-2 torr pressure, the coverage shows a break in the curvature that could be interpreted as a monolayer coverage of sites having a relatively uniform heat of adsorption. This patch of sites accounts for only 41% of the physical surface. At pressures below this break point , the coverage was irreversible. ACKNOWLEDGMENT~—This investigation was supported in part by Research Grant No. AP00818 from the Air Pollution Control Office, Environment Protection Agency. ————it Nitric Oxide —-e—_ Desorbed NO Nitric Oxide Coverage 0 10 20 30 Pressure (torr) Fig. 6. Adsorption/desorption isotherm of nitric oxide on dehydrated gypsum at 24°C. Present Address: Institute for Environmental Studies and Department of Chemistry, University of South Florida, Tampa, Florida 33620. 108 FLORIDA SCIENTIST [VOL 56 LITERATURE CITED ANDREU, P., E. SCHMITZ AND H. NOLLER. 1964. Mechanism of contact eliminations V. HCl elimination from gaseous cis- and trans-chlorostilbene on oxide and salt catalysts. Z. Physik. Chem. F42 (5/6) Pp 270-279. Through Chem. Abstr. 62:7163f. BALANDIN, A. A.,V. I. SPITSYN, N. P. DOBROSEL’SKAYA AND I. FE. MIKHAILENKO. 1961. Radioactive catalysts. Dehydration of cyclohexanol over magnesium sulfate and calcium chloride. Dokl. Akad. Nauk. SSSR 137, Pp.628-630. Through Chem. Abstr. 56:12355e. BENSON, R. F. 1978. Ph.D. Dissert. Univ. of Ark. Fayetteville, Ark. BLYHOLDER, G. D. AND G. W. CAGLE. 1971. Infrared spectra of hydrogen sulfide, carbon disulfide, sulfur dioxide, methanethiol, and ethanethiol absorbed on iron and nickel. Environ. Sci. Technol. 5. 158-164. BOBINOVA, L.M. AND E. P. MOROZOVA. 1963. Polymerization of ethylene oxide in the presence of metallic sulfates. Tr. po Khim. I Khim. Teknol. 296-299. Through Chem. Abstr. 61:7106h. CADLE, R. D. 1966. Particles in the Atmosphere and Space, Reinhold Publishing Corp., New York, N. x. CHANG, W. F. 1987. Reclamation, reconstruction and reuse of phosphogypsum for building materials, Publication No. 01-014-048, Florida Institute of Phosphate Research, Bartow, F'. Corn, M. 1976. Aerosols and the primary air pollutants- Nonviable particles, their occurrence, properties and effects. Pp.78-168. In: STERN, A. (ed.). Air Pollution, Vol. 1. 2nd ed., Academic Press, New York, N. Y. Davis, L. L. 1989. Gypsum, In: Minerals Yearbook, Bureau of Mines, U. S. Dept. of Interior DIx, J. S. 1965. Ethylenimine. U. S. Patent No. 3,205,224. FURUKAWA, S. AND K. NARUCHI. 1962. Esters for plasticizers. Chiba Diagaki Kogakuba Kewnkyu Hokoku 13, No. 23: 15-24. Through Chem. Abstr. 58:627b. HAAGEN-SMIT, A. S. 1972. Light side of smog, Chem. Tech. 2(6): 330-5. AND L. G. WAYNE. 1976. Atmospheric reactions and scavenging processes. Pp.#235-288. In: STERN, A. (ed.). Air Pollution Vol. 1., 3rd ed., Academic Press, New York, N. Y. HODGMAN, C. D. ed.,1960. Handbook of chemistry and physics, 42nd ed.,Pp 2612-2631. Chemical Rubber Company, Cleveland, Oh. KELLOG, W. W., R. D. CADLE, E. R. ALLEN, A. L. LAZRUS, AND E. A. MARTELL. 1972. Sulfur cycle. Science 175: 587-596. KYLOV, O. V., AND Y. E. SINYAK. 1962. Polymerization of ethylene oxide. Neftekhimiya: 2. 688-696. Low, M. J. D., A. J. GOODSEL AND N. TAKEZAWA. 1971. Reactions of gaseous pollutants with solids I. Infrared study of the sorption of sulfur dioxide on calcium oxide. Environ. Sci. Technol. 5: 1191- 1195. MAXTED, E. B. 1956. Poisoning of metallic catalysts. Pp.129-178. Adv. Catal., Vol.3, Academic Press, New York, N. Y. PECHKOVSKI, V. V. 1964. Reaction of calcium oxide with sulfur dioxide under reducing conditions. Zh. Prikl. Khim 37: 240-246. Through Chem. Abstr. 60:11601f. RAZOUK, R. I., R. SH. MICHAIL, AND A SH. SALEM. 1965. Adsorption of cyclohexane on dehydrated gypsum, J. Chem. U. A. R. 6(1): 1-10. Through Chem. Abstr. 63:1237h. ROBINSON, E. AND R. C. ROBBINS. 1968. Sources, Abundance and Fate of Gaseous Atmospheric Pollutants, Final Report of Project PR-6755 of the Stanford Research Institute, Menlo Park, California. SPRUNG, S. 1965. The behavior of sulfur during firing of cement clinker. Schriftenreike Fementind 31: 70-74. Through Chem. Abstr. 64:425e. TANABE, K. AND T. TAKESHITA. 1967. Catalytic activity and acidic property of solid metal sulfates. Pp.315-349. In: Adv. Catal. Vol. 17, Academic Press, New York, N. Y. THOMAS, J. M. AND W. J. THOMAS. 1967. Introduction to the Principles of Heterogeneous Catalysis. Academic Press, New York, N. Y. WASHBURN, E. W. ed., 1930. International critical tables of numerical data, physics, chemistry and technology, Vol. VII. Pp 295-296 for the National Research Council (U. S.). M°Graw-Hill New York, N. Y. Florida Scient. 56(2):98-108.1993. Accepted: November 8, 1992. No. 2, 1993] SIVER AND WUJEK—SCALED CHRYSOPHYCEAE 109 Biological Sciences SCALED CHRYSOPHYCEAE AND SYNUROPHYCEAE FROM FLORIDA: IV. THE FLORA OF LOWER LAKE MYAKKA AND LAKE TARPON “)PETER A. SIVER AND ® DANIEL E. WUJEK ‘) Botany Department,Connecticut College New London, CT 06320; ®) Department of Biology, Central Michigan University Mt. Pleasant, MI 48859 ABSTRACT: Twenty-nine taxa of silica-scaled Chyrsophyceae and Synurophyceae are reported from the spring flora of two relatively eutrophic and humic stained lakes, Lower Lake Myakka and Lake Tarpon. Samples from 1986, 1987, 1989 and 1992 were analyzed. Twenty-seven and fifteen of the taxa were found in Lower Lake Myakka and Lake Tarpon, respectively. Affinities with tropical and temperate floras are noted in both lakes. This represents the first report of Synura petersenii f. truttae Siver and S. australiensis Playfair in Florida, and the second report of the latter taxon in the United States. EXCEPT for the work of Wujek and coworkers (Wujek, 1983, 1984; Wujek and Gardiner, 1985; Wujek and Bland, 1988, 1990) and Siver (1991), very little is known about the flora of silica-scale bearing Chrysophyceae and Synurophyceae from Florida. It is clear, however, based on the previous work, that a rich flora of Chysophyceae and Synurophyceae taxa does exist in Florida waterbodies. Wujek (1984) and Wujek and Bland (1990) reported over forty taxa of Chrysophyceae and Synurophyceae from localities in the central, southern and west coast regions. As many as nineteen taxa were found in a single sample from several waterbodies. The works of Wujek (1983), Wujek and Bland (1988), Wujek and Gardiner (1985) and Siver (1991) contain reports of newly described specific or subspecific taxa, again indicating the need for further investigation. Recently, silica-scale bearing algae have become important as bioindicators of changes in the quality of lake water, especially from a paleolimnological perspective (Siver, 1993a; Smol, 1993 and references therein). Before scaled Chrysophyceae and Synurophyceae could be used as bioindicators in a similar manner in Florida, it is mandatory that the flora be adequately described. This paper is the fourth in a series aimed at describing the flora of Chrysophyceae and Synurophyceae in Florida. The spring floras from two relatively eutrophic and humic stained lakes are described over a six-year period utilizing scanning electron microscopy. All previous work on scale-bearing algae from Florida, except for that of Siver (1991), had been done with transmission electron microscopy (TEM). One of our goals is to assemble a collection of SEM micrographs of the various taxa to complement previously published TEM micrographs. METHODS AND MATERIALS—Plankton samples were collected from the two lakes during the month of March in 1986, 1987, 1989 and 1992. Collections from Lower Lake Mayakka were made from the center of the lake at a depth of 1 m. Collections from Lake Tarpon were made at the end of the walking 110 FLORIDA SCIENTIST [VOL 56 pier located in the community park along the western shore at a depth of 0.5 m. Both water samples and plankton net samples (10 um mesh) were taken and fixed on site with Lugol’s solution. The water and plankton net samples were prepared for analysis with scanning electron microscopy according to Siver (1987). The relative proportions of each taxon were qualitatively estimated from the concentrated water samples and scored according to the following system. If taxa were among the most dominant species of algae in the sample they were scored as “abundant”. Taxa that were sub-dominant, but still important components, were listed as “common”. Taxa that were rare, but whole cells were observed, were listed as “rare”. Taxa where only a few isolated scales were observed, were scored as “very rare”. RESULTS AND DISCUSSION—A total of 29 specific and subspecific taxa of Chrysophyceae and Synurophyceae were recorded from the two localities, including 17 taxa of Mallomonas, five of Synura, four of Spiniferomonas, two of Paraphysomonas and Chyrsosphaerella brevispina (Table 1). A total of 27 and 15 of the taxa were found in Lower Lake Myakka and Lake Tarpon, respectively. Mallomonas caudata (Fig. 1) and Synura petersenii f. truttae (Fig. 2-3) were the most common taxa found in Lower Lake Myakka during the study. Mallomonas caudata was abundant during each of the four years, while S. petersenii f. truttae was either common or abundant. Mallomonas matvienkoae v. myakkana (Fig. 4-5) was observed to be common in 1987, 1989 and 1992. Four additional species, Mallomonas portae-ferreae (Fig. 6-7), M.corymbosa (Fig. 8-9), Synurauvella and Paraphysomonas vestita (Fig. 10) were also found to be common or abundant in Lower Lake Myakka during at least one of the years. All other taxa from Lower Lake Myakka were rare or very rare (Table 1). Taxa in this category, but found during at least two of the years included Mallomonas mangofera f. foveata (Fig. 11), M. tonsurata (Fig. 12), M. striata v. serrata (Fig. 13), M. pseudocoronata (Fig. 14), M. heterospina, Synura petersenii f. petersenii, S. australiensis (Fig. 15) and Spiniferomonas coronacircumspina. Other rare taxa found in Lower Lake Myakka and pictured in this paper include M. akrokomos (Fig. 16), M.annulata (Fig. 17), M. peronoides (Fig. 18), M. bronchartiana (Fig. 19), M. cyathellata v. cyathellata (Fig. 20) and M. papillosa (Fig. 21). The most common taxa in Lake Tarpon were similar to those recorded in Lower Lake Myakka (Table 1). In addition, all but two of the taxa observed in Lake Tarpon were also found in Lower Lake Myakka, indicating the similarities in the two floras. Mallomonas caudata, M. corymbosa and M. tonsurata were the most common species in Lake Tarpon over the study period. Synura petersenii f. truttae and Paraphysomonas vestita were also observed to be at least common during one of the years. Six additional taxa, Mallomonas annulata, M. pseudocoronata, M.matvienkoae v. myakkana and v. matvienkoae, Synura petersenii and Spiniferomonas coronacircumspina were also found during more than one year (Table 1). A note needs to be made concerning Mallomonas matvienkoae v. myakkana, described by Siver (1991) from Lake Myakka. We recognize that this taxon may be the same as M. matvienkoae v. grandis Diirrschmidt & Cronberg described from Sri Lanka (Diirrschmidt and Cronberg, 1989). Scales of v. myakkana differ from the _ type in possessing a cluster of three to five pores in the posterior region and a dense covering of papillae over the distal area. According to the original description of v. grandis, it differs from the type in possessing several (three to five) large subcircular No. 2, 1993] SIVER AND WUJEK—SCALED CHRYSOPHYCEAE LT TABLE 1. The occurrence and relative abundance of 29 taxa of scaled Synurophyceae and Chrysophyceae in Lower Lake Mayakka and Lake Tarpon during March of 1986, 1987, 1989 and 1992. “A” = abundant; “C” = common; “R” = rare; “VR” = very rare. Taxon Mayakka Tarpon 86 87 89 92 86 87 89 92 Mallomonas . akrokomos Ruttner . annulata (Bradley) Harris . bronchartiana Compere . caudata Krieger corymbosa Asmund & Hilliard . cyathellata v. cyathellata Wujek & Asmund heterospina Lund mangofera f. foveata Diirrschmidt R matvienkoae v. matvienkoae (Matvienko) Asmund & Kristiansen R R . matvienkoae v. myakkana Siver C C C R R R . papillosa Harris & Bradley R . peronoides (Harris) Momeu & Péterfi’ VR . portae-ferreae v. reticulata Gretz, Sommerfeld & Wujek A pseudocoronata Prescott R striata v. serrata Harris & Bradley R . tonsurata Teiling emend. Krieger R . transsylvanica Péterfi & Momeu VR VR >> DD Da DD Den a0] ma a rr DB SS oir oS ISS ES USES SESS SS Eo}! sh el tes Synura australiensis Playfair R R curtispina (Petersen & Hansen) Asmund VR S. petersenii f. petersenii Korshikov S. petersenii f. truttae Siver C S. uvella Stein ron @) > wa @) Spiniferomonas S. alata Takahashi VR S. bourrellyi Takahashi R S. coronacircumspina (Wujek & Kristiansen) Nicholls R R R R S. trioralis Takahashi R R 112 FLORIDA SCIENTIST [VOL 56 TABLE 1. CONTINUED Taxon Mayakka Tarpon Paraphysomonas P. takahashi Cronberg & Kristiansen R R P. vestita (Stokes) de Saedeleer C A R R R Cc Chrysosphaerella C. brevispina Korschikov em. Harris & Bradley VR * Note that Diirrschmidt and Cronberg (1989) recently combined M. bangladeshica with M. peronoides. pores in the proximal region of the scale; all other features of the scales are similar to the type (Diirrschmidt and Cronberg, 1989). Since Diirrschmidt and Cronberg (1989) did not state whether scales of v. grandis possessed the dense covering of papillae as do those of v. myakkana we have decided at this point to recognize both taxa. In addition, scales of v. myakkana (6.5 um to 7.5 um in length) are slightly larger than those of the type, but smaller than those listed for v. grandis (8 um to 9 um; see Diirrschmidt and Cronberg, 1989). Further work will be needed in order to determine if indeed both taxa should remain separate or whether a combination is in order. Regardless, it is important to note that both taxa have only been reported from warm tropical or subtropical localities (Diirrschmidt and Cronberg, 1989; Cronberg, 1989; Saha and Wujek, 1990; Siver, 1991). Siver (1991) discussed the difficulties with the identification of taxa in the genus Mallomonas contained within the Series Tonsuratae, especially M. tonsurata and M. corymbosa. Several of the cells found in the study lakes proved to be difficult to identify because they had characteristics of both M. tonsurata and M. corymbosa. Because clear differences between the two species are not always present, we believe that further work is warranted in determining the status of these two taxa. Synura petersenii f. petersenii is often reported to be the most common taxon of Synura (e.g. Wee et al., 1976; Wujek et al., 1981; Siver, 1987; Nicholls and Gerrath, 1985). Kristiansen (1986) stated that S. petersenii may be the most widespread and common taxon of scaled chrysophyte in the world. The form found to be most common in both of the study lakes, Synura petersenii f. truttae, was only recently described from a shallow humic stained locality in Connecticut (Siver, 1987). To date, it has been found in only a few localities (Siver, 1987). Thus, it is interesting that this form of S. petersenii was so common in both Lower Lake Myakka and Lake Tarpon. It is also No. 2, 1993] SIVER AND WUJEK—SCALED CHRYSOPHYCEAE 113 CEE Wiyy YHrn“s) ~ \ CDR AG WGK Ss = or \ et Wo KS AA AGE . \ =—_ oS i 1 4 F L O R ID A S CIEN T 1S T [V O L 5 6 No. 2, 1993] SIVER AND WUJEK—SCALED CHRYSOPHYCEAE 115 Figure Legends Fic. 1-8. FIG. 1. Mallomonas caudata. Scales and bristles. 1,960X. FIG. 2-3. Synura petersenii f. truttae. Note that some scales have toothed spines, some pointed spines and others rounded central ridges. Also note the network of struts and ribs forming a series of pores. 10,200X and 14,600X, respectively. FIG. 4- 5. Mallomonas matvienkoae f. myakkana. Fig. 4. Interior and exterior views of body scales. Note the series of papillae covering the anterior 2/3 of the scale. 6,000X. FIG. 5. Morphology of bristle tip. 7,980X. FIG. 6-7. Mallomonas portae-ferreae. Features of scales and bristles. Note the scales in Fig. 6 that appear to lack the secondary rib pattern. 3,740X and 4,440X, respectively. FIG. 8. Mallomonas corymbosa. Whole cell. Specimen closely related to M. tonsurata, but the bristle pattern is that of M. corymbosa (see text). 1,500X. FIG. 9-16. FIG. 9. Mallomonas corymbosa. Note domed and domeless scales. 4,800X. FIG. 10. Paraphysomonas vestita. Remains of a whole cell. 3,930X. FIG. 11. Mallomonas mangofera f. foveata. Note anterior collar and body scales. 5,670X. FIG. 12. Mallomonas tonsurata. Portion of a whole cell showing domed and domeless scales. 5,900X. FIG. 13. Single scale of Mallomonas striata v. serrata. 8,745X. FIG. 14. Mallomonas pseudocoronata. Note winged body scales. 3,050X. FIG. 15. Synura australiensis. Portion of a cell showing long and slender scales. 2,370X. FIG. 16. Mallomonas akrokomos. Isolated apical scale. 9,515X. Fic. 17-21. FIG. 17. Isolated scale of Mallomonas annulata. 11,160X. FIG. 18. Isolated scale of Mallomonas peronoides (larger scale). 8,200X. FIG. 19. Mallomonas bronchartiana. Body scales. 3,040X. FIG. 20. Mallomonas cyathellata v. cyathellata. Domed and domeless scales. 5,780X. FIG. 21. Mallomonas papillosa. Body scales. 6,650X. 116 FLORIDA SCIENTIST [VOL 56 of interest that both of the Florida localities are humic stained, similar to the waterbody where f. truttae was first described. Preliminary data from other localities in Florida indicate that it is quite widespread and much more common than in more northern climates (Siver, unpub. data). Many of the taxa found in both study lakes are relatively rare in more northern climates, but are common in tropical waters. In a survey of work done in tropical localities, Cronberg (1989) concluded that many of the species of scaled Chrysophyceae and Synurophyceae can be grouped into three categories. First, a group of organisms, including M. matvienkoae v. grandis, taxa in the Peronoides group, M. bronchartiana and M. mangofera f. reticulata Cronberg, appear to be primarily restricted to tropical localities. The second group, including M. portae- ferreae, M. guttata Wujek and Synura australiensis, had a wider distribution, but were primarily found in tropical regions. Third, a number of the more common taxa found in the tropics, including Mallomonas tonsurata, M. crassisquama, M. striata v. striata, Paraphysomonas vestita, Synura petersenii f. petersenii and S. echinulata, are cosmopolitan in nature. Many of the taxa listed in these three groups were found in the study lakes (Table 1). Mallomonas cyathellata v. cyathellata, found in both study lakes, is also commonly found in tropical waterbodies (Cronberg, 1989; Saha and Wujek, 1990), and more temperate areas during warm periods (Wujek and Asmund, 1979). Mallomonas striata v. serrata (Cronberg, 1989) and M. annulata (Diirrschmidt and Croome, 1985; Diirrschmidt and Cronberg, 1989) have also been reported as common taxa in tropical waters. We conclude that the flora of Lower Lake Myakka and Lake Tarpon has a definite tropical element. There were a number of rare or very rare species, including Mallomonas heterospina, M.transsylvanica, and Chrysosphaerella brevispina, found in the study lakes that are often associated with more northern climates and coldwater condi- tions. Each of these taxa are consistently found under low temperatures, often less than 10°C, and have also been reported from either eutrophic and/or humic localities (Siver, 1991; 1993b). Thus, it appears that there is also a temperate element of the floras from the study lakes; we believe that cells of these taxa were residual from the winter. Since all of the samples in this study were taken during March it makes sense that both winter and summer organisms would be found. More work to include differences between seasons is warranted. In addition, more work needs to be done in order to determine the distributions of species along environmental gradients in Florida waterbodies. ACKNOWLEDGMENTS — Special thanks to Norton A. Siver and Scott W. Siver for help in field sampling. We also wish to thank Bill Quinnell and Jim Romanow for photographic and EM assistance, respectively. LITERATURE CITED CRONBERG, G. 1989. Scaled chrysophytes from the tropics. Beih. Nova Hedwigia 95: 191-232. DiéRRSCHMIDT, M. AND R. CROOME. 1985. Mallomonadaceae (Chrysophyceae from Malaysia and Australia. Nord. J. Bot. 5: 285-298. AND G. CRONBERG. 1989. Contribution to the knowledge of tropical chrysophytes: Mallomonadaceae and Paraphysomonadaceae from Sri Lanka. Arch. Hydrobiol. Suppl. 82: 15-37. No. 2, 1993] SIVER AND WUJEK—SCALED CHRYSOPHYCEAE 147 KRISTIANSEN, J. 1986. Silica-scale bearing chrysophytes as environmental indicators. Br. Phycol. J. 21: 495-436. . NICHOLLS, K.H. AND J.F. GERRATH. 1985. The taxonomy of Synura (Chrysophyceae) in Ontario with special reference to taste and odour in water supplies. Can. J. Bot. 63: 1482-1493. SAHA, L.C. AND D.E. WUJEK. 1990. Scale-bearing chrysophytes from tropical Northeast India. Nord. J. Bot. 10: 343-354. SIVER, P.A. 1987. The distribution and variation of Synura species (Chrysophyceae) in Connecticut, U.S.A. Nord. J. Bot. 7: 107-116. . 1991. The Biology of Mallomonas: Morphology, Taxonomy and Ecology. Kluwer Academic Publishers, Dordrecht. 230 pp. . 1993a. The distribution of chrysophytes along environmental gradients: their use as biological indicators. Proc. Third International Chrysophyte Symposium (In press). . 1993b. Morphological and ecological characteristics of Chrysosphaerella longispina and C. brevispina (Chrysophyceae). Nord. J. Bot.: In press. SMOL, J.P. 1993. Application of chrysophytes to problems in paleoecology. Proc. Third International Chrysophyte Symposium (In press). WEE, J.L., J.D. DODD AND D.E. WUJEK. 1976. Studies on silica-scaled chrysophytes from Iowa. Proc. Iowa Acad. Sci. 83: 94-97. WUJEK, D.E. 1983. A new fresh-water species of Paraphysomonas (Chrysophyceae: Mallomonadaceae). Trans. Amer. Microsc. Soc. 102: 165-168. . 1984. Chrysophyceae (Mallomonadaceae) from Florida. Florida. Scient. 47: 161-170. AND B.C. ASMUND. 1979. Mallomonas cyathellata sp. nov. and Mallomonas cyathellata var. kenyana var. nov. (Chrysophyceae) studied by means of scanning and transmission electron microscopy. Phycologia 18: 115-119. , M.M. WEIS AND R.A. ANDERSEN. 1981. Scaled Chrysophyceae from Lake Itasca region. II. Synura, Chrysosphaerella, Spiniferomonas. J. Minn. Acad. Sci. 3: 5-7. AND W.E. GARDINER. 1985. Chrysophyceae (Mallomonadaceae) from Florida. II. New species of Paraphysomonas and the prymnesiophyte Chrysochromulina. Florida Scient. 48: 59-63. AND R.G. BLAND. 1988. Spiniferomonas and Mallomonas: descriptions of two new taxa of Chrysophyceae. Trans. Amer. Microsc. Soc. 107: 301-304. AND R.G. BLAND. 1990. Chrysophyceae (Mallomonadaceae and Paraphysomonadaceae) from Florida. III. Additions to the flora. Florida Scient. 54: 41-48. Florida Scient. 56(2):109-117.1993. Accepted: January 8, 1993 118 FLORIDA SCIENTIST [VOL 56 Biological Sciences PREDATION ON ARTIFICIAL GROUND NESTS IN SOUTHWEST FLORIDA KIMBERLY J. BABBITT' AND JEFFREY L. LINCER? Sarasota County, Ecological Monitoring Division, 1301 Cattlemen Road., Sarasota, FL 34232 ABSTRACT: We examined predation on artificial ground nests in southwestern Florida from May to July 1989. No predation was found on artificial ground nests in pine flatwoods (N=180) during three, 8- day trials. In contrast, 100% predation occurred on nests (N=30) in wetland and hammock habitat during one, 8-day trial. Predation was attributed to mammalian predators, most likely feral hogs and raccoons. Although feral hogs are thought to be important predators of eggs in ground nests, the results of this study indicate that feral hogs were not important predators on ground nests in the pine flatwoods habitat where most ground-nesting occurs. Our limited observations indicate that predation levels by feral hogs, and other mammals could have greater impacts in wetland and hammock habitats, and further study in these habitats is warranted. NEST predation is an important source of nesting mortality for many bird species (Ricklefs, 1969). Depredation on ground nests is commonly higher than on nests above the ground (Ricklefs, 1969; Loiselle and Hoppes, 1983; Wilcove, 1985), and several studies have found mammals to be significant ground-nest predators (e.¢., Henry, 1969; Boag et al., 1984; Yahner and Morrell, 1991). It has been speculated that feral hogs (Sus scrofa) may be predators on the eggs of ground-nesting birds (Belden and Frankenberger, 1977; Wood and Lynn, 1977). The feral hog is an important game species in Florida; however, rooting by hogs has been shown to disrupt both plant (Bratton, 1975) and animal communities (Singer et al., 1984). Matschke (1965) and Henry (1969) used artificial nests to examine the potential impacts of feral hogs on nesting of the wild turkey (Meleagris gallopavo) and ruffed grouse (Bonasa umbellus) in the Appalachian Mountains. Matschke concluded that hogs were important predators, while Henry found that they were minor predators. No evidence of feral hog depredation on actual wild turkey nests was found during studies at Lykes Fisheating Creek Wildlife Management Area and Refuge in Glades Co. (Williams and Austin, 1988). To the authors’ knowledge, no studies have specifically addressed hog predation on ground nests in Florida. Such information could have important management implications for feral hogs and the species they may influence. In this paper, we report on hog predation on artificial ground nests in pine flatwoods, wetland, and hammock habitats in southwestern Florida. Emphasis was on pine flatwoods which provide nesting habitat for several ground-nesting birds in Florida, including 2 important game species, the wild turkey and the northern ‘Present address: Department of Wildlife and Range Sciences, University of Florida, Gainesville, FL 32611 *Present address: BioSystems Analysis, Inc., 13220 Evening Creek Drive, Suite 119, San Diego, CA. 92128 No. 2, 1993] BABBIT AND LINCER—PREDATION ON ARTIFICIAL GROUND NESTS 119 bobwhite (Colinus virginianus) (Hirth and Marion, 1979; Williams and Austin, 1988). METHODS—This study was conducted from May to July 1989 on the T. Mabry Carlton, Jr. Memorial (formerly the Ringling-MacArthur) Reserve, a 10,117-ha area of county-owned land in Sarasota County, Florida. The Reserve is bordered on the north by Myakka River State Park, on the west by the Myakka River, and on the south and east by private ranchland. The most abundant habitat type on the Reserve is pine flatwoods (42%). Over 25% of the Reserve is isolated wetlands and slough systems. The remaining habitat is mesic and hydric hammock, dry prairie, and semi-improved pasture (Biological Research Associates, 1986). Pine flatwoods were characterized by an overstory of slash pine (Pinus elliottii) and an understory dominated by saw-palmetto (Serenoa repens) interspersed with wax myrtle (Myrica cerifera) and gallberry (Ilex glabra). Ground cover in open areas consisted of wire grass (Aristida stricta) and scattered forbs. Hammocks were dominated by an overstory of cabbage palm (Sabal palmetto) and live oak (Quercus virginiana), with an open understory and sparse ground cover. Wetlands were characterized by outer shallow areas with grassy species, particularly maidencane (Panicum hemitomon) with deeper centers dominated by pickerelweed (Pontederia cordata). Location of study areas in flatwoods habitat was based on two criteria: (1) burned within the past five years to control for differences in cover density and height of saw-palmetto and (2) areas large enough to allow continuous transects uninterrupted by wet prairies. Two locations, separated by approximately three km, were selected, and five, 300-m transects were established in each area. Transects were > 100 m apart and each was divided into six, 50-m intervals. Artificial nest site location was determined using a stratified random procedure (George, 1987). Three random numbers were chosen to establish: (1) the distance along each 50-m interval, (2) the direction from the transect (right or left), and (3) the distance from the transect (0-20 m). A single nest was located within each 50 m interval (i.e., six nests per transect), and new random nest locations were chosen for each trial. Nests were placed at the base of the saw- palmetto plant nearest the selected point. The artificial nest consisted of two chicken eggs placed in a shallow depression covered with vegetation litter. As Matschke (1965) found that predation on artificial ground nests was skewed to later dates (after three weeks), suggesting that predators were attracted to decaying eggs, we chose an 8-day trial period to avoid possible bias due to egg spoilage. We conducted three trials separated by at least seven days. The first trial began on 25 May and occurred during a drought period when many of the wetlands adjacent to the study areas were dry. Because use of an area may be related to moisture conditions, we did not begin the second trial until 13 June when there was standing water in adjacent wetlands. The third trial was begun on 29 June. A total of 180 artificial nests was involved in the three trials. Although we were most interested in predation levels in pine flatwoods habitat, low predation rates in the flatwoods and observations of higher hog use of nearby wetland and hammock habitats during the first two flatwoods trials prompted us to conduct a single eight-day trial in these habitats. On each of 10 transects, extending from the hammock into the wetland, we placed one nest in the middle of the hammock, one at the hammock/wetland edge and one an equal distance into the wetland (i.e., if the nest at the middle of the hammock was 15 m from the edge then the nest in the wetland was 15 m from the edge). Transects were shorter than those in the flatwoods, ranging from 47 to 70 m, as their length was dictated by hammock width. Transects were spaced 40 to 80 m apart. The nests were similar to those in the flatwoods and were placed at the base of cabbage palms in the hammock, at the base of cabbage palms or wax myrtles at the hammock/wetland edges and at the base of maidencane clumps in the wetland. The wetland/hammock trial began on 25 June. During the study we conducted diurnal surveys (N =8) for ground-nesting bird species and potential predators in the flatwoods study sites. We also noted any hog rooting activity on the study sites. In addition, we conducted three nocturnal surveys for hogs and other potential nest predators in all three habitats. Artificial nest locations were examined for any remaining eggs or egg fragments and for presence of tracks. RESULTS—No predation on the 180 artificial ground nests was found in the pine flatwoods study areas during any trial. In contrast, all 30 nests in wetland and hammock habitats were predated. Only very small shell fragments remained at or near destroyed nest sites, suggesting mammalian rather than avian predators (Davis, 1959). No tracks were found at the nest sites; however, numerous hog tracks and a 120 FLORIDA SCIENTIST [VOL 56 single raccoon (Procyon lotor) track were found in the wetland/hammock study areas. In addition, hogs (N=7) and raccoons (N=2) were observed in the wetlands/ hammock study areas during nocturnal surveys. Trails were evident throughout the pine flatwoods, and during nocturnal surveys one female hog was observed crossing a road and entering into one of the study areas. No evidence of rooting or any other signs of hog activity (e.g., tracks, scat) were found in the pine flatwoods study areas. Potential mammalian predators known to occur on the Reserve but not observed during surveys include river otter (Lutra canadensis), nine-banded armadillo (Dasypus novemcinctus), striped skunk (Mephitis mephitis), Virginia opposum (Didelphis virginiana), and the cotton rat (Sigmodon hispidus). The common yellowthroat (Geothlypis trichas) and rufous-sided towhee (Pipilo erythrophthalmus) were common ground/near-ground nesters on the flatwoods study sites. In addition, a fledgling common nighthawk (Chordeiles minor) was found on one of the transects. Northern bobwhites and wild turkeys nest in the flatwoods on the Reserve but were not observed within the study areas, although the former were observed nearby. DIscuss1ON—Although results obtained from artificial nests do not necessarily reflect predation levels on real nests, the magnitude of the difference between predation in the pine flatwoods and the hammock and wetland habitats indicates that the potential for predation on actual nests may differ among these habitats. Differential habitat use by nest predators may explain the lack of predation we observed in pine flatwoods. Hogs are known to exhibit seasonal shifts in habitat use in relation to foraging activities and thermoregulatory needs such as cover and water (Baber and Coblentz, 1986). Belden and coworkers (1985) found that spring and summer forage of hogs at Fisheating Creek Wildlife Refuge consisted mainly of grasses and hydric plants, particularly pickerelweed and pennywort (Centella asiatica). Our observations during nocturnal surveys indicate that hogs were foraging mainly in wetland and hammock habitats during this study. Thus high predation levels in the moist habitat, and lack of predation in the dryer habitat, could be related to a high concentration of predators in these areas. Predation levels might be different later in the summer. Although hogs travelled through flatwoods, we did not find any evidence of rooting activity in this habitat during the study. Because the diet of the hog consists primarily of vegetation (Belden et al., 1985; Wood and Roark, 1980), predation on ground nests is probably opportunistic. A high level of predation was found in the hammock and wetland areas where hogs were apparently concentrating most of their foraging activities. Raccoons also forage in hydric and mesic habitats, and have been identified as significant predators on Florida sandhill crane (Grus canadensis pratensis) nesis (Bennett and Bennett, 1990). An alternative explanation for differences in predation levels among habitats is that differences in habitat structure led to differences in nest visibility. Although we covered nests in hammocks and wetlands with litter as in flatwoods, the vegetation cover above the nest probably was less than in flatwoods. This was particularly true No. 2, 1993] BABBIT AND LINCER—PREDATION ON ARTIFICIAL GROUND NESTS 121 for nests at the base of cabbage palms. The absence of predation in pine flatwoods is not easily explained. American crows were observed in the flatwoods study areas, and are important avian nest predators in other areas (Yahner and Scott, 1988; Yahner et al., 1989; Sullivan and Dinsmore, 1990). However, the highest levels of predation on artificial nests found in these studies were in above-ground nests or ground nests that were left uncovered. Lack of predation by corvids in pine flatwoods may be related to the high level of cover provided by saw-palmetto. Because they are visual predators, crows may not detect nests in areas of high cover without observing adult activity. High cover may also reduce mammalian predation on ground nests (Boag et al., 1984). In addition, because we used chicken eggs, which are large compared to the eggs of most species nesting in the area, we may have excluded some smaller snakes from the potential ground nest predator guild. As few ground-nesting bird species nest in hammock or wetland habitats, and no predation by hogs was found in the flatwoods habitat, management efforts to control hog predation on ground nesting birds may not be necessary unless a specific problem is identified. ACKNOWLEDGMENTS—We thank C. May and B. Perry for valuable field assistance, and C. Champe, B. Frankenberger, D. Maehr and an anonymous reviewer for helpful review comments. LITERATURE CITED BABER, D. W., AND B. E. COBLENTZ. 1986. Density, home range, habitat use, and reproduction in feral pigs on Santa Catalina Island. J. Mamm. 67:512-525. BELDEN, R.C., W. B. FRANKENBERGER, and D. H. AUSTIN. 1985. A simulated harvest study of feral hogs in Florida. P-R Proj. W-41-R, Study XIII, Fl. Game and Fresh Water Fish Comm. Tallahassee, FL. 83 pp. BELDEN, R.C., me D W. B. FRANKENBERGER. 1977. Management of feral hogs in Florida - past, present and future. Pp. 5-10. In: WOOD, G. W. (ed.). Research and management of wild hog populations. Belle W. Baruch Forest Science Institute of Clemson Univ., Georgetown, SC. BENNETT, A. J. AND L. A. BENNETT. 1990. Productivity of Florida sandhill cranes in the Okefenokee Swamp, Georgia. J. Field Ornith. 61:224-231. BIOLOGICAL RESEARCH ASSOCIATES. 1986. Technical Memorandum No. 10 Description of the floral communities of the Ringling-Macarthur Reserve. Sarasota, FL. 32 pp. BoaG, D. A., S. G. REEBS, and M. A. SCHROEDER. 1984. Egg loss among spruce grouse inhabiting lodgepole pine forests. Can. J. Zool. 62:1034-1037. BRATTON, S. P. 1975. The effect of the European wild boar, Sus scrofa, on gray beech forest in the Great Smoky Mountains. Ecology 56:1356-1366. Davis, J. R. 1959. A preliminary progress report on nest predation as a limiting factor in wild turkey populations. Pp. 138-145. Proceedings of the first national wild turkey management symposium. Memphis, TN. GEORGE, T. L. 1987. Greater land bird densities on island vs. mainland: Relation to nest predation level. Ecology 68:1393-1400. HENRY, V. G. 1969. Predation on dummy nests of ground-nesting birds in the southern Appalachians. J. Wildl. Manage. 33:169-172. HIRTH, D. H. AND W. R. MARION. 1979. Bird communities of a south Florida flatwoods. Florida Sci. 493-142-151. LOISELLE, B. A. AND W. G. HOPPES. 1983. Nest predation in insular and mainland lowland rainforest in Panama. Condor 85:93-95. 122 FLORIDA SCIENTIST [VOL 56 MATSCHKE, G. H. 1965. Predation by European wild hogs on dummy nests of ground-dwelling birds. Proc. Annu. Conf. Southeast. Assoc. Game and Fish Comm. 19:154-156. RICKLEFS, R. E. 1969. An analysis of nesting mortality in birds. Smithson. Contrib. Zool. 9:1-48. SINGER, F. J., W. T. SWANK, AND E. E. C. CLEBSCH. 1984. Effects of wild pig rooting in a deciduous forest. J. Wildl. Manage. 48:464-473. SULLIVAN, B. D., AND J. J. DINSMORE. 1990. Factors affecting egg predation by American crows. J. Wildl. Manage. 54:433-437. WILCOVE, D. S. 1985. Nest predation in forest tracts and the decline of migratory songbirds. Ecology 66:1211-1214. WILLIAMS, L. E. AND D. H. AUSTIN. 1988. Studies of the wild turkey in Florida. Tech. Bull No. 10. Fla. Game and Freshwater Fish Comm. Univ. of Florida Press., Gainesville, FL. 232 pp. WOOD, G. W., AND T. E. LYNN, Jr. 1977. Wild hogs in southern forests. South. J. Appl. For. 1:12-17. AND D.N. ROARK. 1980. Food habits of feral hogs in coastal South Carolina. J. Wildl. Manage. 44:506-511. YAHNER, R. H. AND D. P. SCOTT. 1988. Effects of forest fragmentation on depredation of artificial nests. J. Wildl. Manage. 52:158-161. ,T. E. MORRELL, AND J. S. RACHAEL. 1989. Effects of edge contrast on depredation of artificial avian nests. J. Wildl. Manage. 53:1135-1138. AND T. E. MORRELL. 1991. Depredation of artificial avian nests in irrigated forests. Wilson Bull. 103:113-117. Florida Scient. 56(2):118-122.1993. Accepted: March 8, 1993 REVIEW Edward A. Fernald and Elizabeth D. Purdum (eds), James R. Anderson and Peter A. Krafft, Jr. (cartographers). Atlas of Florida, University Press of Florida, Gainesville, 1992. Pp. 288, 9 x 12. Price: $39.95. MOST Floridians are immigrants to Florida. When the population increased from about 6.8 million (1970) to 11.7 million (1986), between 89 and 92% of the increase was due to met migration (UF Economic Leaflet 46(6) 1-4.). This book is for us, as well as for those who wish an outstanding overview of Florida life and history in beautiful visual form. This is the superb, informative volume that you would expect from the State Geographer (Dr. Fernald) and an anthropologist and experienced editor (Dr. Purdum), who collaborated with 17 other contributors. Between them, they covered Natural Environment, History and Culture, Population, Economy, Recreation and Tourism, Infrastructure and Planning. Pictures, impressive charts, beautiful maps, references, a thorough index, and tight writing — there is something here for all Floridians or those interested in Florida. — Dean F. Martin, University of South Florida, Tampa. No. 2, 1993] PALMER ET AL.—XPS SPECTRA 123 Chemical Sciences XPS SPECTRA OF [PT(DIEN)X]Y COMPLEXES Jay W. PALMER, WILLIAM E.. SWARTZ, Jr. DavIpD KING? AND JOSEPH A. STANKO Department of Chemistry, University of South Florida, Tampa, Florida 33620 ABSTRACT: X-ray photoelectron binding energies of inner shell electrons of platinum and ligand nitrogen atoms were obtained in the series of platinum(II) diethylenetriamine (dien) complexes, [Pt(dien)X]Y, where X = CN, NO,, I, Cl, and NO, and where selected counter ions, Y = I, Cl, or NO,. The ordering deduced for the coordinated anions X in [Pt(dien)X]Y complexes was compared with the ordering for the same ligands in the spectrochemical series, the nephelauxetic series and the trans-effect of coordination complexes. It was found that the XPS Pt(4f°”) and Pt(4f””) values for these anions correlated quite well with the ordering of the spectrochemical series, while the XPS Pt(4d*”) values correlated very well with the ordering in the nephelauxetic series. The XPS(dien)N(1s) values fit also quite well with the ordering of the trans-effect; however, no evidence was seen with the ordering of the secondary trans-effect. IN recent years, an increasing number of people have developed cancer (Davis et al., 1990). Much of this probably is the result of an older population, but also may be the product of a more polluted environment. This is especially true in the state of Florida, which not only has a large population of older people and increased pollution, but also has many sun worshipers who have increased risks in acquiring skin cancer. Because of the greater number of cancer patients, there has been an increasing interest in the aqueous solution chemistry of the diethylenetriamine complexes of platinum(II), [Pt(dien)X]Y, where dien equals NH,-CH,-CH,NH-CH,CH,-NH,,. Much of this work has been done in attempting to explain the mechanism of action of Cisplatin, [cis-Pt(NH,),Cl,], an antitumor agent commonly used in Florida as well as the rest of the world. This mechanism of action has been found to be based on the interaction of the Pt(NH,),”* moiety with the nucleobases of DNA (Meischen et al., 1976; Rosenberg, 1979; Fichtinger-Schepman et al., 1985; Banard, 1989). The main product of these interactions seems to be bifunctional interstrand cross-linking between two neighboring guanine bases through the N7 atoms (Miller and Marzilli, 1985, Sherman and Lippard, 1987; Van der Veer et al., 1987). A limitation of Cisplatin and its platinum(II) ethylenediamine (en) complex, [Pt(en)Cl,], analog is their concentration-dependent nephrotoxicity (Krakoff, 1979; Dentino et al., 1978). Reagents such as sodium diethyldithiocarbamate (Na(ddtc)) or thiourea can be used to reduce this nephrotoxicity (Von Hoff et al., 1979; Lempers and Reedijk, 1978); therefore, it becomes important to understand their mechanism of action. Studies of the reactions of a monodentate complex, [Pt(dien)CI]Cl, with 9-ethylguanine and glutathione have been reported (Van der Veer et al., 1987; ‘Martin Marietta Specialty Components, P.O. Box 2908, Largo, Florida 34649 *Solar Energy Research Institute, Branch 213 16-3, Golden, CO 80401 124 FLORIDA SCIENTIST [VOL 56 Lempers et al., 1988). These results then were compared with those of Cisplatin (Lempers and Reedijk, 1990). Other aqueous chemistry of the diethylenetriamine complexes of platinum for the most part has also dealt in part with the interest in the use of Cisplatin in cancer chemotherapy. Equilibrium and kinetic studies involved dimerization of [Pt(dien)(H,O)]?* (Erickson et al., 1987; Beattie, 1983); spontaneous solvolysis reactions (Kotowski and Van Eldik, 1984); rates and mechanism of substitution reactions of dimethyl sulfoxide and [ Pt(dien)H,O]* (Romeo and Cusumano, 1981); kinetics of anation (Kotowsk et al., 1980), and photosubstitution reactions (Bartocci et al. 1969). In spite of the fact that these complexes have been studied extensively for many years, confusion on the mechanism of substitution reactions is still apparent. Earlier work on anation of [Pt(dien)(OH,)* had suggested that a 3-coordinate intermediate was involved and the replacement of water in the complex was a function of the entering ligand which decreased in the order OH >> F ~ SCN > Br > Cl > NO, (Gray and Olcott, 1962). More recent work (Beattie, 1983), however, suggests that an equilibrium is involved. Still more recent studies (Erickson et al., 1987) indicate that a hydroxo bridged dimerization reaction is a part of the process. An earlier study (Basolo et al., 1960) had suggested that the relative reactivities of replaceable ligands in [Pt(dien)X]Y complexes decreased in the following order NO, > Cl > Br >I >N, > SCN > NO, > CN. This order parallels both the order of increasing bond strength and the trans effect. In a related study, Palmer and Basolo (1960) measured the base-catalyzed hydrogen exchange of the amine hydrogens in a series of [Pt(dien)X]Y complexes. This was done by following the IR spectrum of the complexes dissolved in acetate-buffered, D,O solutions in the N-H overtone region. The rate of hydrogen exchange as a function of coordinated X anion, was found to decrease in the order: SCN’ > I > NO, >> Br > Cl. Furthermore, in that study the hydrogen exchange rate showed considerable curvature indicating non-equivalence of hydrogens in the complex. The rapid exchange corresponded to approximately 20% (1 out of 5) of the N-H hydrogens, or probably to the hydrogen on the ligand nitrogen trans to the X group. This “secondary” trans effect was independently confirmed by Watt and Cude (1968b) in a series of experiments involving deprotonation and subsequent methylation of the [ Pt(dien)X] ions in liquid ammonia. In an attempt to shed more light on this dilemma, we examined the XPS spectra of this series of complexes. XPS spectra, even though core energy levels are being probed, would be expected to show the influence of electron charge density shifts between the metal and the ligand and thus provide another measure of bonding interactions (Jolley, 1977). It is of interest to determine if the order of the platinum core-binding energies can be correlated with the different ligand relationships observed in platinum(II) complexes such as the spectrochemical series, the nephelauxetic series, and the trans-effect series. MATERIALS AND METHODS—Synthesis—|[Pt(dien)X]Y complexes were prepared according to methods described in the literature, (vida infra). The platinum(II) starting materials were the ammonium No. 2, 1993] PALMER ET AL.—XPS SPECTRA 125 salt of the tetrachloro-complex, (NH,),PtCl,, which was used for one set, and the dichloride salt PtCl, for asecond set. Both were obtained from Aldrich Chemical Co. The diethylenetriamine was manufactured by Fluka Chemic AG. Because of the difficulty in obtaining some of the more soluble complexes, two methods of preparation were used. The first followed the procedure originally developed by Mann (Mann, 1934; Basolo et al., 1960; Watt and Cude, 1968a) in which the dien ligand is first protonated and then heated with the tetrachloroplatinate(II) anion, PtCl,” to synthesize [Pt(dien)CI]Cl. In the second type of preparation, modification of the Mann procedure was used (Watt and Cude, 1968a). This synthesis required the prior preparation of platinum(II) iodide, PtI,.H,O, from platinum chloride, which then was treated with the free amine form of the dien ligand to form the product, | Pt(dien)I]I. The two complexes, [Pt(dien)CN]I, [Pt(dien)NO,]I, were synthesized by first preparing the aquo complex, [Pt(dien)H,O](NO,)’, by treating the iodo-iodide complex with the stoichiometric amount of silver nitrate solution to remove and precipitate silver iodide. The insoluble AgI precipitate was removed by filtration. The filtrate was treated with solutions containing stoichiometric amounts of sodium salts of nitrite and cyanide to displace the water ligand. However, since the nitrate salts of the cyano- and nitro-complexes are very soluble, the solutions were further treated with stoichiometric solutions of Nal to produce the less soluble [Pt(dien)CN ]I and [Pt(dien)NO,]I complexes. A portion of the [Pt(dien)H,O](NO,), solution was evaporated to dryness; however, the removal of water ligand from the complex and conversion to the [Pt(dien)NO,]NO, complex was accomplished in the high vacuum of the XPS unit during analysis. No elemental analysis was done on this complex. Analyses of the complexes were done by Desert Analytics, Tucson, Arizona. Calc. for [Pt(dien)CI]Cl: C, 13.01; H, 3.56; N, 11.38. Found: C, 12.73; H, 3.16; N, 10.04. Calc. for [ Pt(dien)I]I: C, 8.90; H, 2.38; N, 7.61. Found: C, 9.54; H, 2.53; N, 8.15. Calc. for [| Pt(dien)CN|I: C, 13.30; H, 2.91; N, 12.42. Found: C, 13.24; H, 2.92; N, 12.07. Calc. for [Pt(dien)NO, JI: C, 10.19; H, 2.79; N, 11.89. Found: C, 9.81; H, 2.61; IN LOA2. X-ray Photoelectron Spectra. The XPS binding energies were recorded ona GCA McPherson ESCA 36 photoelectron spectrometer via excitation with Al(K, ) x-rays (E = 1486.6 eV). The typical base pressure in the sample chamber was 1 x 10” Torr. The platinum complexes were finely ground and spread onto double sided sticky-tape affixed to 1 x 1 cm aluminum mounts. Because the binding energy shifts from compound to compound were small, 100 spectral scans were taken of each sample. In addition, four to five samples of each complex were run and the results averaged. All spectra were calibrated against the carbon Is binding energy taken as 285.0 eV. For platinum, in addition to obtaining binding energy on the 4f core levels, the 4d levels were also obtained which were much broader than the 4f core levels. The nitrogen Is core levels were taken for all types of nitrogen. RESULTS AND DiscussioN—Selected XPS binding energy data for five diethylenetriamine complexes of platinum(II) are given (Table 1). These data represent the Pt(4f°”), Pt(4f””), Pt(4d°”) and (dien) N(1s) binding energies. The peak widths at half-height are shown in parentheses. TABLE 1. Some measured binding energies (eV) of [Pt(dien)X]Y complexes’ xX Y PtPt(4f”) Pt(4f”) Pt(4d°”) (dien)N(1s) NO, NO 76.5(2.0) 73.3(2.1) 316.4(3.8) 400.3(2.5) Cl Cl 76.8(2.1) 73.5(2.2) 316.5(4.4) 400.3(2.4) NO, I 78.1(2.2) 74.8(2.2) 317.1(3.8) 400.1(2.5) CN I 78.0(2.2) 74.8(2.2) 317.9(3.4) 400.9(3.2) I I 78.4(2.3) 74.9(2.3) 318.1(4.0) 400.8(2.8) ‘Values in parentheses are peak widths at half-heights. 126 FLORIDA SCIENTIST [VOL 56 Both the Pt(4f°”) and Pt(4f”) binding energies for the [Pt(dien)X]Y complexes increase in the order NO, < Cl < NO, ~ CN < I. With the exception of I, this follows the so-called spectrochemical or Fajans-Tscuhida series (Schaeffer and Jorgensen, 1958) for octahedral complexes. The Pt(4d*”) binding energies increase in the order NO,, Cl, NO,, CN, l. This is the order of the nephelauxetic series (Schaeffer and Jorgensen, 1958; Figgis, 1966), except that the position for NO, does not appear to be reported in the literature. This ligand ordering observed for the core-binding energies of the Pt(4d” *) electrons may be an indication of the degree of covalent back-bonding of the platinum 5d electrons with ligand orbitals. Because 5d core-binding energies are around 4-6 eV, quite far removed from the standard C(1s) core-binding energy of 285.0 eV, the Pt(4d°”) core-binding energies were measured. This way a broad d electron peak could be measured. It is expected that the 4d electron delocalization is a reflection of the 5d electron delocalization to the ligand. With the exception of NO,,, the XPS Pt(4d°”) widths at one-half height decrease in the order of CI, > T > NO, > CN. This may mean that the platinum 5d electron charge density is being directed into suitable orbitals on the coordinated anion, X, in the order given but to an increasing extent. Inspection of the data for the XPS (dien)N(1s) core-binding energies shows that with the exception of NO, the energies increase in the order of NO, ~ Cl << CN’ ligands. This is the order oe the increasing trans effect (Basolo etal. 1960). The width of the peak at one-half height indicates that those complexes containing CN’ and I are the broadest. However, part of this broadening in CN” may be the result of (CN)N(1s) which occurs at 399.0. Considerable effort was done in trying to curve- fit the peaks to determine if a splitting could be observed for the (dien)N(1s). That is, one peak for the two dienN(1s) cis-nitrogens to the X ligand and one smaller peak for the one trans-nitrogen. No splitting of the peaks was found. Perhaps further effort with curve-fitting computer programs would be helpful here to determine if splitting does occur. If splitting of the (dien)N(1s) peak is present, then this would confirm the secondary trans effect seen in previous studies. There does seem to be a problem with the NO, ligand in studies of (dien)N(1s). It was observed that towards the end of the scans in a run, the (NO, N(1s) peak had decreased considerably or disappeared. Perhaps this is the result of an oxidation- reduction reaction occurring in the presence of the XPS x-rays, i.e., the NO, ion was reduced by the amine hydrogen. Because of the close proximity of these ligands, considerable hydrogen bonding probably takes place. Alternatively, in the presence of x-rays in a high vacuum, hydrogen from the amine ligand and oxygen from the NO, ligand may be removed as water. Previously, it had been observed that XPS x-rays can cause the reduction of complexed platinum IV to platinum II (Gilbert et al., 1982). These data indicate that the XPS core-binding energies for platinum(II) and its ligands may be useful in relating chemical and physical properties of the different ligands in a series of platinum(II) complexes. ACKNOWLEDGMENTS— We gratefully acknowledge the efforts of Julia R. Burdge in attempting to obtain evidence for the secondary trans effect in these complexes. No. 2, 1993] PALMER ET AL.—XPS SPECTRA 127 LITERATURE CITED BANARD, C. F. J. 1989. Platinum anti-cancer agents, Platinum Metals Rev. 33:162-167. BARTOCCI, C., F. SCANDOLA, AND V. BALZANI. 1969. Mechanism of photosubstitution Reactions of square-planar platinum(II) complexes. J. Am. Chem. Soc. 91:(25):6948-6951. BASOLO, F., H. B. GRAY, AND R. G. PEARSON. 1960. Mechanism of substitution reactions of complex ions. J. Am. Chem. Soc. 82:4200-4203. BEATTIE, J. K. 1983. On a dissociative pathway in the anation of aquadiethyleneplatinum(II) ion. Inorg. Chim. Acta. 76:L69. Davis, D.L., D. HOEL, J. FOx, AND A.D. LOPEZ. 1990. International trends in cancer mortality in France, West Germany, Italy, Japan, England and Wales, and the United States. pp. 5-48. In: DAVIS, D.L. AND D. HOEL (eds.), 1990. Trends in Cancer Mortality in Industrial Countries. Ann. N.Y. Acad. Sci. 609. DENTINO, M., F.C. LUFT, M. N. YUM, S. D. WILLIAMS, L. H. EINHORN. 1978. Longterm Effect of cis- diamminedichloride platinum(CDDP) on renal function and structure in man. Cancer. 41:1274- 1281. ERICKSON, L. E., H. L. ERICKSON, and T. Y. MEYER. 1987. Equilibrium and kinetic studies of monoaquo complexes of platinum(II). Inorg. Chem. 26:997-999. FICHTINGER-SCHEPMAN, A. M. J., J. L. VAN DER VEER, J. H. J. DEN HARTOG, P. H. M. LOHMAN, AND J. REEDIJK. 1985. Adducts of the antitumor drug cis-diamminedichloroplatinum(II) with DNA: Formation, identification and quantification. Biochemistry 24:707-713. FIGGIS, B. N. Introduction to Ligand Fields, Interscience Publishers, John Wiley and Sons, New York. 1966. 242-245. GILBERT, R.A., J.A. LLEWELLYN, W.E. SWARTZ, JR., AND J.W. PALMER. 1982. Application of factor analysis to the resolution of overlapping XPS spectra. Applied Spectroscopy. 36:428-430. Gray, H. B. AND R. J. OLCOTT. 1962. Kinetics of the reactions of diethylenetriammeaquoplatinum(II) ion with different ligands. Inorg. Chem. 1:481-485. JOLLEY, W.L. 1977. The application of x-ray photoelectron spectroscopy in inorganic chemistry, Electron Spectroscopy: Theory, Techniques and Applications, 1:119-49. KOTOWSKI, M., D. A. PALMER, and H. KELM. 1980. Kinetics of the anation of aquadiethylenetriamine platinum(II) ions. Inorg. Chim. Acta. 44:L113-L114. AND R. VAN ELDIK. 1984. Spontaneous solvolysis reactions of some Pd(II)-dien complexes in aqueous solution: Equilibrium and kinetic data. Inorg. Chem. 23:3310-3312. KRAKOFF, I. H. 1979. Nephrotoxicity of cis-dichlorodiaminneplatinum(II). Cancer Treat. Rep. 63:1523- 525 LEMPERS, E. L. M., K. INAGAKI AND J. REEDIJK. 1988. Reactions of | PtCl(dien)|Cl with glutathione, oxidized glutathione and 5-methyl glutathione. Formation of an S-bridged dinuclear unit. Inorg. Chim. Acta. 102:201-207. AND J. REEDIJK. 1990. Reversibility of binding of cis-platin-methionine in proteins by diethyldithiocarbonate or thiourea. Inorg. Chem. 29:217-222. MANN, F. G. 1934. The constitution of complex metallic salts. J. Chem. Soc. 1934:466-474. MEISCHEN, S. J., G. R. GALE, L. M. LAKE, C. J. FRANGAKIS, M. G. ROSENBLUM, E. M. WALKER Jr., L. M. ATKINS AND A. B. SMITH. 1976. Antileukemic properties of organoplatinum complexes, J. Natl. Cancer Inst. 57:841-845. MILLER, S. K. AND L. G. MARZILLI. 1985. Interaction of platinum antitumor agents with guanine nucleosides and nucleotides. Inorg. Chem. 24:2421-2425, PALMER, J. W. AND F. BASOLO. 1960. Effect of ligands on rates of hydrogen exchange of substituted metal amines. J. Phys. Chem. 64:778-780. . ROMEO, R.AND M. CUSUMANO. 1981. Rates and mechanism of substituted reactions of dimethylsulfoxide and aquo-(diethylenetriamine)platinum(II) ions. Inorg. Chim. Acta. 49:167-171. ROSENBERG, B. 1979. Anti cancer activity of cis-dichlorodiammineplatinum(II) and some relevant chemistry. Cancer Treatment Reports. 63:1433-1438. SCHAEFFER, C. E. AND C. K. JORGENSEN. 1958. The nephelauxetic series of ligands corresponding to increasing tendency of partly covalent bonding. J. Inorg. Nuclear Chem. 8:143-147. SHERMAN, S. E. ANDS. J. LIPPARD. 1987. Structural aspects of platinum anticancer drug interactions with DNA. Chem. Rev. 87:1153-1181. VAN DER VEER, J. L., H. VAN DER ELST, AND J. REEDIJK. 1987. Separation, characterization, and stability of products from cis-PtCl,(NH,,), and PtCl(dien)Cl with 9-ethylguanine, formed under neutral or 128 FLORIDA SCIENTIST [VOL 56 alkaline conditions. Inorg. Chem. 26:1536-1540. VON HoFF, D. D., R. SCHILSKY, C. M. REICHERT, R. L. REDDICK, M. ROZENCWERG, R. C. YOUNG, AND F. M. MUCCIE. 1979. Toxic effects of cis-dichlorodiammineplatinum(II) in man. Cancer Treat. Rep. 63:1527-1531. WarTT, G. W. AND W. A. CUDE. 1968a. Diethylenetriamine complexes of platinum(II) halides. Inorg. Chem. 7:335-340. AND W. A. CUDE. 1968b. The secondary trans effect in platinum(II) complexes. J. Am. Chem. Soc. 90:(23):6382-6384. Florida Acient. 56(2):123-128.1993. Accepted: March 19, 1993 REVIEW Eugene Garfield, Science Reviews, Journalism, Inventiveness and Other Es- says, ISI Press, Philadelphia, 1992. Pp. xx + 451. Price: $35.00. INVENTOR of Current Contents® and Science Citation Index®, the author is also founder and Chairman of the Institute for Scientific Information®. This volume, like the preceding ones in the series, Essays of an Information Scientist, covers a range of interesting and usefully presented topics. They can be treasured for themselves or they can be read as useful review articles that provide an entree into the subject for the reader who wishes more. The author of Current Comments, a weekly series of essays, switched to a biweekly format with guest authors, but with an introduction by Dr. Garfield. The range is notable, as usual, and includes citation analysis (comparisons, uses, impact), computers, science and poetry, science and politics and the CDC (Center for Disease Control), scientists (with interesting excerpts from David Cassidy’s Uncer- tainty: The Life and Science of Werner Heisenberg), tinnitus (a personal favorite), and electronic publishing (the Florida Scientist is safe). The connectiveness? All are informative, carefully documented, and usefully collected in an attractively printed volume. In addition for collectors like me, over 85 Commentaries from The Scientist® written by the author in 1986-1992 are conveniently included, and I can part with a portion of my clipping collection. This volume, like its predecessors, is recommended to those who enjoy good writing, good editing, and fascinating topics. — Dean F. Martin, University of South Florida, Tampa. | / INSTRUCTIONS TO AUTHORS Individuals who publish in the Florida Scientist must be active members in the Florida Academy of Sciences. Submit a typewritten original and two copies of the text, illustrations, and tables. All typewrit- ten material—including the abstract, literature citations, footnotes, tables, and figure legends— shall be double-spaced. Use one side of 81/2 x 11 inch (21'/2 cm X 28 cm) good quality bond paper for the original; the copy may be xeroxed. Margins should be at least 3 cm all around. Number the pages through the Literature Cited section. Avoid footnotes and do not use mimeo, slick, erasable, or ruled paper. Use metric units for all measurements. Assistance with production costs will be negotiated directly with authors of pages which exceed 10 printed pages. Excess over 10 pages by an author ina single volume will result in charges for additional pages. ADDRESS follows the author’s address. ABSTRACT—AI] manuscripts shall have a short, concise, single-paragraphed abstract. The abstract follows immediately the author’s address. ACKNOWLEDGMENTS are given in the body of the text preceding immediately the Literature Ciied section. LITERATURE CITED section follows the text. Double-space every line and follow the format in the current issue. MANuscriPrTs of 5 or less, double-spaced typewritten pages should conform to the short arti- cle protocol. See recent issue for proper format of both long and short articles. TABLES shall be typed on separate sheets of paper, double-spaced throughout. Each table shall contain a short heading. Do not use vertical rulings. Maximum character width, including spaces, of tables is 98.0. This includes a minimum space of 5 characters between columns. Tables are charged to authors at $45.00 per page or fraction. ILLUSTRATIONS—AIl drawings shall be done in good quality India ink, on good board or draft- ing paper. Letter by using a lettering guide or equivalent. Typewritten letters on illustrations are unacceptable. Drawings and photographs should be large enough to allow 1/3 to 1/2 reduction in size. Photographs shall be glossy prints of good contrast. Whenever possible, mount photographs in lots size. The author’s name and figure number should be penciled lightly on the back of each figure. Figure legends must be listed on a separate page and not on the drawing or photograph. The leg- end must be double-spaced. Illustrations are charged to authors at $25.00 per page or fraction. We ask authors to prepare accepted manuscripts in personal computer formats. Please do not submit a manuscript on disk. Once reviewers have recommended a manuscript be accept- ed, you will be given directions for submitting the final manuscript on disk. PROOF must be returned promptly. Notification of address change and proofreading are the author’s responsibility. Alterations after the type has been set will be charged to the author. REPRINTS may be ordered from the printer on forms provided when the corrected proofs are returned to the Editor. FLORIDA ACADEMY OF SCIENCES INSTITUTIONAL MEMBERS FoR 1993 Archibold Biological Station St. Thomas University, Miami, FL The British Library, Boston Spa, Weatherby, UK U.S. Department of Commerce, National Brevard Community College, Florida Advanced Marine Fisheries (Panama City) Technology Center U.S. Environmental Protection Adminis- Florida Institute of Technology tration Library, Gulf Breeze, FL Harbor Branch Oceanographic Institution, Inc. University of Central Florida, Division of Newfound Harbor Marine Institute Sponsored Research St. John’s River Community College, Palatka University of South Florida St. Petersburg (Florida) Public Library Corporate Members Lockheed Space Operations, Titusville Membership applications, subscriptions, renewals, changes of address, and orders for back numbers should be addressed to the Executive Secretary, Florida Academy of Sciences, P.O. Box 033012, Indialantic, FL 32903-0012 PUBLICATIONS FOR SALE by the Florida Academy of Sciences Complete sets. Broken sets. Individual numbers. Immediate delivery. A few numbers reprinted by photo-offset. All prices strictly net. Special bargain: $500 for nearly complete sets. Prices quoted include domestic postage. PROCEEDINGS OF THE FLORIDA ACADEMY OF SCIENCES (1936-1944) Volumes 1-7—$10.00 per volume; single numbers $3.50 QUARTERLY JOURNAL OF THE FLORIDA ACADEMIC OF SCIENCE (1945-1972) Volumes 8-35—$10.00 per volume; single numbers $3.50 FLORIDA SCIENTIST (1973-75) Volumes 36-38—$10.00 per volume; single issues $3.50 except for symposium numbers priced separately. Volumes 39-47 —$13.00 per volume; single numbers $4.25 except for symposium numbers priced separately. Volumes 48 onward—$20.00/volume; single numbers $5.00 except for symposium numbers priced separately. Florida’s Estuaries—Management or Mismanagement P—Academy Symposium FLORIDA SCIENTIST 37(4)—$5.00 Land Spreading of Secondary Effluent—Academy Symposium FLORIDA SCIENTIST 38(4)—$5.00 Solar Energy—Academy Symposium FLORIDA SCIENTIST 39(3)—$5.00 (includes do-it-yourself instructions) Anthropology—Academy Symposium FLORIDA SCIENTIST 43(3)—$7.50 Shark Biology—Academy Symposium FLORIDA SCIENTIST 45(1)—$8.00 Future of the Indian River System—Academy Symposium FLORIDA SCIENTIST 46(3/4)—$15.00 Second Indian River Research Symposium FLORIDA SCIENTIST 53(3)—$15.00 Individual orders should be sent with payment. A statement will be sent in re- sponse to a bona fide purchase order over $10.00 from a recognized institution. Address all orders to: The Florida Academy of Sciences, Inc. P.O. Box 033012 Indialantic, Florida 32903-0012 Telephone: (407) 723-6835 1993 ISSN: 0098-4590 Scientis Volume 56 Summer, 1993 Number 3 ‘Flori CONTENTS Notes on the Impact of the December 1989 Freeze on Local Populations of Rivulus marmoratus in Florida, with Additional Distribution Records in SME Re TAS ee At ele eee ical sic r busty beau ua ash Rah Daun bales 129 D. Scott Taylor A TTELPS -cenccceceled telat yer hee Ae een Gn ne ag en TEE NE HO ma EU CCL NEI rl Carl A. Luer 134 Vegetative Cover in Florida Based on 1985-1989 Landsat Thematic Mapper Beg eS etcesig hee sare ana cebbases Aly sau aadeshi saotavilyndcaveuantcs 135 Randy S. Kautz, D. Terry Gilbert, and Gregory M. Maulden Tree Planting and Preservation Practices at Single-Family Residences: Policy DENEVE TOUS ea atte a AB SA Soe ae RR eA 155 Lisa B. Beever, Tim Eckert, and Jeffrey S. Magnun Selection of Nest Cavity Volume and Entrance Size by Honey Bees in Florida 163 Roger A. Morse, James N. Layne, P. Kirk Visscher, and Francis Ratnieks Movement of Fluridone in the Upper St. Johns River, Florida ................... 168 Andrew J. Leslie, Don C. Schmitz, R. L. Kipker, and D. L. Giradin Chemical Differences Between Stressed and Unstressed Individuals of Bald- WSS CT AROG INU CISELCIILIN) cnn esi cas ct Al esesdee Loin eed vsdeacsdesdesteistiatadsssss 178 Donna Hall Miller, Sydney T. Bacchus, and Harvey A. Miller Dolomite Extraction from Phosphate Pebble by Aqueous Carbonic Acid- PMMMUMTOSHATUUET SUILALE BULLOL. «.csisct oct oovasecnesogennnssensdesrcovehavacdsactsavcaradveseGoveccess 185 Yixue Pan, Robert F. Benson, and Dean F. Martin QUARTERLY JOURNAL OF THE FLORIDA ACADEMY OF SCIENCES FLORIDA SCIENTIST QUARTERLY JOURNAL OF THE FLORIDA ACADEMY OF SCIENCES Copyright© by the Florida Academy of Sciences, Inc. 1993 Editor Dr. DEAN MARTIN Co-Editor: MRS. BARBARA B. MARTIN Institute for Environment Studies Department of Chemistry University of South Florida 4202 East Fowler Avenue Tampa, Florida 33620-5250 THE FLORIDA SCIENTIST is published quarterly by the Florida Academy of Sciences, Inc., a non-profit scientific and educational association. Membership is open to indi- viduals or institutions interested in supporting science in its broadest sense. Applica- tions may be obtained from the Executive Secretary. Both individual and institutional members receive a subscription to the FLORIDA SCIENTIST. Direct subscription is avail- able at $40.00 per calendar year. Original articles containing new knowledge, or new interpretation of knowledge, are welcomed in any field of Science as represented by the sections of the Academy, viz., Biological Sciences, Conservation, Earth and Planetary Sciences, Medical Sciences, Physical Sciences, Science Teaching, and Social Sciences. Also, contributions will be considered which present new applications of scientific knowledge to practical problems within fields of interest to the Academy. Articles must not duplicate in any substantial way material that is published elsewhere. Contributions are accepted only from members of the Academy and so papers submitted by non-members will be accepted only after the authors join the Academy. Instructions for preparation of manuscripts are inside the back cover. Officers for 1992-93 FLORIDA ACADEMY OF SCIENCES Founded 1936 President: Dr. Patrick J. GLEASON Treasurer: Dr. MARVIN IvEY 1131 North Parkway 14452 Hillview Drive Lake Worth, Florida 33460 Largo, Florida 34644 President-Elect: Dr. Patricia M. Dooris Executive Secretary: Mrs. Betty PREECE Institute for Environmental Studies P.O. Box 033012 Department of Chemistry Indialantic, Florida 32903-0012 University of South Florida Tel: (407)723-6835 Tampa, FL 33620-5250 Program Chair: Dr. DEL DELUMYEA Secretary: Ms. MARCELA GUITERREZ-MAYKA Millar Wilson Laboratory for Chemical Research 701 E. River Dr. Jacksonville University Temple Terrace, FL 33617 Jacksonville, Florida 32211 Published by the Florida Academy of Sciences, Inc. P.O. Box 033012 Indialantic, Florida 32903-0012 Printing by C & D Printing Company, St. Petersburg, FL Florida Scientist QUARTERLY JOURNAL OF THE FLORIDA ACADEMY OF SCIENCES DEAN F. Martin, Editor BARBARA B. MartIN, Co-Editor Volume 56 Summer, 1993 Number 3 Biological Sciences NOTES ON THE IMPACT OF THE DECEMBER 1989 FREEZE ON LOCAL POPULATIONS OF RIVULUS MARMORATUS IN FLORIDA, WITH ADDITIONAL DISTRIBUTION RECORDS IN THE STATE D. SCOTT TAYLOR Brevard Mosquito Control District, 2870 Greenbrooke St., Palm Bay, FL 32905 ABSTRACT: Limited data are available on the effects of the December, 1989 freeze on fishes of Indian River Lagoon, Florida. Air and water temperature data for this event are compared with data from 3 prior freezes in the last 2 decades. While air temperatures were clearly lower in 1989, water temperatures did not reach the extremes of previous events. Collection records for Rivulus marmoratus, a tropical killifish inhabiting high marsh habitats along Indian River Lagoon in Brevard and Indian River Counties, are compared before and after the 1989 freeze. An apparent impact on R. marmoratus populations is documented. Additional collection records are cited on both coasts of the state, adding to the distribution records for this little-known species. THE distribution of tropical fish species at the northern limits of their range is limited by temperature. The east central Florida area falls in the middle of the biogeographic transitional zone from warm-temperate to sub-tropical, and low temperature events can have drastic impacts on aquatic organisms in this area of Indian River Lagoon (IRL). Several severe freezes in the last 2 decades have been documented in a variety of reports: 1977 (Snelson and Bradley, 1978, Gilmore et al., 1978), 1983 and 1985 (Provancha et al., 1986) and 1989 (Snodgrass, 1991). The effects of some low-temperature events (e.g. 1977) on aquatic organisms have been examined in detail. However, the effects of one of the worst freezes of this century (December, 1989) went largely unstudied, except for notes on an exotic cichlid (Snodgrass, 1991), presumably because the event took place over a holiday season. This report (1) compiles available air and water temperature information available on this event for the east central Florida area and compares it with previous events, (2) documents the apparent impacts on local populations of Rivulus marmoratus Poey, a tropical killifish at the northern end of its range in IRL in 130 FLORIDA SCIENTIST [VOL 56 Brevard and Indian River Counties, FL and, (3) adds to the distribution records for R. marmoratus on both coasts of Florida. METHODS—December, 1989 climatological data were obtained from NOAA records (NOAA, 1989). Water temperatures for IRL during the period December 24-31, 1989 were obtained from power plants in Brevard, Indian River and St. Lucie Counties, and for Mosquito Lagoon from field data supplied by the Bionetics Corporation at Kennedy Space Center. Collection records for R. marmoratus at 6 sites in Brevard and Indian River counties for the period 1/5/88 to 12/1/89 were compared with collections from 12/28/89 to 10/25/91. Collections were made in burrows of the great land crab (Cardisoma guanhumi) with hook and line and hand net (Taylor, 1988) or traps (Taylor, 1990). All fish were released after collection. Additions to the reported collection sites of this species throughout both coasts of central and south Florida are also reported. RESULTS AND DISCUSSION—Minimum air temperatures during the period December 24-25, 1989 were -7.2° C (Titusville), -6.1° C (Melbourne), -5.0° C (Vero Beach) and -7.2° C (Ft. Pierce). Temperatures at all stations except Melbourne dropped below 0° C for the next 4 days (NOAA, 1989). The following minimum intake water temperatures for power plants adjacent to the IRL were reported: Titusville: December 26, 1989, 8.2° C (Clameta, 1991); Titusville: December 25, 1989, 7.7° C (Robinson, 1991); Ft. Pierce: December 26, 1989, 11.1° C (Lawson, 1991). In Titusville and the south end of Mosquito Lagoon, water temperatures remained below 10° C for 5-7 days (Clameta, 1991; Robinson, 1991; Provancha, 1992). The 1977, 1983, and 1985 freeze events each broke existing low temperature records for Florida, and the reports of Snelson and Bradley (1978) and Provancha and co-workers (1986) for northern Brevard County and Gilmore and co-workers (1978) in Indian River and St. Lucie Counties provide some interesting comparisons for these events with the 1989 freeze. Temperature data for all 4 events are presented in Table 1. Clearly, the 1989 event resulted in lower air temperatures, but IRL water temperatures did not reach the extremes of the previous years. Additional data from Gilmore and co-workers (1978) support this, indicating that 1977 water tempera- tures in Ft. Pierce were below 10° C for at least 72 h, while in 1989 the Ft. Pierce power plant readings did not fall below 10° C. However, some inherent difficulties are apparent in comparing water temperature data for the 4 events. Power plants report intake water temperature, which may be affected by discharge plumes of heated water. In addition, varying wind or tidal flow conditions between events may affect movement of heated plumes and, thus, intake temperatures. The 1977 Vero Beach water temperatures of Gilmore and co-workers (1978) were obtained from a power plant intake, as were some of the 1989 data. All of the remaining data for 1977, 1983 and 1985 were field collected. Table 2 presents the collection data for R. marmoratus before and after the 1989 freeze event. A vigorous statistical interpretation of the data is not possible due to unequal sample sizes and some variability in collection techniques. For example, a “visit” with either hook and line or net could consist of sampling from 5- 12 burrows, with some of the same burrows sampled at each visit. Similarly, the trap effort varied, with trap sets in a single burrow ranging from 6-24 h. In spite of these procedural shortcomings, there is an apparent drastic decline of R. marmoratus populations No. 3, 1993] TAYLOR—IM PACT OF 1989 FREEZE 131 TABLE 1. Comparison of air and IRL/Mosquito Lagoon water temperatures for four freeze events in east central Florida* Variable/Locale Titusville Vero Beach Ft. Pierce Water Titus./Mosq. Lag. Vero Beach Ft. Pierce OW, -3.9° -3.9° “near 4°” 6.0° 6.0° Temperature (C), Year 1983 -5.5° 205 1985 -6.6° O05 1989 -7,2° -5.0° ete 7.7°/5.0° SILI °1977, 1983 and 1985 data from: Gilmore and co-workers, (1978); Snelson and Bradley, (1978); Provancha and co- workers, (1986). TABLE 2. Collection rates of Rivulus marmoratus at six sites in Brevard and Indian River Counties, Florida before and after the December, 1989 freeze. Pre-freeze (1/5/88-12/1/89) H&L or net* x #/visit Site (# visits) 1 18.6 (5) 2 2207) 3 AOm Gl) 4 SOF (2) 5 3) 6 2a) Gl) ee *Hook and line or hand net collection: x number collected per “visit” **Traps set 1/burrow: x number per trap Traps*° x #/trap (# traps) 1.6 (10) 0.3 (10) 1.3 (26) Post-freeze (12/28/89-10/25/91) H&L or net x #/visit (# visits) Traps x #/trap (# traps) 0 (8) 132 FLORIDA SCIENTIST [VOL 56 following the freeze. No dead fish were found immediately after the freeze (only one site checked on 12/28/89). The mortality reports for IRL fishes of Snelson and Bradley (1978), Gilmore and co-workers (1978) and Provancha and co-workers (1986) all concur in that, with few exceptions, the larger individuals of vulnerable species were killed first. Additionally, none of these investigators reported smaller species (eg. cyprinodonts, poeciliids, gobiids) killed in significant nunmbers. Gilmore and co-workers (1978) suggest that small, cryptic species were killed but not observed. The implications of both temperature and mortality data may not directly apply to R. marmoratus. Rivulus marmoratus has not been reported north of southern Brevard County (Taylor, 1992) and mortality of this species was not reported in any of the above studies. Taylor and co-workers (1992) indicated that R. marmoratus become moribund at 9° C in the laboratory, yet the crab burrow habitat provides some thermal buffering (Taylor, 1992). Therefore, the IRL open water temperature data may not accurately reflect the “pore” water temperatures found in a crab burrow. Huehner and co-workers (1985) reported a tendency for R. marmoratus to seek terrestrial shelter at 19-20° C in aquaria, but whether this behavior would continue at lower temperatures or occur at all in crab burrows is unknown. Recent reports (Huehner et al. 1985; Taylor, 1990; Taylor and Snelson, 1992) have added to knowledge of the distribution of R. marmoratus in Florida. Better recognition of the specific habitats of this fish has resulted in the following recent additions to the recorded distribution in the state. Pinellas County, Feather Sound, Old Tampa Bay (Lat. 27° 55'). Old mosquito ditches in mangroves. Two specimens. August, 1985. Collector: Jeff Brown Charlotte County, Charlotte Harbor (Lat. 26° 57'). Old mosquito ditch in mangroves. One specimen. January, 1992. Collector: William Dunson. Broward County, West Lake (Lat. 26° 02'). Old mosquito ditches and ephem- eral pond in mangroves. Thirty specimens. December, 1986-January, 1989. Collec- tor: Robert Whitman. Martin County, St. Lucie Inlet State Park (Lat. 27° 09'). Crab burrow in mangroves. One specimen. November, 1992. Collector: D. Scott Taylor Specimens from Pinellas and Charlotte Counties were examined by the author and identification confirmed. The record from Pinellas County is of particular interest, as this collection represents the northernmost collection on the west coast of Florida. The corresponding northern limit on the east coast (Brevard County) is only about 20 km north of the old Tampa Bay latitude. Snelson (1978) had previously reported an apparently unsubstantiated collection from St. Petersburg (Pinellas Co.). A limited collection effort by the author at the Old Tampa Bay site in August, 1992 did not produce any R. marmoratus. Although there is no documentation of the December 1989 freeze impacts in Tampa Bay, the damage to mangroves did not appear significant enough to have harmed the R. marmoratus population. In the 1977 freeze, Gilmore and co-workers (1978) indicated fewer fish killed in Tampa Bay than in IRL. The collection from Charlotte County suggests that the species is contiguous on the west coast from previously confirmed collections in the Ft. Myers-Naples area . No. 3, 1993] TAYLOR—IMPACT OF 1989 FREEZE 133 (Taylor and Snelson, 1992) north to Tampa Bay. The uncertainty of collections adjacent to Florida Bay (Taylor and Snelson, 1992) has also been clarified by recent extensive collections of this species by W. F. Loftus, W. P. Davis and D. S. Taylor (unpublished data). Distributional gaps on the east coast between Dade and Brevard Counties are partially filled in by the Broward and Martin County collections cited above. The species is now known from all east coast counties north to Brevard, with the exception of Palm Beach County. As a clearer picture of distribution develops for R. marmoratus in Florida, the affiliation of this species with mangrove habitats is reemphasized. Temperature limitations clearly affect the distribution of both R. marmoratus and mangroves. The decline in collection rates for the Brevard and Indian River County sites following the December, 1989 freeze indicates a significant impact on local populations. In the 3 sites where R. marmoratus were collected following the freeze, specimens were not found until 8-17 months after the event. The thermal “buffering” offered by crab burrows was apparently not sufficient to protect all of the fish from the extreme cold. Naturally occurring low population levels at this northern limit of the fish’s range and its inherent slow reproductive rate prevented rapid rebound of freeze-damaged populations. In addition, many crabs (Cardisoma guanhumi) were found dead in their burrows following the freeze. Provancha and co-workers (1986) point out that the localized effects of past freezes varied widely and absolute minimum temperature reached was not the only criteria for judging these effects. Nevertheless, the December 1989 event will no doubt be recorded as one of the worst of this century. LITERATURE CITED CLAMETA, A. 1991. Orlando Utilities Corp., Titusville, Pers. Commun. GILMORE, R. G., L. H. BULLOCK AND F. H. BERRY. 1978. Hypothermal mortality in marine fishes of south-central Florida January, 1977. Northeast Gulf Sci. 2:77-97. HUEHNER, M. K., M. E. SCHRAMM, AND M. D. HENS. 1985. Notes on the behavior and ecology of the killifish Rivulus marmoratus Poey 1880 (Cyprinodontidae). Florida Scient. 48(1):1-7. LAWSON, K. 1991. Ft. Pierce Utilities Authority, Ft. Pierce, Pers. Commun. Noaa. 1989. Climatological Data, Florida, December 1989. 93(12). PROVANCHA, M. J. 1992. Bionetics Corp., Kennedy Space Center, Pers. Commun. , P. A. SCHMALZER, AND C. R. HALL. 1986. Effects of the December 1983 and January 1985 freezing air temperatures on select aquatic poikilotherms and plant species of Merritt Island, Florida. Florida Scient. 49(4):199-212. ROBINSON, R. 1991. Florida Power and Light Corp., Titusville, Pers. Commun. SNELSON, F. F., JR. 1978. Rivulus, Rivulus marmoratus Poey. Pp. 18-19. In: Gilbert, C. R. (ed.), Rare and Endangered Biota of Florida. Vol. 4. Fishes. Univ. Press of Florida, Gainesville, FL. AND W. K. BRADLEY, JR. 1978. Mortality of fishes due to cold on the east coast of Florida, January, 1977. Florida Scient. 41:1-12. SNODGRASS, J. W. 1991. Winter kills of Tilapia melanotheron in coastal southeast Florida. Florida Scient. 54(2): 85-86. TAYLOR, D. S. 1988. Observations on the ecology of the killifish Rivulus marmoratus Poey (Cyprinodontidae) in an infrequently flooded mangrove swamp. Northeast Gulf Sci. 10(1):63-68. . 1990. Adaptive specializations of the cyprinodont fish Rivulus marmoratus. Florida Scient. 53(3):239-248. . 1992. Diet of the killifish Rivulus marmoratus collected from land crab burrows, with further ecological notes. Environ. Biol. Fishes. 33:389-393 AND F. F. SNELSON, JR. 1992. Mangrove rivulus, Rivulus marmoratus. Pp. 200-207. 134 FLORIDA SCIENTIST [VOL 56 GILBERT, C. R. (ed.), In: Rare and Endangered Biota of Florida. Vol II. Fishes. Univ. Press of Florida, Gainesville, FL. , S. A. RITCHIE, AND E. JOHNSON. 1992. The killifish Rivulus marmoratus: a potential biocontrol agent for Aedes taeniorhynchus and brackish water Culex. J. Amer. Mosquito Control Assoc. 8(1):80-83. Florida Scient. 56(3):129-134.1993. Accepted: April 2, 1993. FLORIDA ACADEMY OF SCIENCES 1993 ANNUAL MEETING OUTSTANDING STUDENT PAPER AWARDS AGRICULTURAL SCIENCES José R. Espaillat, Agronomy Department, University of Florida, “The Interaction of Fusarium oxysporum and Meloidogyne incognita Race 1 on Sage, Rosemary, and Lavender.” Graduate Co-award. Diego A. Diz, Agronomy Department, University of Florida, “Defoliation Effects on Seed Yield Components and Ease of Harvesting in Pearl Millet x Elephantgrass Hybrids.” Graduate Co-Award. ANTHROPOLOGICAL SCIENCES Kenneth J. Winland, Department of Anthropology, Florida Atlantic University, “Paleopathology and Paleodemography of the Highland Beach (8PB11) Population.” Graduate Award. Jini M. Hanjian, Florida Mental Health Institute, University of South Florida, “The Anthropologi- cal Perspective in the Study of the Children’s Mental Health Service System: An Assessment.” Honorable Mention, Applied Anthropology. James P. Pepe, Department of Anthropology, Florida Atlantic University, “Reconstruction of Prehistoric Cosmology and Social Organization: An Example from the Southeastern Ceremonial Complex.” Honorable Mention, Archaeology. ATMOSPHERIC AND OCEANOGRAPHIC SCIENCES No information available as of 6/10/93. BIOLOGICAL SCIENCES No information available as of 6/10/93. ENVIRONMENTAL CHEMISTRY AND CHEMICAL SCIENCES Charles D. Norris, Department of Chemistry, University of South Florida, “Application of Supported Chelating Agents for Extraction of Cadmium.” Graduate Co-Award. Elsie Gross, Department of Chemistry, University of South Florida, “Isolation and Characteriza- tion of a Hydroxamate Siderophore Produced by Lyngbya majuscula.” Graduate Co-Award. M. Stacey Thomson, Department of Chemistry, University of South Florida, “Speciation and Determination of N-Nitrosodimethylamine and NO, Species in Ambient Air.” Graduate Co-Award. FLORIDA COMMITTEE ON RARE AND ENDANGERED PLANTS AND ANIMALS Steven A. Johnson, Department of Biology, University of Central Florida, “Site Fidelity of the Florida Green Turtle at the Archie Carr National Wildlife Refuge.” Graduate Award. GEOLOGICAL AND HYDROLOGICAL SCIENCES Richard A. Hisert, Department of Geology, University of Florida, “The Underground Flow Path of the Santa Fe River in O’Leno State Park.” Graduate Award. Kenneth M. Lord, Department of Geology, University of Florida, “Basement Features of the Floridian Plateau Delineated from Digitally Filtered Bouguer Gravity Anomoly Maps.” Honorable Mention. MEDICAL SCIENCES No information available as of 6/10/93. Cont. on page 162 No. 3, 1993] KAUTZ ET AL._VEGETATIVE COVER IN FLORIDA 135 / Conservation Sciences VEGETATIVE COVER IN FLORIDA BASED ON 1985-1989 LANDSAT THEMATIC MAPPER IMAGERY RANDY S. Kautz’, D. TERRY GILBERT”, AND GREGORY M. MAULDIN”) Office of Environmental Services, Florida Game and Fresh Water Fish Commission, 620 S. Meridian St., Tallahassee, FL 32399-1600 ®Teon County Courthouse, 301 S. Monroe St., Room P308C, Tallahassee, FL 32301 ABSTRACT: The survival of many species of Florida wildlife is threatened by widespread habitat loss. In order to assist in the identification of critical wildlife habitats before they are developed for other uses, a map of 22 land cover types (i.e., wildlife habitats) was created for the entire state of Florida using georeferenced Landsat Thematic Mapper imagery dated 1985-1989. During the study period, lands supporting relatively natural vegetative cover occurred over 58% of Florida, and lands showing evidence of disturbance occurred over 42% of the state. Pinelands were the most abundant upland cover type, freshwater marsh was the most abundant wetland type, and grassland/agriculture was the most abundant disturbed land type. FLORIDA'S native species of wildlife are in trouble. Since 1513, when Ponce de Leon first arrived in Florida, at least 9 vertebrate taxa have been driven to extinction (Kautz, 1993). Five of these species have disappeared in the last 50 years alone. With 56 taxa (excluding fishes and whales) federally listed as endangered or threatened species, Florida is home to more listed species than all states except California (U.S. Fish and Wildlife Service, 1989). Moreover, the populations of 44% of Florida’s 668 vertebrate species are thought to be declining (Millsap et al., 1990). Until recently, many of these species were considered common and presumed to be secure. The single factor most commonly cited as the cause of declines in wildlife populations is loss of habitat due to development. In the last 50 years, 21% of Florida’s forest lands and 56% of its herbaceous wetlands were destroyed to make room for a population that grew from 1.7 to 12.9 million, to accommodate over 39 million tourists each year, and to make available an additional 1.7 million acres for agricultural use (Duda, 1987; Winsberg, 1992; Kautz, 1993). Virtually every acre of forest or wetland that was eliminated provided habitat for species of wildlife that are now declining. Scott and co-workers (1987) recognized that, in order to identify remaining habitats that are critical to the long-term protection of biological diversity, current vegetation maps are a must. In addition, due to the need to map areas as large as a state and due to the speed at which development is destroying remaining habitats, the techniques used to create current vegetation maps must be both rapid and cost- effective. Remotely sensed spectral data collected by the Landsat satellite is well suited to this purpose. 136 FLORIDA SCIENTIST [VOL 56 Early attempts to map vegetation using Landsat satellite Multispectral Scanner (MSS) data met with mixed results (Fox et al., 1983; Franklin et al., 1986). However, the Landsat Thematic Mapper (TM) sensor, which was launched into orbit in 1982, presented new opportunities due to increased spatial and spectral resolution. A 1986 feasibility study showed that Landsat TM imagery could successfully be used to map wildlife habitats in Florida (Wickham and Kautz, 1989). Subsequent efforts have proven the value of Landsat TM data, and many states are now using Landsat TM imagery to map vegetation as part of the U.S. Fish and Wildlife Service’s Gap Analysis program (Scott et al., 1993). This paper describes a 3.5-year project that used Landsat TM imagery to map land cover, as a surrogate for wildlife habitats, throughout the entire state of Florida. The project was completed between July 1, 1987, and December 31, 1990, as a cooperative effort among three state agencies. The Florida Game and Fresh Water Fish Commission designed the project, provided funding out of the Florida Nongame Wildlife Trust Fund, and contributed biological expertise to remote sensing experts. The Florida Department of Transportation’s remote sensing section performed all of the classification and interpretation work. The Florida Department of Natural Resources provided technical assistance in data management. METHODS—The first step in the project was the identification of the cover types that would be mapped. The criteria established for selecting cover types for mapping were: (1) each cover type should be readily identifiable by field biologists familiar with Florida, (2) a high degree of correlation should exist between each cover type and the species of wildlife that use it as habitat, (3) the selected types should have spectral signatures that can be separated from one other using Landsat TM imagery, and (4) the number of cover types mapped should be large enough to adequately represent the range of types present in Florida but not so numerous that classification accuracy would be compromised. A review of existing land cover classification systems revealed those with as few as 2 upland and wetland vegetation types in Florida (Shelford, 1963) to those with as many as 42 (Florida Natural Areas Inventory and Florida Department of Natural Resources, 1990). The Florida Department of Transpor- tation (1985) lists over 500 categories of land use and land cover in Florida, more than 100 of which could be useful indicators of wildlife habitat. Between these extremes are the 17 vegetation types mapped by Davis (1967) and the 26 types described by the U.S. Soil Conservation Service (undated). For the purposes of this study, the number and kinds of vegetation types appearing in the latter 2 classification systems seemed to best fit the criteria listed above. The first 6 months of the project were spent evaluating which of the vegetation types described by Davis (1967) and the Soil Conservation Service (undated) could be separated spectrally using Landsat TM data. It was found that 22 separate Florida cover types could be mapped successfully. Of the 22 cover types mapped during the project, 17 were natural vegetation types, 3 were disturbed vegetation types, 1 was barren land, and 1 was water. The term “natural,” used here to describe many of the vegetation types mapped during the project, must be taken loosely because many of the types have experienced some type of human-related disturbance. For example, the pinelands class includes many areas that are in short-rotation planted pines and are not natural pine stands. Other forest types, such as sandhill and mixed hardwood-pine, may suffer from or be the product of fire suppression and may not support the biological diversity characteristic of pristine natural stands. On the other hand, the vegetation types considered here as disturbed types (i.e., shrub and brush, grassland, exotic plants) generally represent early successional stages following severe disturbance or are lands in agricultural use. However, in some cases they represent natural shrub-dominated communities that could not be differentiated from disturbed lands with the methodology used in the project. Brief definitions for each of the 22 cover types appear in the Glossary. The computer hardware used for image processing included a Gould MPX32-SEL minicomputer, Comtal 65SER image processing system, Calcomp 9000 digitizer, a Dell 286 microcomputer, a Dell 386 microcomputer, and a Tektronix 4896 color inkjet plotter. ATLAS Remote Imaging System (Version 1.13) image processing software, developed and marketed by Delta Data Systems, Inc., Picayune, MS, No. 3, 1993] KAUTZ ET AL.—VEGETATIVE COVER IN FLORIDA ibe 57 was used for all image processing tasks. Classification of the Landsat TM data was performed separately for each of the multi-county regions served by Florida’s 11 Regional Planning Councils (RPC). RPCs provide regional coordination of growth management and land-use planning for three or more contiguous counties (Figure 1). All Florida counties have prepared a mandatory comprehensive land use plan, a required component of which is a current vegetation map. Mapping the state by region divided the state into more manageable mapping units and simultaneously provided interim products useful to ongoing land use planning rograms. a All or portions of 17 Landsat TM scenes, with imagery dates ranging from 1985 in the Florida Keys to 1989 in the Panhandle, were used to map current Florida vegetation (Figure 2). The pixel size of each scene was resampled to 30 m. The first stage of image processing was to mosaic together the four quadrants of each Landsat scene. Each scene was then geocorrected to fit the Universal Transverse Mercator (UTM) coordinate system, North American Datum 1927 (NAD27). The source for planimetric control was U.S. Geological Survey 7.5-minute quadrangle maps. Map precision is roughly equal to one-half of a TM pixel. The root- mean-square error of the control network for each scene was less than 15 m. Jefferson County, Florida, is the only county bisected by the two UTM Zones (16 and 17) that occur in the state. The western portion of the county lies within Zone 16 and is part of the Apalachee Region. All of Jefferson County was mapped to Zone 16 rather than dividing it between two UTM zones and two RPCs. New raster matrices representing the area included within each RPC were extracted from the geocorrected scenes. This often involved mosaicking together portions of two or more Landsat scenes North Central Apalachee East Central Withlacoochee Treasure Coast Fic. 1. Locations of the 11 regional planning councils in Florida. 138 FLORIDA SCIENTIST [VOL 56 6/3/88 10/23/87 ; A1 / noes |g FIG. 2. Dates and locations of the Landsat Thematic Mapper scenes used to map Florida land cover. with different dates. The planimetric offsets were customized for each region such that the UTM coordinates were correct throughout the region. The raster matrix covering each region was divided into subscenes of approximately 1,024 lines by 1,024 elements consisting of all 7 geocorrected TM bands. The smaller subscenes were created in an effort to minimize heterogeneity within spectral classes while still providing an adequate sample of pixels for accurate signature training. Contrast enhancement algorithms were applied to all Landsat TM bands in each subscene to maximize separability between spectral classes. Care was taken to maintain the spectral response relationships encoded in the image data. Vegetation-index and brightness-index ratio bands were generated for each subscene to better define the variance in spectral responses among vegetation types and to better separate bare lands from disturbed vegetation. Vegetation-index ratio bands were created by (1) dividing band 4 (near infrared) by band 3 (visible red), (2) dividing band 5 (mid-infrared) by band 2 (visible green), and (3) dividing band 7 (far infrared) by band 1 (visible blue). The 4/3 ratio band was used to separate vegetated areas from non-vegetated areas and to enhance differentiation of vegetation types based on high-contrast visible green wavelengths. The 5/2 ratio band also is useful for separating vegetated from non-vegetated areas, but it allows for differentiation of vegetation types based on high-contrast infrared wavelengths. The 7. ratio band occasionally proved useful in the differentiation of forested wetland types, particularly bay swamps. A brightness-index ratio band was created by dividing band 3 by band 2, and was used to identify No. 3, 1993] KAUTZ ET AL._VEGETATIVE COVER IN FLORIDA 139 | bare areas. Correlation matrices were developed to identify which raw and ratio bands were highly correlated. The raw bands and ratio bands which provided maximum separability of spectral classes were used to assemble multi-channel data sets, consisting of 3-7 raw and ratio bands, for signature training. The raw bands most commonly used were 1, 3, 4, 5, and 7. Raw information in band 2 generally was highly correlated with information in other bands and was not was not used for signature training or classification because it did not enhance separability of vegetation classes. The optimized multi-channel data sets for each subscene were processed with an unsupervised signature training algorithm to generate training statistics. Although a supervised method generally provides the analyst greater control over the process, this method was deemed too labor intensive for such a large project. The homogeneity parameters used to define the training statistics (i.e., upper and lower standard deviation and covariance) were set to conform to the distribution of spectral response values relative to the land-cover classes for each of the input bands. The number of spectral classes to be output was set to the system default of 59. The n-dimensional spectral distances among the training signatures were calculated and analyzed to verify the validity of the signatures. Like classes were either merged or deleted. If the covariance of aclass fell below a threshold that was dependent on the vegetation types present in the subscene area, the statistics were deleted and regenerated under modified homogeneity conditions before the classified image was generated. Once a set of valid signatures was obtained, it was processed through a canonical correlation algorithm to produce a new set of uncorrelated training signatures. A Euclidean distance classifier was used to generate,a classified image of each subscene of TM data based on the canonical signatures. Each of the classified subscene images was compared to black and white aerial photography, and each of the spectral classes in the classified image was assigned to one of the 22 land cover classes. Collateral information, such as cover-type maps with high spatial accuracies, were used when available to assist in the interpretation of classification results. Spectral classes containing more than one cover type were compared to aerial photography and interactively edited to improve map accuracy. Once the preliminary image interpretations were completed, each subscene was reviewed to locate potential ground-truth sites which generally consisted of extensive areas of natural vegetation. Distur- bance features, such as clearcuts, pastures, and old fields, are common in Florida, and few if any areas remain in a truly pristine condition. Thus, despite the emphasis on remote natural areas, disturbances often were present on the sites selected for ground-truthing, and, in fact, they proved to be useful landmarks that aided navigation in the field. Other criteria used for selecting ground-truth sites were: (1) areas where heterogeneity of vegetation types confounded the interpretation of aerial photography; (2) areas where two or more cover types appeared to be confused by the classifier; and (3) areas in which certain cover types could not be discerned from the aerial photography. An effort was made to select ground-truth points that were evenly distributed to ensure that final corrections were not biased by field observations from only a few parts of a region. A color-coded land cover map was prepared for each site selected for ground-truthing. All ground- truth maps were printed at a scale of approximately 1:24,000 and covered a rectangular area of 2,430 ha. Each map was mounted on poster board and covered with a piece of mylar for use in the field. The UTM coordinate of the center point of each map was recorded and converted to latitude and longitude coordinates. Access to the many remote sites selected for ground-truthing was obtained via a helicopter equipped with a Loran unit into which the latitude and longitude coordinates of each site had been entered. While flying at altitudes of 50-200 m over the area represented on a ground-truth map, notes were made on the mylar overlay of classes that were correctly and incorrectly mapped. The number of ground-truth maps checked in this manner ranged 21-56 per region (see Table 1). A total of 454 ground- truth maps, distributed as shown in Figure 3, was inspected in the field during the course of the project. The total area ground-truthed was 1,102,700 ha, which is 8% of the land area of the state. The information obtained during ground-truthing was used to correct inaccuracies contained within each subscene. Corrections were made not only to mapped areas actually visited in the field, but also to other subscene areas which had the same spectral classes as those observed to be incorrect at the ground-truth sites. Field observations of cover types difficult to interpret from aerial photography proved invaluable in correcting classification and interpretation errors occurring in subscenes not subject to ground-truthing. Once all corrections had been made, the subscenes covering a region were then mosaicked together. This frequently resulted in unmatched cover types along subscene boundaries. In what amounted to a huge undertaking, problem areas were edge-matched to the extent possible. At this point in the process, aregion map was considered acomplete and useable product. The steps described above were repeated for each of Florida’s 11 RPC regions. Once cover types in all 11 regions 140 FLORIDA SCIENTIST [VOL 56 TABLE 1. Number of points surveyed, total area surveyed, and dates of survey for ground-truth points in each Florida region. Total Hectares Region points surveyed Survey Dates Northeast Florida 48 116,600 Aug 08-12, 1988 Withlacoochee 42 102,000 Sep 19-24, 1988 Tampa Bay 34 82,600 Nov 01-04, 1988 Central Florida 48 116,600 Jan 09-13, 1989 Southwest Florida 56 136,000 Apr 10-14, 1989 May 22-26, 1989 Treasure Coast 21 51,000 May 22-26, 1989 South Florida 42 102,000 Jul 24-28, 1989 North Central Florida 47 114,200 Nov 13-17, 1989 Apalachee 46 111,700 Feb 19-23, 1990 West Florida 35 85,000 Apr 16-20, 1990 East Central Florida 35 85,000 Jul 23-27, 1990 Totals 454 1,102,700 had been mapped, unmatched cover types along adjacent region boundaries were edge-matched. Before this could be accomplished, however, it was first necessary to digitize the boundaries of all 67 Florida counties from 1:24,000 scale U.S. Geological Survey topographic maps. County boundaries were used to extract image data from each composite region data set to create a cover map of each individual county. Then, adjacent counties along region boundaries were mosaicked together. Unmatched cover types along region boundaries were edge-matched by referring to aerial photography and reinterpreting the classifications. Once edge-matching along region boundaries was complete, final data sets for each region were recompiled from the individual, edge-matched county image maps. Final project products were: 1) a digital cover map of each of Florida’s 11 regions, and 2) a separate digital cover map of each of Florida’s 67 counties. Unfortunately, due to the large size of Florida, limited funding and manpower, and demanding time schedule, no standard accuracy analysis was completed for the final cover map. A secondary objective of this project was to compare the extent and distribution of pre-settlement and present day cover types. To accomplish this, the Davis (1967) 1:1,320,000 scale map of potential natural vegetation in Florida was digitized, and the area of each of the vegetation types appearing on the map was calculated. The Davis (1967) map is presumed to reflect the pre-settlement area of each of 17 major vegetation types occurring in Florida. To the extent possible, area estimates derived from the Davis (1967) map were compared with the results of this mapping effort. However, a one-to-one correspondence does not exist between Davis (1967) types and the types mapped by this project, so only those comparisons that were actually possible were made. RESULTS AND DISCUSSION—Natural vegetation types—During the 1985-1989 No. 3, 1993] KAUTZ ET AL—_VEGETATIVE COVER IN FLORIDA 14] / FIG. 3. Approximate distribution of points used to ground-truth Florida land cover maps developed from Landsat Thematic Mapper imagery. study period, natural vegetation types found on upland sites covered 4.89 million ha, or 35%, of Florida, and wetland vegetation types covered 3.32 million ha, or 24% (Table 2). Together, land vegetated by upland and wetland plant communities covered 8.21 million ha, or 58%, of the landscape. ; Pinelands—The most abundant vegetation type in Florida is pinelands, a category which includes all lands dominated by pines except sandhills and sand pine scrub. Pinelands cover 2.65 million ha (19%) of the landscape. Pinelands are almost exclusively confined to the Panhandle and the northern third of the Peninsula. Florida pinelands occur predominantly on flatwoods and clay-hill sites. Sandhill, sand pine scrub, and xeric oak scrub—During the 1985-1989 study period, sandhill covered 344,500 ha (2.4%) of Florida, sand pine scrub covered 97,100 ha (0.69%), and xeric oak scrub covered 73,800 ha (0.52%). The remaining sandhills in Florida occur as highly fragmented patches found on well-drained upland ridges in north Florida and the Panhandle. The largest remaining tract of sand pine scrub is in the Ocala National Forest in north-central Florida. Scattered patches of xeric oak scrub occur along the southern end of the Lake Wales Ridge, 142 TABLE 2. Area and percent cover of 22 Florida land cover types based on 1985-1989 Landsat Thematic Mapper satellite imagery. Land Cover Class NATURAL VEGETATION TYPES Uplands Pinelands Upland Hardwood Forests Dry Prairie Sandhill Mixed Hardwood-Pine Forests Sand Pine Scrub Xeric Oak Scrub Tropical Hardwood Hammock Coastal Strand Uplands Subtotal Wetlands Freshwater Marsh & Wet Prairie Mixed Hardwood Swamp Cypress Swamp Shrub Swamp Mangrove Swamp Coastal Salt Marsh Bay Swamp Bottomland Hardwood Forest Wetlands Subtotal Natural Vegetation Subtotal FLORIDA SCIENTIST Area (ha) 2,648,814 932,085 561,114 344,477 222,539 97,056 73,753 6,175 5,377 4,891,390 1,095,865 772,830 662,675 272,384 221,221 196,486 63,650 36,334 3,321,445 8,212,835 Cover (%) 18.80 6.60 4.00 2.40 1.60 0.69 0.52 0.04 0.04 34.69 7.80 5.50 4.70 1.90 1.60 1.40 0.45 0.26 23.61 58.30 No. 3, 1993] KAUTZ ET AL—VEGETATIVE COVER IN FLORIDA 143 TABLE 2. (Con’t) Area Cover Land Cover Class (ha) (%) DISTURBED LAND TYPES Grassland & Agriculture 2.537, 168 18.00 Barren & Urban Land 1,678,673 11.90 Shrub & Brush Land M6SO17 11.70 Exotic Plant Communities 16,302 0.12 Disturbed Land Subtotal 5,883,160 41.72 TOTAL LAND AREA 14,095,995 100.02 OPEN WATER 1,732,697 — along old sand dunes just inland of the Atlantic and Gulf coasts, and in clearcut areas of the Ocala National Forest. A comparison with other work (Davis, 1967; Kautz, 1993) shows that sandhill and scrub area have declined markedly over time. Area calculations from the digitized version of the Davis (1967) general vegetation map of Florida indicates the potential for 2.79 million ha of sandhill (i.e., longleaf pine/xeric oak) in Florida prior to settlement (Table 3). Comparison with area estimates based upon the imagery data indicates an 88% loss of the original sandhill vegetation from Florida. Perhaps coincidentally, a recent compilation of U.S. Forest Service data shows that longleaf pine, the dominant tree of sandhills, declined 88% in the last 50 years (Kautz, 1993). Area calculations from the Davis (1967) map also show that sand pine scrub originally covered 418,100 ha of pre-settlement Florida. Although Davis (1967) is not clear on the point, his sand pine scrub class apparently would include our xeric oak scrub class. Adding the sand pine and xeric oak scrub classes from this study yields a total of 170,809 ha of scrub remaining in Florida in the 1985-1989 study period. Comparing the Davis (1967) map estimate of scrub area with our results shows that scrub communities have declined 59% since settlement. The loss of scrub is a matter of great interest to conservationists because of the large number of rare and endemic species found only in scrub habitats (Myers, 1990). All pine forests—Adding the areas in the pinelands, sandhill, and sand pine scrub classes allows for an estimate of the total area of Florida supporting forests dominated by pines. During the study period, pine forests covered 3.09 million ha 144 FLORIDA SCIENTIST [VOL56 TABLE 3. Estimates of the pre-settlement area and percent cover of 17 natural plant communi- ties in Florida derived from the Davis (1967) general map of natural vegetation. Area Cover Plant Community (ha) (%) Pine Flatwoods 4,846,500 34.4 Longleaf Pine/Xeric Oak 2,789,000 19.8 Hardwood Swamp 1,119,400 7.9 Wet and Dry Prairies 825,300 5.9 Mixed Pine-Hardwood 791,300 5.6 Hardwood Forest 666,900 AG, Everglades Sawgrass 428,100 3.0 Mangrove/Coastal Marsh 421,500 3.0 Sand Pine Scrub 418,100 3.0 Rockland Marshes 411,900 2.9 Everglades Prairies 340,700 2.4 Coastal Strand 249,800 1.8 Freshwater Marsh 185,300 1.3 Scrub Cypress 185,100 3 Cypress Swamp 164,500 2 Pine Rocklands 154,800 a Cabbage Palm Hammock 97,800 0.7 Total 14,096,000 100.0 (22%) of Florida, and accounted for 63% of all upland vegetation types and 38% of all natural vegetation types. A forest inventory completed in 1987 showed that slash pines dominate in 69% of all pine forests (Bechtold et al., 1990). In addition, 53% of all pine forests are in pine plantations, a condition which provides poor quality wildlife habitat relative to natural stands (Harris et al., 1974; Umber and Harris, 1974; Repenning and Labisky, 1985; McComb et al., 1986). Upland hardwood forests—In this study, the major upland forest classes No. 3, 1993] KAUTZ ET AL.—-VEGETATIVE COVER IN FLORIDA 145 dominated by hardwoods were upland hardwood forests and mixed hardwood-pine forests. During the study period, upland hardwood forests covered 0.93 million ha (6.6%) and mixed hardwood-pine forests covered 0.22 million ha (1.6%) of Florida. Together, the two hardwood classes covered 1.15 million ha, and accounted for 24% of all upland communities and 14% of all natural vegetation types. In comparison, Kautz (1993) reported 0.77 million ha of oak-hickory (i.e., upland hardwood) forests, 0.49 million ha of oak-pine forests, and a total of 1.26 million ha of upland hardwood forests in Florida in 1987. In pre-settlement times, extensive hardwood forests covered 1.56 million ha of Florida as estimated by summing the Davis (1967) mixed pine-hardwood, hardwood forest, and cabbage palm hammock classes. Assuming that these Davis (1967) classes roughly correspond to the upland hardwood forest and mixed hardwood-pine forest classes mapped during this inventory, it is estimated that 0.40 million ha, or 26%, of Florida’s hardwood forests have been lost to development since the advent of European man. In pre-settlement times, hardwood forests occurred primarily on upland ridges formed by limestone outcrops in the northern third of the Peninsula, and mixed hardwood-pine forests occurred on the rolling clay-hills of the Panhandle (Davis, 1967). However, these forests have long since been converted to other uses, mainly pasture, agriculture, and silviculture. Today, the largest remaining stands of hard- wood forest are the hydric hammocks along the Gulf coast from Wakulla to Hernando County, and the bluff forests along the Suwannee, Aucilla, and Chipola rivers in north Florida. Other notable stands of hardwoods occur along the Brooksville Ridge in Hernando County; at the northern end of Green Swamp; along the lower Hillsborough, upper Myakka, and upper Peace rivers; and scattered throughout Big Cypress Swamp. Many of the remaining upland hardwood and mixed pine-hard- wood forests are small, highly fragmented patches scattered throughout north Florida. Dry prairies—Historically, dry prairies were found in broad flat areas of south- central Florida along Fisheating Creek, the Kissimmee River, and the upper St. Johns River (Davis 1967). In this study, any area of Florida supporting vegetation similar to that of dry prairies was mapped as dry prairie regardless of whether or not it fell within the historic distribution as defined by Davis (1967). During the mapping period, areas supporting dry prairie vegetation covered 0.56 million ha (4%) of Florida, and accounted for 11% of upland vegetation types and 7% of all vegetation types. Much of the original dry prairie habitat has been converted to improved pasture for the production of cattle. Area calculations from the Davis (1967) map show that there were 0.83 million ha of prairie habitat in pre-settlement Florida. However, area estimates from the Davis (1967) map are not directly comparable with the results of this study because (1) Davis (1967) included in his prairie class wet depressions that were mapped as freshwater marsh in this study, and (2) some of the areas mapped as dry prairie in this study were located outside of the areas depicted on the Davis (1967) map. Dry prairies are important habitats for several rare or threatened species of wildlife including gopher tortoise (Gopherus polyphemus), 146 FLORIDA SCIENTIST [VOL 56 crested caracara (Polyborus plancus audubonii), burrowing owl (Athene cunicularia), and Florida grasshopper sparrow (Ammodramus savannarum floridanus). Tropical hardwood hammock and coastal strand—The least extensive vegeta- tion types mapped during the project were tropical hardwood hammock and coastal strand. Tropical hardwood hammocks covered only 6,175 ha of Florida and coastal strand covered only 5,377 ha during the study period. Neither of these vegetation types has ever been very abundant in Florida, even prior to settlement. The largest tracts of remaining tropical hardwood hammock are in the Florida Keys, especially on North Key Largo. Due to the popularity of beaches and sand dunes as residential and recreational areas, coastal strand presently is restricted to small patches along undeveloped high energy shorelines. Freshwater marsh and wet prairie—The most abundant wetland community in Florida is freshwater marsh and wet prairie (Table 4). During the study period, freshwater marshes and wet prairies covered 1.10 million ha (7.8%) of Florida. They comprised 33% of all Florida wetlands and 13% of all natural vegetation types. The largest remaining expanses of freshwater marsh in Florida occur in the Everglades, in Big Cypress Swamp, along the upper St. Johns River, and along the western shore of Lake Okeechobee. The extent and distribution of remaining freshwater marshes is a major concern because Florida has lost 1.57 million ha of herbaceous wetlands to development in the last 50 years (Kautz, 1993). During the same period, south Florida wading bird populations declined 90% as a result of drainage, alteration of hydroperiod and flow patterns, introduction of nutrients and pollutants, and invasion of exotic pests in marsh habitats (Bildstein et al., 1991). A comparison of these results with the Davis (1967) map was not possible because Davis lumped wet and dry prairies into a single vegetation type. Coastal salt marsh—Coastal salt marsh covered 196,500 ha (1.4%) of Florida during the study period. Salt marshes are most abundant along the Gulf coast from Hernando to Wakulla County; behind the barrier islands of Nassau, Duval, and St. Johns counties; and in a zone transitional between the freshwater marshes and mangrove swamps at the tip of south Florida. Total marsh—Adding the coastal salt marsh and freshwater marsh classes together yields a total marsh coverage of 1.29 million ha in Florida during the 1985- 1989 study period. This value compares well with the total marsh area of 1.25 million ha reported by Kautz (1993) based on U. S. Forest Service inventory data from 1987 and the value of 1.29 million ha in 1984 as obtained by summing the palustrine emergent, palustrine aquatic bed, and estuarine intertidal emergent classes of Frayer and Hefner (1991). Forested wetlands—The forested wetland classes included in this study were mixed hardwood swamp, cypress swamp, bay swamp, and bottomland hardwood forest. During the study period, mixed hardwood swamp covered 0.77 million ha — (5.5%) of Florida, cypress swamp covered 0.66 million ha (4.7%), bay swamp covered 0.064 million ha (0.45%), and bottomland hardwoods covered 0.036 million ha (0.26%). Collectively, forested wetlands covered 1.54 million ha (11%) of Florida, and accounted for 46% of all wetland communities and 19% ofall plant communities. No. 3, 1993] KAUTZ ET AL._VEGETATIVE COVER IN FLORIDA 147 Large tracts of forested wetlands are most common in Big Cypress Swamp, Green Swamp, the lowlands of Volusia and Flagler counties, Pinhook Swamp in Baker and Columbia counties, and the floodplains of many Florida rivers. Although the study results were not subjected to an accuracy assessment, in all likelihood, the area of bay swamps in Florida was underestimated by this study. Interpretive skills in the identification of bay swamps from aerial photography improved as the project progressed. As a result, cover maps of the regions of the state mapped early in the project (i.e., east central Florida, northeast Florida, Withlacoochee, and south Florida) show few bay swamps whereas regions mapped last (i.e., north-central Florida and the Panhandle) show more. Where bay swamps were omitted, they most likely were mapped as mixed hardwood swamp. Shrub swamps—Shrub swamps covered 0.27 million ha (1.9%) of Florida during the study period. Shrub swamps dominated by titi are abundant in the wet flatwoods of the Apalachicola National Forest in the Panhandle and in Pinhook Swamp between Osceola National Forest and the Georgia state line. Shrub swamps dominated by willows and wax myrtle occur in wetlands that have experienced recent or repeated disturbance or have suffered from an absence of periodic fires. Willow and wax myrtle shrub swamps are typical of the phosphate-mined regions of central Florida, riverine sand bars and canal banks throughout the state, the water conser- vation areas of the Everglades, and abandoned wet pastures. Mangrove swamps—Mangrove swamps covered 0.22 million ha (1.6%) of Florida during the study period. Mangroves are most abundant along the coast of southwest Florida from Charlotte Harbor south through the Florida Keys. Small but notable patches of mangrove swamp occur along the eastern shore of Tampa Bay, along the western shore of Biscayne Bay, in the Indian River just north of Ft. Pierce Inlet, at the northern end of Mosquito Lagoon, and in the vicinity of Ponce de Leon Inlet. Coastal wetlands—On his map of general vegetation types, Davis (1967) lumped mangrove swamps and coastal salt marshes together into a single coastal wetlands class. Area calculations from the Davis (1967) map indicate the potential for 421,500 ha of coastal wetlands in Florida at the time of settlement. Adding the areas of coastal salt marsh and mangrove swamp from this study yields a current estimate of 417,700 ha of coastal wetlands in Florida. Comparing the Davis (1967) map with the results of this study suggests that 3,800 ha (0.9%) of Florida’s coastal wetlands have been lost to development since settlement. Large areas of mangrove swamp were destroyed in the vicinities of Marco Island, Boca Ciega Bay, Terra Ceia Bay, and Miami Beach to create prime waterfront property. Disturbed lands—Disturbed land cover types include grassland and agricul- ture, shrub and brush land, urban and barren land, and exotic plant communities. During the study period, disturbed lands covered 5.88 million ha (42%) of Florida. Since most disturbed lands occur on uplands (with some exceptions, such as the Everglades Agricultural Area), it is possible to estimate the extent to which native upland vegetation has been disturbed by adding and then comparing the areas of disturbed lands and natural vegetation types found on uplands. This calculation reveals that 54% of all uplands are in a disturbed condition whereas only 46% of 148 FLORIDA SCIENTIST [VOL 56 Florida’s uplands support natural vegetation. Grassland and agriculture—During the study period, the most extensive type of disturbed land was the grassland and agriculture class. Grassland and agriculture covered 2.54 million ha (18%) of Florida and accounted for 43% of all disturbed lands. Grassland and agriculture are most common in the south-central portion of the Peninsula, on the upland ridges stretching from the Tampa Bay area north to the Georgia line, and in Jackson County in the Panhandle. Barren and urban lands—Barren and urban lands covered 1.68 million ha (12%) of the state, accounting for 29% of all disturbed lands. Barren and urban lands are most common along the southeast coast; around Charlotte Harbor and Tampa Bay; in the vicinities of Orlando, Jacksonville, and Pensacola; in the phosphate mining region of west central Florida; and along the Lake Wales Ridge down the center of the Peninsula. Shrub and brush lands—Shrub and brush land covered 1.65 million ha (12%) of Florida, accounting for 28% of all disturbed lands. Shrub and brush lands are commonly associated with commercial timber operations in north Florida and the Panhandle. In these areas of the state, the practice of clearcutting and planting pines results in many sites in an early successional stage dominated by shrubs and sapling trees. In addition, citrus groves often were mapped as shrub and brush land, especially in the vicinity of Lake Apopka where hard freezes in the mid 1980s killed citrus trees over a wide area. Exotic plants—Exotic plants covered 16,300 ha (0.12%) of the state, accounting for only 0.28% of disturbed lands. Large stands of exotic plants are most common in extreme south Florida. Exotic plants were probably underestimated in our study because of the difficulty of identification on aerial photography, small patch size, and rapid proliferation. Water—The total area of water within the study area was 1.73 million ha. However, this is an arbitrary number due to the manner in which it was calculated. Because approximately three-fourths of Florida is surrounded by the Atlantic Ocean and the Gulf of Mexico, the area figure for water is a function of distance from land used define the extent of Florida. For example, one could choose the 4.85 km territorial limit, only that area of Florida landward of mean sea level, or some distance in between as the boundary for this calculation. A distance of 0.4 km from the coast was arbitrarily chosen as the distance from shore used to calculate the area of water covering Florida. Problems with mapping forested wetlands—The Federal Geographic Data Committee (1992) recently concluded that mapping wetlands with satellite data provides results inferior to those obtained using conventional aerial photography interpretation techniques. A comparison of the results of this study with those of the U.S. Fish and Wildlife Service’s National Wetlands Inventory (Frayer and Hefner, 1991) would seem to bear this out, at least with respect to the mapping of forested wetlands. Frayer and Hefner (1991) documented the presence of 4.23 million ha of ~ wetlands in Florida in 1984, or 0.91 million ha more than the 3.32 million ha of wetlands recorded by this study. The differences in the results of these two studies appears to be in the areas of No. 3, 1993] KAUTZ ET AL.—VEGETATIVE COVER IN FLORIDA 149 forested freshwater wetlands and wet pine flatwoods. Frayer and Hefner (1991) reported the presence of 2.21 million ha of forested freshwater wetlands in Florida versus the 1.54 million ha recorded by this study. Similarly, Frayer and Hefner (1991) found 0.47 million ha of palustrine scrub/shrub wetlands (which include wet pine flatwoods) whereas this study documented only 0.27 million ha of shrub swamp. Together, the differences in forested freshwater wetlands and wet pine flatwoods account for 0.87 million ha of the 0.91 million ha difference in total wetland area between the two studies. The underestimation of forested freshwater wetlands in this study was probably the result of including hydric hammocks, a type of wetland hardwood forest, in the upland hardwood forest class. This was done as a matter of convenience due to the difficulty of separating hardwood hammocks into xeric, mesic, and hydric types using Landsat TM data alone. Likewise, in this study it was necessary to create a single comprehensive class for all lands vegetated by pine forests (i.e., pinelands) due to the similar spectral characteristics of closed-canopy pine forests of all types. As a result, wet pine flatwoods sites, which function ecologically as wetlands, were included in the pinelands class whenever a site had a closed canopy of pines, even though an understory of wetland shrubs was present. Despite the problems with mapping forested freshwater wetlands, both studies produced almost identical estimates of the area of herbaceous and coastal wetlands in Florida. This suggests that Landsat TM data is a very useful tool for mapping freshwater marshes, coastal salt marshes, and mangrove swamps at the community level. Problems caused by community and landscape heterogeneity—Most of the vegetation types mapped during this project could be mapped successfully using Landsat TM data as long as specific sites presented homogeneous spectral informa- tion to the Landsat sensor. However, problems in classification arose when there was a high degree of spectral heterogeneity within community types or at the landscape level. Classification algorithms frequently produced numerous spectral classes within the same community type as a result of such factors as variations in the extent of canopy closure, clustering of a single plant species in one location, subtle variations in soil moisture and soil type, and degree of flooding in wetlands. For example, closed canopy stands of longleaf pines in sandhills exhibited a pinelands signature, and clumped stands of turkey oaks on sandhill sites often were confused with the shrub and brush class. Similarly, bare patches of sand within scrub communities exhibited the spectral characteristics of the barren lands class even though bare areas are a common component of the scrub community. Labor-intensive on-screen editing was necessary to reassign confused classes to the correct community type. Heterogeneity at the landscape level occurs when a high degree of interspersion exists among two or more community types over a large area. Examples include the ridge and swale systems of relict sand dunes that support alternating scrub and marsh communities (e.g., St. Vincent Island in the Panhandle); the mosaic of shrub swamp, cypress swamp, and pine forests in the wet flatwoods of the Apalachicola National Forest; and the mosaic of sawgrass, wet prairie, flag, and cattail associations in the 150 FLORIDA SCIENTIST [VOL 56 Everglades. Landscape level heterogeneity presents two types of problems. First, many pixels are likely to straddle more than one community or association type such that the spectral information in the mixed pixels represents none of the vegetation types actually present on the ground. As a result, the classification algorithm generates spectral classes that have no real meaning with respect to the vegetation types being mapped. Second, in heterogeneous landscapes, the persons interpreting the spectral classes from aerial photography and the persons engaged in ground-truthing often are not be able to discern which vegetation classes are actually represented by the confused spectral classes. In most cases of high landscape heterogeneity, it was necessary to make a judgement call concerning the placement of confused spectral classes into specific vegetation classes, and entire areas were circumscribed with polygons and placed into the appropriate, often more generalized, vegetation class. Seasonality of image data—Most of the imagery used in this project was from the months of March and April. However, a couple of scenes from the months of June and October were used, and imagery taken in December and January was used for the Keys. The differences in seasonality of the data presented few problems, largely because subscenes from the same date were processed individually. Only after classification were subscenes from different dates mosaicked together to create vegetation maps of larger regions. The only real problem presented by using imagery ranging in time from 1985 to 1989 was that the final statewide map represents the status of land cover in Florida over a “smudge” in time rather than a single year, and that will complicate future comparisons with these results. With the exception of the December and January scenes, all of the imagery was taken either early in the growing season or in the fall prior to leaf fall. During these times, photosynthetic activity was sufficient to provide reliable spectral signatures for the vegetation types being mapped. Even though the Keys scenes were taken in winter, vegetation growth still occurs at this time of year because of the proximity of the area to the tropics, and reliable spectral signatures are obtained. For the state as a whole, imagery taken in the spring seems to provide the most usable spectral information for vegetation classification. At this time of the year, vegetative growth is well underway throughout the state, whereas in the winter photosynthetic activity is minimal and deciduous vegetation is lacking leaves. Spring time also marks the end of the winter dry season in south Florida, a time when water tables are at their lowest. As a result, confusion of vegetation types due to extensive areas of standing water are minimized. Imagery taken in the summer typically is not useable because of too much cloud cover, high moisture content in the atmosphere, and extensive flooding of the flat areas of south Florida due to summer rains. ACKNOWLEDGMENTS—Many persons assisted with this project. Technical assistance in data management, image processing, and aerial photography interpretation was provided by Khaleda Hatim, Doug Hallman, Gail McAuley, Will Rice, Jesse Day, and Judy Elert. Administrative assistance was provided by Brad Hartman, Don Merkel, Alan Shopmyer, and Ken Haddad. Special credit goes to helicopter pilot Lance Ham who participated in ground-truthing efforts throughout the state. Ground- truthing assistance was provided by Rick Gooch, Kim Dryden, Brian Barnett, Ray Fernald, Mike Allen, Lee Edmisten, and Larry Perrin. John Hefner, Kevin Pope, and an anonymous reviewer provided many useful suggestions for improving the final manuscript. No. 3, 1993] KAUTZ ET AL._VEGETATIVE COVER IN FLORIDA 151 LITERATURE CITED BECHTOLD, W. A., M. J. BROWN, AND R. M. SHEFFIELD. 1990. Florida’s forests, 1987. U.S.D.A. Forest Service, Southeastern Forest Experiment Station, Resource Bull. SE-110, Asheville, NC. 61 pp. BILDSTEIN, K. L., G. T. BANCROFT, P. J. DUGAN, D. H. GoRDON, R. M. ERWIN, E. NOL, L. X. PAYNE, AND S. E. SENNER. 1991. Approaches to the conservation of coastal wetlands in the western hemisphere. Wilson Bull. 103(2):218-254. DAVIs, J. H. 1967. General map of natural vegetation of Florida. Institute of Food and Agricultural Sciences, Agriculture Experiment Stations, Circular S-178, Univ. Florida, Gainesville, FL. 1 p. Dupa, M. D. 1987. Floridians and wildlife: sociological implications for wildlife conservation in Florida. Florida Game and Fresh Water Fish Commission, Nongame Wildl. Progr. Tech. Rep. No. 2, Tallahassee, FL. 130 pp. FEDERAL GEOGRAPHIC DATA COMMITTEE. 1992. Application of satellite data for mapping and monitoring wetlands. U.S. Geological Survey, FGDC Wetlands Subcommittee Tech. Rep. 1, Reston, VA. 32 pp. FLORIDA DEPARTMENT OF TRANSPORTATION. 1985. Florida land use, cover and forms classification system. State Topographic Bureau, Thematic Mapping Section, Procedure No. 550-010-001-a, Tallahassee, FL. 79 pp. FLORIDA NATURAL AREAS INVENTORY AND FLORIDA DEPARTMENT OF NATURAL RESOURCES. 1990. Guide to the natural communities of Florida. FNAI and FDNR, Tallahassee, FL. 111 pp. Fox, L., III, K. E. MAYER, AND A. R. FORBES. 1983. Classification of forest resources with Landsat data. J. Forest. 83:283-287. FRANKLIN, J., T. L. LOGAN, C. E. WOODCOCK, AND A. H. STRAHLER. 1986. Coniferous forest classification and inventory using Landsat and Digital Terrain Data. IEEE Transactions on Geoscience and Remote Sensing, GE-24(1):139-149. FRAYER, W. E. AND J. M. HEFNER. 1991. Florida wetlands, status and trends, 1970's to 1980's. U.S. Fish and Wildlife Service, Southeast Region, Atlanta, GA. 31 pp. Harris, L. D., L. D. WHITE, J. E. JOHNSTON, AND D. G. MILCHUNAS. 1974. Impact of forest plantations on North Florida wildlife and habitat. Proc. Annu. Conf. Southeast. Assoc. Game and Fish Comm. 28:659-667. KAUTZ, R. S. 1993. Trends in Florida wildlife habitat 1936-1987. Florida Scient. 56(1):7-24. McComp, W.C.,S. A. BONNEY, R. M. SHEFFIELD, AND N. D. CosT. 1986. Den tree characteristics and abundance in Florida and South Carolina. J. Wildl. Manage. 50(4):584-591. MILLsAP, B. A., J. A. GORE, D. E. RUNDE, AND S. I. CERULEAN. 1990. Setting priorities for the conservation of fish and wildlife species in Florida. Wildl. Monogr. No. 111. 57 pp. MYERS, R. L. 1990. Scrub and high pine. Pp. 150-193. In: MYERS, R. L. AND J. J. EWEL (eds.), Ecosystems of Florida. Univ. Central Florida Press, Orlando, FL. REPENNING, R. W. AND R. F. LABISKy. 1985. Effects of even-age timber management on bird communities of the longleaf pine forest in northern Florida. J. Wildl. Manage. 49(4):1088-1098. SCOTT, j. M., B. CsuTI, J. D. JACOBI, AND J. E. ESTES. 1987. Species richness, a geographic approach to protecting future biological diversity. BioSci. 37(11):782-788. , F. DAVIS, B. CSUTI, R. NOSS, B. BUTTERFIELD, C. GROVES, H. ANDERSON, S. CAICCO, F. D’ERCHIA, T. C. EDWARDS, JR., J. ULLIMAN, AND R. G. WRIGHT. 1993. Gap analysis: a geographic approach to protection of biological diversity. Wildl. Monogr. No. 123. 41 pp. SHELFORD, V. E. 1963. The ecology of North America. Univ. Illinois Press, Urbana, IL. 610 pp. U. S. FISH AND WILDLIFE SERVICE. 1989. Endangered Species Tech. Bull. 14(8):10-11. U. S. SOIL CONSERVATION SERVICE. Undated. 26 ecological communities of Florida. U.S.D.A. Soil Conservation Service, Gainesville, FL. 146 pp. UMBER, R. W. AND. D. HARRIS. 1974. Effects of intensive forestry on succession and wildlife in Florida sandhills. Proc. Ann. Conf. Southeast. Assoc. Game and Fish Comm. 28:686-693. WICKHAM, J. AND R. S. KAUTZ. 1989. Mapping natural plant communities in Florida using Landsat Thematic Mapper satellite data. Pp. 399-405. In: DRISCOLL, D. AND E. O. Box (eds.), Proc. 11th ISB-Congress. SPB Academic Publ., The Hague, Netherlands. WINSBERG, M. D. 1992. Recreation and tourism. Pp. 208-235. In: FERNALD, E. A., and E. D. PURDUM (eds.), Atlas of Florida. Univ. Press of Florida, Gainesville, FL. Florida Scient. 56(3):135-154.1993. Accepted: April 9, 1993. 152 FLORIDA SCIENTIST [VOL 56 GLOSSARY. DEFINITIONS OF THE 22 FLORIDA LAND COVER TYPES MAPPED USING LANDSAT THEMATIC MAPPER IMAGERY. Pinelands—All forested lands dominated by pines (Pinus spp.) except sand pine scrub and sandhill. Pinelands include both natural and planted stands of pines. Typical species: Slash pine (P. elliottii), longleaf pine (P. palustris), sand pine (P.clausa), pond pine (P. serotina), and loblolly pine (P. taeda). Upland hardwood forests—All types of mixed hardwood forests found on xeric, mesic, and hydric sites throughout Florida with the exception of tropical hardwood forests. Although hardwoods predominate, scattered pines may be present. Typical species: American beech (Fagus grandifolia), magnolia (Magnolia grandiflora), live oak (Quercus virginiana), water oak (Q. nigra), sand-live oak (Q. geminata), laurel oak (Q. hemisphaerica), cabbage palm (Sabal palmetto), pignut hickory (Carya glabra), American holly (Ilex opaca), spruce pine (P. glabra), red maple (Acer rubrum), sweetgum (Liquidambar styraciflua), and blue-beech (Carpinus caroliniana). Dry prairie—Broad, largely treeless savannas that occur on nearly level, poorly drained soils in south-central Florida. Native grasses, herbaceous plants, and shrubs dominate, and widely scattered slash pines with up to 10-15% canopy cover may be present. Typical species: saw palmetto, fetterbush (Lyonia ferruginea), gallberry (Ilex glabra), wiregrass (Aristida stricta), and bluestems (Andropogon spp.). Sandhill—Forest communities dominated by longleaf pine, turkey oak (Q. laevis), and wiregrass and occurring on rolling, deep, well-drained sands. Mixed hardwood-pine forests—A forest community in which both hardwoods and pines are abundant but neither is dominant. Although this community type is common on the rolling clay hills of the Panhandle, it also occurs throughout Florida, usually a result of past land use practices. Typical species: water oak, southern red oak (Q. falcata), pignut hickory, magnolia, loblolly pine, slash pine, and shortleaf pine (P. echinata). Sand pine scrub—A forest community dominated by sand pines with an understory of scrub oaks growing on excessively drained, sterile soils. Ground cover typically consists of lichens and litter, and bare areas are common. Typical species: Chapman’s oak (Q. chapmanii), myrtle oak (Q. myrtifolia), sand-live oak, and silk bay (Persea humilis). Xeric oak scrub—A shrub community found on sites similar to sand pine scrub but dominated by any of several species of low growing oaks. Typical species: myrtle oak, Chapman’s oak, scrub oak (Q. inopina), sand-live oak, and Florida rosemary (Ceratiola ericoides). Tropical hardwood hammock—Mixed hardwood forests dominated by tropical species and found only in extreme south Florida where they grow in shallow soils overlying limestone outcrops. Typical species: gumbo limbo (Bursera simaruba), pigeon plum (Coccoloba diversifolia), white stopper (Eugenia axillaris), wild tama- rind (Lysiloma latisiliqua), mastic (Mastichodendron foetidissimum), willow bustic No. 3, 1993] KAUTZ ET AL._VEGETATIVE COVER IN FLORIDA 153 (Bumelia salicifolia), and strangler fig (Ficus aurea). Coastal strand—A community of low-growing vines, grasses, herbaceous plants, and occasional trees or large shrubs growing on sand dunes that parallel the Atlantic Ocean and Gulf of Mexico. Soils are sterile and excessively drained, and the community is subject to wind stress and salt spray. Typical species: sea oats (Uniola paniculata), railroad vine (Ipomoea stolonifera), saw palmetto, and beach cordgrass (Spartina patens). Freshwater marsh and wet prairie—Periodically inundated freshwater wetland communities dominated by a variety of herbaceous plant species growing on hydric sand, clay, marl, and organic soils. This class also includes beds of floating aquatic vegetation along rivers, lakes, and canals. Typical species: pickerelweed (Pontederia lanceolata), arrowhead (Sagittaria latifolia), cattail (Typha spp.), spikerush (Eleocharis spp.), maidencane (Panicum hemitomon), fire flag (Thalia geniculata), sawgrass (Cladium jamaicense), sand cordgrass (Spartina bakeri), white water lily (Nymphaea odorata), yellow lotus (Nelumbo lutea), and water hyacinth (Eichhornia crassipes). Mixed hardwood swamp—Periodically flooded forested wetlands vegetated by a variety of hardwood species, none of which is dominant. Typical species: bald cypress (Taxodium distichum), pond cypress (T. ascendens) water tupelo (Nyssa aquatica), blackgum (N. biflora), swamp ash (Fraxinus spp.), dahoon holly (Ilex cassine), and red maple. Cypress swamp—Regularly inundated forested wetlands dominated by either bald cypress or pond cypress. Cypress swamps typically occur as borders along large rivers, streams, and lakes, or in wet depressions as circular domes or linear strands. Shrub swamp—Wetland communities dominated by dense, low-growing, woody shrubs; saplings; or small trees. Shrub swamps most often are the product of natural or man-made disturbance, such as fire, clear cutting, siltation, or changes in hydroperiod. Typical species: willow (Salix spp.), buttonbush (Cephalanthus occidentalis), black titi (Cliftonia monophylla), little-leaf titi (Cyrilla parviflora), and pond pine (P. serotina). Saplings of common wetland trees, such as red maple, sweebay magnolia (Magnolia virginiana), and blackgum, may also be present. Mangrove swamp—Dense, low-growing forests that occur in the brackish intertidal wetlands of south Florida. Typical species: red mangrove (Rhizophora mangle), black mangrove (Avicennia germinans), white mangrove (Laguncularia racemosa), and buttonwood (Conocarpus erectus). Coastal salt marsh—Herbaceous wetland communities found statewide in brackish intertidal waters of the coastal zone. Typical species: smooth cordgrass (Spartina alterniflora), black needlerush (Juncus roemarianus), salt grass (Distichlis spicata), glasswort (Salicornia virginica), saltwort (Batis maritima), sea oxeye (Borrichia frutescens), and sea lavender (Limonium carolinianum). Bay swamp—Forested wetlands dominated by broadleaved evergreen trees growing in peaty or mucky soils maintained at saturation by groundwater seepage. Typical species: sweetbay magnolia, swamp bay (Persea palustris), and loblolly bay (Gordonia lasianthus), pond cypress, blackgum, slash pine, pond pine, and dahoon holly. Bottomland hardwood forest—Forested wetland community composed of a 154 FLORIDA SCIENTIST [VOL 56 diverse assortment of hydric hardwoods which occur on the rich alluvial soils of the floodplains of several panhandle rivers. Typical species: water hickory (Carya aquatica), overcup oak (Q. lyrata), swamp chestnut oak (Q. michauxii), American sycamore (Platanus occidentalis), Florida elm (Ulmus floridanus), and swamp ash. Grassland and agriculture—Any areas supporting ground cover only, including improved pastures, golf courses, lawns, road shoulders, cemeteries, recently aban- doned cropland, and agricultural fields planted to corn, hay, sugar cane, beans, water melons, tobacco, etc. Barren and urban land—Highly reflective, unvegetated areas of all types, including beaches, active strip mines, bare soil, tilled agricultural fields, roads, bridges, parking lots, rooftops, landfills, airports, industrial parks, etc. Shrub and brush land—Any areas supporting a diverse assortment of early successional stage shrubs and sapling trees, usually as a result of disturbance (e.g., land clearing, clear cutting, fire) 5-10 years prior to the date of the imagery. Typical species: wax myrtle (Myrica cerifera), salt myrtle (Baccharis halimifolia), winged sumac (Rhus copallina), elderberry (Sambucus canadensis), saw palmetto, black- berry (Rubus spp.), gallberry (Ilex glabra), bluestems (Andropogon spp.), and dogfennel (Eupatorium spp.). Exotic plants—Upland or wetland areas dominated by non-native trees and shrubs. Typical species: melaleuca (Melaleuca quinquenervia), Brazilian pepper (Schinus terebinthefolius), Australian pine (Casuarina spp.), and eucalyptus (Euca- lyptus spp.). Open water—Open waters of all types, including the Atlantic Ocean, Gulf of Mexico, bays, lagoons, lakes, rivers, streams, ponds, tidal creeks, and estuaries. / No. 3, 1993] BEEVER ET AL._TREE PLANTING AND PRESERVATION 15: Ol Urban and Community Planning TREE PLANTING AND PRESERVATION PRACTICES AT SINGLE-FAMILY RESIDENCES: POLICY CONSIDERATIONS Lisa B. BEEVER ?, TIM ECKERT °, AND JEFFREY S. MANGUN © Division of Environmental Sciences, Lee County, P.O. Box 398, Ft.Myers, Florida 33902; ®) Lee County Soil and Water Conservation District, P.O.Box 787, Ft. Myers, Florida 33902; ®) Division of Forestry, Ft.Myers District, 10941 S.R.80, Ft. Myers, Florida 33905 ABSTRACT: Trees in residential areas have been related to the public health, safety and welfare. Many local jurisdictions across the country have considered and adopted tree planting requirements for new homes. Unincorporated Lee County possesses no regulation for tree planting for most single-family lots. In anticipation of new regulations, tree planting and preservation practices without regulation were tested. On the average lot, over 15 trees were planted or preserved, and over half of these trees were exotic. Planting practices varied between communities within Lee County. Community variation affected the cost effectiveness of regulatory requirements when compared to a public tree planting program. AS NATIVE plant communities are cleared and houses built, the benefits that trees provide are curtailed. “Where at one time fifty to one hundred trees grew to the acre, we now have three to five houses each with one small tree or palm ... the continued practice of such destruction, without some adequate replacement, goes a long way toward deteriorating the climate of southern Florida” (Sturrock, 1968). Local governments have had a long history of addressing the destruction of trees by encouraging the preservation and planting of trees through regulations. Since the 1880's, city councils (such as Orlando) have protected and planted trees. Such municipalities as Fort Myers encouraged “all owners of lots on any street ... to plant shade and ornamental trees ...” (Anonymous, 1886). This interest in requiring protection and planting of trees on single-family lots has persisted in many jurisdictions. The importance of trees in developed areas has been tied to climate, soil conservation, aesthetics, noise reduction, air quality, energy savings, and wildlife habitat. The direct connection between tree planting on single-family lots and the public safety, health, and welfare has led some local jurisdictions to require the planting/ preservation of trees in association with building permits. For example, the City of Cape Coral, Florida requires two shade trees (or one shade tree and three palms) before Certificates of Occupancy are issued. Collier County requires one native tree for every 3000 square feet (279 square meters) of property, with a minimum of two trees and a maximum of 15 trees per lot (Kirby, 1992). Of states with community forestry assistance programs, 31 of 46 (67%) provide assistance to local communities for development of shade tree ordinances (Casey and Miller, 1988). ' Current address: Charlotte County-Punta Gorda Metropolitan Planning Organization, 28000 Airport Road, A-6, Punta Gorda, FL 33982-2411. 156 FLORIDA SCIENTIST [VOL 56 When Lee County, Florida considered extending its 2 tree per lot requirement from the subdivision ordinance to the building code which applies to previously platted lots, four questions emerged. First, do homeowners plant and preserve trees without regulation? Second, do homeowners plant/preserve the desirable tree species that would be required? Third, do planting trends vary between communi- ties within the county? Finally, would a public tree planting program be more cost effective than a regulatory program? METHODS—Research was conducted in unincorporated Lee County, Florida to address these questions. The first 445 Lee County single-family residence building permits from 1989 were identified by address. The identified building permits constitute 18.5% of the 2405 single-family building permits issued for 1989. The street addresses were found ona map and grouped by general location of the county. The sample of building permits was derived from a short period of time rather than randomized through the year. The chosen method controlled a bias that could result from comparing property owners that had up to an additional year to plant trees. The survey sheet included general location, map grid, street address, tree species, whether the species was native, quantity of the particular species, average height, yard location, lawn species, and condition of lawn. The street address and map grid information were used to organize the route for recording the data at 445 site locations. Two investigators (trained foresters from the Soil and Water Conservation District and from the Division of Forestry) collected the data jointly. Logistics, such as navigation through the neighborhoods, could be handled better through a cooperative effort. In addition, better tree identifi- cation could be achieved by having two knowledgeable individuals conduct the survey. The surveys were conducted from March 5, 1992 to July 8, 1992 and consisted of 12 separate trips. Communities that were surveyed included Boca Grande, Bonita Springs, East Fort Myers, Estero, Fort Myers Beach, Lehigh Acres, North Fort Myers, Pine Island, San Carlos Park, and South Fort Myers. In addition to the incorporated cities (Ft. Myers, Cape Coral, and Sanibel), certain other areas were also not represented because of lower growth rates resulting in no building permits in the sample. These areas include Captiva, Alva, Burnt Store, and Southeast Lee County. No distinctions were made between trees that were preserved during construction and trees that were planted after construction, since such distinctions are normally not included in tree planting regulations. No property was entered to avoid trespass restrictions and to remain consistent. Seven of the 445 surveys were not completed due to limited visibility and other field difficulties. RESULTS—Over 82 identified tree species were planted or preserved by 438 property owners. Several more species were grouped under the category of “un- known exotic species”. Of the 82 identified species, 15 species constituted over 80% of the 6927 trees. These species include slash pine (Pinus elliotti) (1490 trees), queen palm (Arecastrum romanzoffianum) (1035), cabbage palm (Sabal palmetto) (597), cajeput tree (Melaleuca quinquenervia) (328), areca palm (Chrysalidocarpus lutescens) (310), citrus (Citrus spp.) (297), pygmy date palm (Phoenix roebelenii) (293), wax myrtle (Myrica cerifera) (277), live oak (Quercus virginiana) (188), cypress (Taxodium distichum) (164), bottlebrush (Callistemon rigidus) (150), sago (Cycas circinalis) (147), australian pine (Casuarina spp.) (116), washington palm (Washingtonia robusta) (116), and laurel oak (Quercus laurifolia) (98). A total of 6927 trees were planted/preserved at an average of 15.8 (standard deviation (s.d.) 14.8, standard error (s.e.) 0.7) trees per lot. Some 6264 (91%) of the individual trees are species recommended by the South Florida Water Management District in its model landscape code (Schnell, 1987). 3225 (46.6%) are species native to the southeastern United States. Thus, an average of 7.4 (s.d. 11.9, s.e. 0.6) native trees per lot were planted/preserved. 2207 (31.9%) trees are native shade trees which No. 3, 1993] BEEVER ET AL.—TREE PLANTING AND PRESERVATION 157 do not include palms or wax myrtle. An average of 8.6 (s.d. 8.7, s.e. 0.4) exotic trees were planted/preserved per lot. For every native tree, 1.16 exotic trees were planted/preserved. Numbers of exotics versus native trees were plotted. All but 18 cases include a mixture of less than fifty native trees and less than thirty exotic trees. This results in no significant correlation between the number of exotic and native trees planted (correlation=.004). When over thirty exotic trees were planted/preserved, fewer than 10 native trees were planted/preserved (11 cases). When over fifty native trees were planted/preserved fewer than 20 exotic trees was planted (6 cases). Only one property owner did not follow this pattern with 60 native trees and 45 exotic trees. DIscussION—Considering the county as a whole, standards set by the South Florida Water Management District (SFWMD) model landscape code and by the Lee County Development Standards Ordinance (DSO) were met, even though neither code had any legal applicability to most of the lots that were studied. The SFWMD model landscape code (Schnell, 1987) recommends at least 1 tree for every 1500 to 2000 square feet (139 to 186 square meters) of residential lot. Assuming the average lot is a quarter acre, 6 to 8 trees are recommended. Of species the SFWMD recommended, an average of 14.5 (s.d. 14, s.e. 0.7) trees per lot were planted/preserved. The average number of trees planted/preserved by property owners is approximately twice that recommended by SFWMD. The DSO requires two trees per newly subdivided residential lot. Only a small portion of all residential lots are regulated under the DSO, due to the abundance of pre-platted lots in Lee County. Of 438 building permits issued, 20 (4.6%) were for subdivisions created under the DSO. The DSO requires only 75% of trees to be native varieties; 25% of required trees may be palms; and 50% of required trees may be native palms. For the purposes of this analysis, however, the stricter standard of two native shade trees per lot was utilized since “half trees” cannot be assessed on a lot by lot basis. One native palm tree and one native shade tree was also deemed to qualify as meeting the requirements. Without regulation, an average of 5.9 (s.d. 8.9, s.e. 0.4) DSO standard trees per lot were planted/preserved per lot. The average planting/preservation of DSO standard trees alone exceeded the DSO standard by two and a half times. Actual planting/preservation of trees, on average, was double that recom- mended in the SFWMD model landscape code and DSO standards. These codes would not be applied by averaging but on a lot by lot basis and as a minimum standard. Of 438 lots, 299 (68.3%) met SFWMD average recommendations and 283 (64.6%) met the DSO standards. Community variations emerged. Communities where more than half of prop- erty owners met the DSO standards for planted/preserved trees included Bonita Springs (75%, n=72), South Fort Myers (72%, n=88), East Fort Myers (70%, n=14), San Carlos (69%, n=55), North Fort Myers (66%, n=64), and Pine Island (64%, n=25). Other communities included Lehigh Acres (47%, n=79), Fort Myers Beach (40%, n=5) and Boca Grande (33%, n=6) (See Fig. 1). Chi Square analysis of community variation is significant at the 0.05 level (X?=20.75). 158 FLORIDA SCIENTIST [VOL 56 U.S S.A.31 q BOCA GRANDE ea PINE ISLAND N.FT.MYERS LEGEND FT.MYERS BEACH Mmm 80-100% 50-60% 70-80% [KX 40-50% Lee County, 60-70% [KN 30-40% BONITA SPRINGS Florida Fic. 1. Percentage of lots meeting proposed DSO standards by community. Although average planting/preservation trends were similar when comparing DSO and SFWMD standards, community variation existed. In all communities, more than half of the property owners planted/preserved trees according to the SFWMD standards, including Boca Grande (100%), South Fort Myers (91%), East Fort Myers (86%), Pine Island (76%), Bonita Springs (74%), Fort Myers Beach (60%), Estero (60%), Lehigh Acres (57%), San Carlos Park (55%) and North Fort Myers (52%) (See Figure 2). Chi Square analysis of community variation is signifi- cant at the .001 level (X?=46.01). It is of interest that the ranking of communities is different for meeting DSO standards as compared to SFWMD standards. The DSO standards require fewer total trees but more native trees than the SFWMD standards. Thus, communities that tend to plant/preserve many exotic trees will rank higher than communities that plant/preserve only a few native trees when considering the SFWMD standards. Communities that plant/preserve only a few native trees will rank higher than the communities that plant/preserve many exotic trees when considering the DSO standards. Over half (52.7%) of property owners met both DSO and SFWMD standards. One-fifth (19.6%) met neither standard. Another 16% met only the SFWMD standards and 11.6% met only the DSO standard. Chi Square analysis of community variation is significant at the .001 level (X?=96.02). The correlations between community and the number of exotic trees (correla- tion=.143) and native trees (correlation=-.1169) were both significant at the .01 No. 3, 1993] BEEVER ET AL—TREE PLANTING AND PRESERVATION 159 U.S.41 1-75 $.R.31 ] BOCA GRANDE PINE_ ISLAND FT.MYERS BEACH LEGEND Mm 80-100% 50-60% NORTH E=] 70-80% RX 40-50% Lee County, 60-70% KN) 30-40% BONITA SPRINGS Florida Fic. 2. Percentage of lots meeting proposed SFWMD standards by community. level. However, the correlation between community and total number of trees was not significant (correlation=-.010). Communities which averaged more native trees than exotic trees included Bonita Springs, East Fort Myers, North Fort Myers, and Pine Island. It is interesting to note that Fort Myers Beach averaged more than six times more exotic trees than native trees. Other communities which averaged more exotic trees than native trees include Boca Grande, Estero, Lehigh Acres, San Carlos Park, and South Fort Myers. As a group, unincorporated Lee County single-family homeowners planted/ preserved at least twice the desirable trees recommended by the SFWMD model landscape code and specified in the Lee County DSO. However, such potential regulations relate to minimum community standards. In both cases, 32-35% of property owners did not meet recommended standards. Governments seeking to reduce this tree deficit between planting practices and planting standards have several available options. The most efficient and effective method should be sought. Alternative planting programs include tree planting requirements for new single-family houses, public property/easement planting programs, private property planting programs, and building voucher systems. Several governments in Florida require a certain number of trees to be planted/preserved before certificates of occupancy are issued for new single-family houses (e.g., City of Cape Coral and Collier County). Some governments fund tree planting on public property and easements (e.g. Lee County and City of Cape Coral). Local governments may negotiate with neighborhood associations to issue contracts to plant trees along 160 FLORIDA SCIENTIST [VOL 56 neighborhood streets where public easement widths are too narrow to accommodate tree planting. Upon issuance of a Certificate of Occupancy, local governments could issue a voucher that allows the new home builder to redeem for a certain number of trees at participating nurseries. The nurseries then send the voucher back to the agency for payment. The cost of the trees could be paid by an additional assessment on the permit fee, but at a reduced rate. The rate can be reduced because the value of the introducing new home owners to area nurseries and the purchase of a large total number of trees. The cost effectiveness of regulation was compared to a public tree planting program. Both methods have associated costs. Regulation requires public invest- ment in permitting, inspection, compliance, and enforcement. Public planting includes the cost of the trees, installation, maintenance, and administration of the contracts. The cost comparison changes depending on the planting practices of and the tree planting standards adopted by the local community. As an example, 75% of Bonita Springs property owners planted/preserved at least 2 desirable trees as specified by the DSO. The other 25% planted 1 desirable tree. The permit review and compliance of the 72 houses would have yielded an additional 18 trees. Given an estimated cost of $50 per building permit, the 18 trees would have cost $3600 or $200 per installed tree. The estimate of $50 per building permit was derived, in part, from the Collier County program which began in January 1992. A total of 1334 permits were inspected from January to October 1992. Kirby (1992) estimated that 2 positions at a base rate of $16,000 annually could complete the permitting and inspections. Given an additional 40% for benefits and $15,000 annually for overhead, the per permit assessment is approximately $50. Applying this method to all of the test communities resulted in the following per installed tree cost to implement DSO specifications: Bonita Springs ($200), South Fort Myers ($176), Estero ($167), San Carlos Park ($146), East Fort Myers ($140), North Fort Myers ($114), Pine Island ($96), Lehigh Acres ($84), Boca Grande ($75), Fort Myers Beach ($84). In considering all of Lee County, an additional 172 trees would have been planted in the sample, or $127 per installed tree, to meet DSO specifications. By applying SFWMD recommendations, the following per installed tree cost for each test community was as follows: Boca Grande ($300), South Fort Myers ($275), East Fort Myers ($233), Bonita Springs ($128), North Fort Myers ($80), Estero ($79), Pine Island ($52), San Carlos Park ($40), Lehigh Acres ($33), Fort Myers Beach ($25). For all of Lee County, the per installed tree cost of implementing SFWMD recommendations is $66. The cost of simply contracting the installation of trees varies between $35 to $75 per installed tree, averaging about $50 per installed tree (Joyce, 1992). This estimate is based on recent experience in Lee County and on a minimum of 6 ft (1.8 m) tall trees of a grade recommended in the Development Standards Ordinance. This figure does not include the governmental personnel cost of administering the bid process and contracts, which can vary greatly depending on the volume of trees per contract and planting plan costs. A rough estimate of $35 to $75 for administration costs results in a total of $75 to $150 for each installed tree. No. 3, 1993] BEEVER ET AL.—TREE PLANTING AND PRESERVATION 161 The cost of a tree planting program is comparable to permitting costs to implement DSO standards. The cost of implementing SFWMD standards through regulation is more cost effective than a public planting program. The per tree cost of regulation becomes lower as the number of required trees increases. The stricter the tree-planting regulation, the more cost effective it becomes. As decision-makers consider new single-family property tree planting require- ments, several issues should be addressed. First, what are the tree standards of the community? The higher this standard, the more cost effective a regulatory program will be. Second, do citizens demand a minimum tree standard regarding neighbor- hood aesthetics (similar to lot mowing requirements)? If citizens are most concerned about consistent planting standards in their neighborhood and by their neighbors, a regulatory approach would be easier to manage. Third, what are the most cost effective and publicly acceptable approaches to planting trees? Other alternatives, such as a public tree planting program, neighborhood grant program or builder voucher system, may be more cost effective or acceptable by the citizens. Fourth, is there demand for more trees on lots that are already built? Regulations that are retroactive are more difficult to enforce than regulations that are implemented through an established permitting process, such as the Certificate of Occupancy process. Because of the community variation in planting/preservation practices of single- family property owners, a review of local planting practices is recommended for communities considering single-family tree planting regulations. CONCLUSIONS—Sturrock (1968) was concerned that the suburban landscape in South Florida does not include as many trees as the native plant communities. Based on the investigation in Lee County, this basic premise remains correct. However, Sturrock’s (1968) sparse picture of one small tree or palm per house is either pessimistic or out-dated. Instead, the standard lot has between 1 and 30 trees, averaging 15 trees. A different dilemma emerges from this investigation. Lee County is becoming an alien terrain. More than half of the trees planted and preserved on single-family lots are exotic. Over two (exotic) queen palms are planted per lot, and they are the second most common trees species. A higher percentage of households had a queen palm planted than any other tree species. Within the local environmental planning profession, the queen palm has come to represent a blight on the suburban landscape, replacing more valuable native trees. Melaleuca, an invasive exotic pest, is the fourth most common tree within the sample. (Most melaleucas were preserved where most queen palms were planted). For over a century, local governments in Florida have addressed the planting and preservation of trees. There are various methods to achieve additional trees in the community. Common methods include regulation and planting programs. The choice in methods should be based on the difference between planting practices and planting standards. As this difference becomes greater, regulation becomes more cost effective. 162 FLORIDA SCIENTIST [VOL 56 LITERATURE CITED ANONYMOUS. 1886. Ordinance for Improvement of the Streets. New-Press. Fort Myers, FL. December a p..2. CASEY, Cy. AND R. W. MILLER. 1988. State government involvement in community forestry: a survey. J. Arboricult. 14:141-144. Kirby, M. 1992. Collier County Government, 2800 North Horseshoe Drive, Naples, Florida. Pers. Commun. Joyce, R. K. 1992. Lee County Government, Natural Resources Management Division, P. O. Box 398, Ft. Myers, Florida. Pers. Commun. SCHNELL, J. F. Jr. 1987. Final report on development of a model landscape code for south Florida. South Florida Water Management District. SOUTH FLORIDA WATER MANAGEMENT DISTRICT (SFWMD)). Xeriscape plant guide II, 48pp. STURROCK, D. 1968. Trees for Southern Florida: a plea for afforestation. Central and Southern Florida Flood Control District. West Palm Beach, FL. Florida Scient. 56(3):155-162.1993 Accepted: April 23, 1993. OUTSTANDING STUDENT PAPER AWARDS (Cont) PHYSICAL AND SPACE SCIENCES Charles R. Bogusch, Department of Electrical Engineering, University of Central Florida, “Femtosecond Pulse Generation in a Ti:Sapphire Laser at Extended Wavelengths.” Graduate Award. SCIENCE TEACHING Nanette Madigan, Millar Wilson Laboratory for Chemical Research, Jacksonville University, “Collection of Batteries for Proper Disposal from Five Schools in Duval County.” Undergraduate Co- Award. J. Matthew Gibson, College of Natural Sciences, Eckerd College, “Earth Science Teacher Enhancement Programs: A Student’s Perspective and Input.” Undergraduate Co-Award. URBAN AND REGIONAL PLANNING Jennifer Horowitz, Millar Wilson Laboratory for Chemical Research, Jacksonville University, “Review of the Historical Record of Air Quality Research in Jacksonville, Florida.” Undergraduate Award. SPECIAL AWARDS AMERICAN ASSOCIATION FOR THE ADVANCEMENT OF SCIENCE AWARD UNDERGRADUATE AWARDEES: Tatiana Wellens, Department of Chemistry, University of South Florida, “Micellar Electrokinetic Capillary Chromatographic Separation of Cationic Porphyrins.” R. Scot Duncan, Department of Biology, Eckerd College, “Predation on Supratidal Mangrove Littorinids by Puffer Fishes (Tetraodontidae) in the Gulf of Nicoya, Costa Rica. CENTRAL FLORIDA CHAPTER, THE EXPLORERS CLUB AWARD Penny E. Cople, Biology Department, Stetson University, The Effects of Satellite Nests on Predation in Artificial Nests of Pseudemys floridana peninsularis. Co-Award. D.L. Lenard, Jr., Department of Biology, University of Central Florida, “Limited Breeding Distribution of the Prothonotary Warbler (Protonotaria citrea) Within the Wekiva Basin.” Co-Award. ENVIRONMENTAL CHEMISTRY AND CHEMICAL SCIENCES AWARD Charles D. Norris, Department of Chemistry, University of South Florida, “Application of Supported Chelating Agents for Extraction of Cadmium.” Graduate Co-Award. Elsie Gross, Department of Chemistry, University of South Florida, “Isolation and Characteriza- . tion of a Hydroxamate Siderophore Produced by Lyngbya majuscula.” Graduate Co-Award. Cont. on page 184 | No. 3, 1993] MORSE ET AL.—HONEY BEES IN FLORIDA 163 Biological Sciences SELECTION OF NEST CAVITY VOLUME AND ENTRANCE SIZE BY HONEY BEES IN FLORIDA Rocer A. Morse"), JaMEs N. Layne”, P. Kirk VIssCHER®, AND FRANCIS RATNIEKS™) “Department of Entomology, Cornell University, Ithaca, New York 14853; Archbold Biological Station, P.O. Box 2057, Lake Placid, Florida 33852; Department of Biology, University of California, Berkeley, California 94720; ‘Department of Entomology, University of California, Riverside, California 92507 Asstract: In south-central Florida, honey bees (Apis mellifera L.) of European descent occupied nest boxes with round entrance holes of about 7.9 cm? (3.2 cm diameter), but rejected those with areas of about 31.7 cm (6.4 cm diameter). They also accepted boxes with nest volumes ranging from 10.2 to 13.2 L, considerably smaller than any recorded natural nest or human-made nest box occupied by honey bee swarms in northern regions. The strong behavioral difference between populations in semitropical and cool-temperate regions may reflect genetic differences or non genetic based flexibility in response to differing environmental conditions. SEELEY and Morse (1976) collected and dissected 21 honey bee nests in trees in the vicinity of Ithaca, New York. Most nest cavities were 30 to 60 L in volume; the entrances had an area of 10 to 40 cm (total entrance area per nest, sometimes the sum of several entrances). Both the cavity volumes and the nest entrance sizes were smaller, on average, than those in the multi-story, human-made hives used in commercial beekeeping. The volume of one standard 10-frame Langstroth hive body, rarely used by itself, is about 42 L. When used in conjunction with a standard bottomboard and cover, about 4 L are added to the volume, and the nest entrance is about 71 cm. In tests in the Ithaca area investigating the nest cavity characteristics preferred by honey bee swarms, bees chose 40-L cavities in preference to cavities of 10 or 100 L; 100-L nests were sometimes accepted, but no swarm ever occupied a bee hive with a volume of 10 L (Seeley, 1977; Seeley and Morse, 1978). When presented a choice between bee hives with entrances 3.2 and 8.9 cm in diameter (7.9 and 62.1 cm in area), the bees always chose the one with the smaller entrance; no swarm occupied a bait hive with an 8.9 cm diameter hole. Nest entrance shape did not influence the choice. In nature, and especially in old buildings, we have occasionally found a honey bee nest ina larger cavity or with a larger nest entrance, which is always at the bottoms of the nest. The acceptance of such extremes may reflect a shortage of more suitable nest sites. This paper reports observations on honey bee utilization of nest boxes erected for vertebrate animals in south-central Florida that further define the preferred nest entrance size for honey bees and reveal a marked geographic difference in the nest cavity volume accepted by swarms. 164 FLORIDA SCIENTIST [VOL 56 METHODS AND MATERIALS—The study area was a 420-ha tract on the Archbold Biological Station, 12 km south of Lake Placid, Highlands County, Florida. The bees in the study area probably swarmed from commercial apiaries in surrounding citrus groves, or at least from colonies descended from these. They were therefore a mix of European races, predominantly A. mellifera ligustica. The lowest winter temperatures in the study area occur in January. The 40-year average daily temperature in January is 15.9°C, and the mean minimum daily temperature is 8.5°C. Cold spells with nighttime temperatures below freezing are usually of short duration. Thus flight by honey bees is generally possible throughout the winter. From November 1972 to July 1984, 125 nest boxes erected on trees were checked monthly or bimonthly for occupancy by vertebrates and invertebrates. A few boxes were checked at irregular intervals following this period. The boxes were originally established in order to facilitate studies of the biology of gray squirrels (Sciurus caroliniensis) and other cavity-nesting vertebrates. The nest boxes were of the design described by Barkalow and Soots (1965), with inside dimensions of 18 cm by 18 cm by 50 cm and a volume of 15 L. A round entrance hole was located 5 cm from the top of the box. Boxes were located in seven of the major habitat types on the Station described by Abrahamson and co-workers (1984). Ninety boxes were placed along 4-wheel drive roads or firelanes and 25 on permanent small mammal trapping grids in the following natural habitat types: southern ridge sandhill, sand pine scrub, scrubby flatwoods, flatwoods, and bayhead. The remaining boxes were in man-modified habitats, including five boxes around the edge of an old field and five in a park-like area of mature trees, shrubs, and lawns. Boxes were attached to the tree trunks 1.5-5.3 m above ground, some in shade and some in sunny locations. Originally all of the nest boxes had entrance holes 6.4 cm (31 cm? area) in diameter, but in March 1979 the entrances of 58 of the 115 nest boxes in natural habitats were reduced to 3.2 cm diameter (7.9 cm? area) in order to exclude use by gray squirrels and encourage use by flying squirrels (Glaucomys volans). Alternate nest boxes along roads and firelanes and on grids were selected for entrance size reduction, giving an essentially random distribution of large- and small-holed boxes with respect to location, habitat, and other environmental factors. Resu_ts—In the period November 1972 to March 1979, when only large-hole boxes were available, no honey bee swarms occupied the nest boxes. However, from April 1979 to July 1984, 26 swarms were recorded in small-hole boxes, while the large-hole boxes remained unused by honey bees. The first occupied boxes were found within a month of reduction of entrance size. An additional four colonies were found in small-hole nest boxes in January 1985. In none of the nest boxes did vertebrates and bees cohabit. Most (89%) of the honey bee colonies in nest boxes first appeared during spring and summer. The number of new colonies appearing in different months (percentages of total in parentheses) were: January, 1 (3.8%); March 4 (15.4%); April, 2 (7.7%); May, 13 (50.0%); July, 4 (15.4%); September, 1 (3.8%); November, 1 (3.8%). The frequency of new colonies in April was low because nest boxes were not usually checked that month. Minimum periods that individual colonies persisted in nest boxes ranged from 1 to 33 months, with a mean of 5.2 months. Length of occupancy of nest boxes by the same colony (based on observed presence in consecutive checks), and the number of colonies known to occupy a box for the given period, were as follows: 1 month, 8; 3 months, 5; 4 months, 1; 5 months, 1; 7 months, 1; 9 months, 1; 11 months, 3; 15 months, 1; 18 months, 1; 33 months, 1. As boxes were routinely checked every other month during most of the study, maximum duration of occupancy could be as much as 2 months more than the periods given above. Removal of some colonies from nest boxes to standard beehives also contributed to an underestimate of colony longevity. Eight (31%) of the 26 colonies in nest boxes survived one winter (recorded in November, January, and March), and one colony persisted through three winters. In No. 3, 1993] MORSE ET AL.—HONEY BEES IN FLORIDA 165 addition, the four nest box swarms found in January 1985 were still active the following March. Although nest boxes were distributed more or less uniformly throughout the study area, most of the boxes occupied by bees were in the southern two-thirds of the area. This difference apparently did not reflect the distribution of bees on the study area, as there were apiaries along both the northern and southern boundaries, and citrus groves in which bees regularly foraged bordered the eastern side. A more likely explanation is a difference in the general structure of the vegetation between the southern and northern parts of the study tract, the southern area being more open and the northern part more heavily forested. The mean +SD and range (in parentheses) of the volumes of 10 of the 30 nest boxes containing honey bee colonies were 12.0£0.4 L (10.2-13.2). The variation in volume reflected the amount of squirrel nest material present in the boxes. One colony in a box with 13.0-L volume had built 6-8 L of comb below the nest box, a behavior only rarely observed in colonies in nest boxes or human-constructed hives. The mean +SD and range of entrance diameters of the sample of occupied nest boxes were 3.3+0.2 cm (3.2-3.8). The slight variation in nest entrances was the result of gnawing by squirrels. Discussion—The observations show unequivocably that the bees preferred small-hole boxes and indicate that the threshold of their ability to discriminate between hole sizes lies somewhere between 7.9 and 31 cm’. This marked responsive- ness to entrance area is an indication that a small entrance may be significant in the survival of the colony. Seeley and Morse (1978) suggested that smaller entrances are easier to defend against the entry of predators and robbing honey bees and probably also help in maintenance of the colony microclimate. The volumes of the boxes occupied, with a mean of 12 Landa minimum of 10.2 L, are some of the smallest reported for nests of European honey bees. Of 47 natural nests examined by Seeley and Morse (1976) and Seeley (1977) in the Ithaca, New York, region, all but 3 had volumes exceeding 20 L, the exceptions being 18.2, 18.8, and 19.6 L. Seeley (1977) also found that swarms selected against 10-L cavities when these were paired with 25-L cavities. Rinderer and co-workers (1982) found that European bees in Louisiana, given a choice among 5-, 10-, and 20-L cavities, selected against the 5-L cavity, but colonized 10- and 20-L cavities approximately equally. In contrast, when given the same choice, Africanized bees in Venezuela selected against both the 5- and 10-L cavities (Rinderer et al., 1982). However, Winston and co- workers (1983) reported colonies of Africanized honey bees in South America using nest cavities of less than 9 L. Seeley (1985) considered rejection by swarms of cavities below a certain minimum volume to be an adaptation to insure a nest capacity sufficient to accomodate the amount of honey needed to sustain the colony throughout the winter. In a study at Ithaca, New York, during the winter of 1985-86, (F. Ratnieks, unpubl. data), all of 18 colonies housed in 17-L boxes perished and only 4 of 18 in 27-L boxes survived. At Guelph, Ontario, some 110 km north and 240 km west of Ithaca, 10 colonies in 21-L boxes all died during the winter, while most colonies in 166 FLORIDA SCIENTIST [VOL 56 larger cavities survived (Morales-Soto, 1986). Studies of winter survival in New York by Seeley and Visscher (1985) showed that colonies in 40 L nests with less than about 20 kg of honey starved in winter. That amount of honey alone occupies about 13 L and would take up considerably more volume as stored by bees. Clearly, the minimum viable colony volume in northern areas is well in excess of 10 L. In contrast, however, the Florida colonies readily occupied nest cavities with volumes ranging from 10.2 to 13.0 L and regularly survived over winter. The observed low threshold of acceptable nest volume in Florida may have a genetic basis, reflecting weaker selection against colonies choosing small nests in southern populations as a result of milder winters. However, circumstantial evidence does not support this interpretation. Highlands County is near the center of the citrus-growing area in south-central Florida; each year in this area there is an influx of large numbers of honey bee colonies, moved in by beekeepers to take advantage of the citrus honey flow. Many of these colonies swarm, and during the citrus bloom swarms can frequently be found hanging in trees near apiaries. Several beekeepers who keep bees in the area are migratory, moving their bees between Florida and various northern states each year. Because of this movement and the structure of the queen rearing and beekeeping industry in the U.S., there is presumably considerable gene flow between honey bee populations of central Florida and other regions of the U.S. In actuality, a locally adapted honey bee population is more likely to occur in the Ithaca region, where feral colonies probably outnumber domestic ones and com- mercial and migratory beekeeping is on a much smaller scale than in Florida. The alternative explanation of the geographic difference in nest-volume thresh- olds—a response to different environmental conditions by genetically similar popu- lations—seems more likely. Such variation has been shown for aggressiveness in Africanized bees (Brandenburgo et al., 1977) and polygyny in Vespula (Ross and Visscher, 1983). There are many natural cavities in the forests in the Ithaca region, whereas in the central Florida study area there are fewer large trees and the available natural cavities, mainly woodpecker holes, tend to be relatively small. The scarcity of larger cavities coupled with milder winter weather which permits foraging and places less demand on stored honey for colony survival may thus lead to use of cavities with lower volumes. Higher winter bee population density and greater competition for nest cavities from other animals in central Florida compared to northern localities also may contribute to the use of smaller cavities in the south. It should be emphasized that our results show only that the Florida colonies use and are able to persist in low volume nest sites and not that they “prefer” them, as the bees were offered no choice between large and small nest cavities as in the case of previous studies (Seeley and Morse, 1978; Rinderer et al., 1982). The results of this study are striking in demonstrating the fine degree of nest- entrance-size discrimination of which bees are capable and the trenchant difference between semitropical and cool-temperate honey bees in nest size selection. The mechanisms enabling bees to precisely measure nest entrance areas and the minimum difference they can detect, as well as the question of the extent to which genetic and environmental factors contribute to the geographic difference in nest size selection await further research. No. 3, 1993] MORSE ET AL.—HONEY BEES IN FLORIDA 167 LITERATURE CITED ABRAHAMSON, W.G., A.F. JOHNSON, J.N. LayNE, AND P.A. PERoni. 1984. Vegetation of the Archbold Biological Station, Florida: An example of the southern Lake Wales Ridge. Florida Scient. 47:209-250. BarKALow, F.S., JR. AND R.F. Soots. 1965. An improved gray squirrel nest box for ecological and management studies. J. Wildl. Manage. 29:679-684. BRANDENBURGO, M.A.M., L.S. GONGALVES, AND W.E. Kerr. 1977. Effects of Brazilian climatic conditions upon the aggressiveness of Africanized colonies of honeybees. Final Rept. U.S.D.A. Proj. 12. 14. 7001-362, 35 pp. Morates-Soro, G. 1986. Effects of cavity size on demography of unmanaged colonies of honey bees (Apis mellifera L.) M.S. thesis, Univ. of Guelph, Ontario, Canada. RINDERER, T.E., K.W. TucKER, AND A.M. Co.ins. 1982. Nest cavity selection by swarms of European and Africanized honeybees. J. Apic. Res. 21:98-103. Ross, K.G. anD P.K. VisscHER. 1983. Reproductive plasticity in yellowjacket wasps: a polygynous, perennial colony of Vespula maculifrons. Psyche 90:179-191. SEELEY, T.D. 1977. Measurement of nest cavity volume by the honey bee (Apis mellifera). Behav. Ecol. and Sociobiol. 2:201-227. . 1985. Honeybee Ecology: A Study of Adaptation in Social Life. Princeton Univ. Press, Princeton, NJ. AND R.A. Morse. 1976. The nest of the honey bee (Apis mellifera L.) Insectes Sociaux 23:495-512. AND R.A. Morse. 1978. Nest site selection by the honey bee, Apis mellifera. Insectes Sociaux 25:323-337. AND P.K. VissCHER. 1985. Survival of honey bees in cold climates: the critical timing of colony growth and reproduction. Ecol. Entomol. 10:81-88. Winston, M.L.,O.R. TayLor, AND G.W. Oris. 1983. Some differences between temperate European and tropical African and South American honeybees. Bee World 64:12-21. Florida Scient. 56(3):163-167.1993. Accepted: April 28, 1993. 168 FLORIDA SCIENTIST [VOL 56 Environmental Sciences MOVEMENT OF FLURIDONE IN THE UPPER ST. JOHNS RIVER, FLORIDA ANDREW J. LESLIE’, DON C. SCHMITZ", ROBERT L. KIPKER"’, AND DAVID L. GIRARDIN® “Bureau of Aquatic Plant Management, Florida Department of Environmental Protection, 3917 Commonwealth Boulevard, Tallahassee, Florida 32399 St. Johns River Water Management District, P. O. Box 1429, Palatka, Florida 32078 ABSTRACT: The aquatic herbicide fluridone (Sonar) applications to control the exotic aquatic macrophyte Hydrilla verticillata (L. f.) Royle were monitored in 1985 and 1987 in several shallow lakes located within the upper St. Johns River, Florida. Herbicide residues within the monitored treatment plots in 1985 and 1987, peaked above 200 g/L within six hours after application but could no longer be detected (detection limit 1 g/L) by 36 to 48 hours. Seven days after treatments in 1985 and 1987, fluridone was detected moving out of the treated lakes; concentrations ranged from 11 to 26 ug/L in 1985 and I to 9 ug/L in 1987. By 14 days after treatment in 1985, fluridone (7 ug/L) was detected at the potable water intake located 8 km downstream of the treatment areas; finished water contained no detectable residues. Fluridone residues (1-4 ug/L) were detected entering the potable water intake and in finished water (1-2 ug/L) in 1987. Concentrations were well below the U. S. Environmental Protection Agency tolerance of 150 ug/L. Fluridone reduced the area covered by hydrilla 40% to 90% in treated lakes in 1985; the plant reduction lasted four months to one year. Fluridone residues were detected for 50 days in 1985 and for 28 days in 1987 within the river system downstream of the treatment plots. Increased rainfall-induced water flow occurred in 1987 versus 1985, and most likely accounted for the greater persistence of fluridone in 1985. Little aquatic vegetation reduction could be demonstrated in 1987, probably due to insufficient concentration/contact time. FLURIDONE, l-methyl-3phenyl-5-[3-(trifluromethyl)phenyl]-4(1H)-pyridinone, a herbicide marketed by Dow-Elanco Products Company, Indianapolis, Indiana, under the trade name Sonar, was applied to three shallow lakes near the headwaters of the St. Johns River, Florida, to control the exotic aquatic macrophyte Hydrilla verticillata (L. f.) Royle. The herbicide fluridone was chosen because it is relatively slow acting (minimizing possible oxygen depletion problems from quick vegetation kills that can occur with contact herbicides), is reported to have low mammalian toxicity (Parka et al., 1978; McCowen et al., 1979; Lilly, undated), and is highly effective on hydrilla (Parka et al., 1978; Grant et al., 1979; Sanders and Theriot, 1979; Schmitz et al., 1987). Fluridone has been reported to disperse out of target areas (West et al., 1983; Schmitz et al., 1987). This characteristic, combined with the high susceptibility of hydrilla to this herbicide, has been exploited to reduce large populations of hydrilla with relatively small applications (Hinkle, 1985). However, there was concern that fluridone applied in the upper St. Johns lakes would travel downstream to the potable water intake (8 km downstream of Lake Sawgrass, Fig. 1) and possibly exceed the U. S. Environmental Protection Agency tolerance of 150 pg/L for residues in potable water. The objectives of this study were to monitor residues of fluridone during actual No. 3, 1993] LESLIE ET AL.—MOVEMENT OF FLURIDONE IN ST. JOHNS 169 a | LAKE WASHINGTON LAKE 8 SAWGRASS jy pI ) : Zee fad 4 10 which were collected by filtration, dried at 170°C and weighed. Results of the gravimetric analysis were compared with results from the analysis of the solution. Effect of ammonium sulfate concentration—F our concentrations of aqueous ammonium sulfate (0, 0.001, 0.01, 0.1, and 1.0 M) were used for leaching solutions. This study was limited to room temperature (24°C) and three different sets of pebbles (IMC 35-60 mesh, FIPR 35-60 mesh, and FIPR -230 mesh). Effect of temperature—Five temperatures (8°, 15°, 24°, 35°, and 45°C) were used with 0.1 M aqueous ammonia to determine the optimum leaching temperature. A hot-water or ice-water bath was used to maintain the temperature, which was constant within +1°C. RESULTS—Choice of pebbles—Three samples of pebbles were used in this study, and their properties are listed (Table 1), but owing to the abundance of the FIPR-pebble sample, that material was used in the more systematic and detailed studies. The 35-60 mesh fraction of both IMC and FIPR pebbles was used because it was the fraction that had the lowest magnesium content and the greatest promise of success. The IMC pebbles were light grey, and the FIPR pebbles were yellowish, and these color differences persisted in leaching solutions (the latter pebbles produced a more yellowish solution). All sets of pebbles are low-grade phosphate pebbles because of the high magnesium and comparatively low phosphate concen- tration (Table 1). Effects of ammonium sulfate concentration—F or all experiments, the pH of the leaching liquor increased with time during the one-hour study, rapidly at first, then approached a maximum. Analyses of the leaching liquor for calcium and magnesium (by AAS) indicate that the pebbles reacted with the mixture of carbonic acid and ammonium sulfate, and this was verified by measuring the weight of the carbonate precipitate when the leaching liquor was separated and treated with concentrated (15 M) aqueous ammonia. The available data suggest that there was a marked increase in calcium or magnesium leaching over the first ten minutes, followed by a slower period in which there were noticeable differences in amount of either leached. The pattern held regardless of the concentration of ammonium sulfate. Data are available for three different sets of pebbles with 3-5 different ammonium TaB_e 1. Chemical composition of experimental samples (%) sample % Composition Phosphate pebbles P.O, % Ca % Mg % Wt. % Total IMC 35-60 mesh 26.68 30.57 0.98 26.50 FIPR 35-60 mesh 28.86 ole ek? 25.45 FIPR -230 mesh 18.40 29.09 4.07 12.52 188 FLORIDA SCIENTIST [VOL 56 sulfate concentrations (Pan, 1992). The data for one set are presented (Fig. 1), showing the percent calcium or magnesium removed in 60 minutes as a function of ammonium sulfate concentration in the leaching solution. These data indicate that the calcium and magnesium were effectively removed from FIPR pebbles, and that the amount removed at 60 minutes increased with increasing ammonium sulfate concentration. On the other hand, other data (Table 2), considering the calcium magnesium weight ratios, indicate that the material removed was not pure dolomite. Dolomite has a calcium/magnesium ratio of 1.67 (wt/wt), and the weight ratios indicate the removal of calcium carbonate. Effect of temperature on leaching—Data are available for calcium and magne- sium concentrations in the leaching liquor (carbonic acid 0.1 M ammonium sulfate) as a function of time and temperature (7-45°C) for the FIPR 35-60 mesh phosphate pebbles. The amounts of either cation leached increased with time and with leaching temperature (Table 3). Raising the temperature favored magnesium removal over calcium, if the results are expressed as percentage of the element in the pebbles (Table 3). DIsCcUssION—Several methods have been used to leach magnesium, in addi- tion to those noted earlier, and five examples are given here. Aqueous sulfur dioxide was effective in leaching magnesium according to Hanson and co-workers (1985), who reported that flow rate of sulfur dioxide and —2e'+-% Mg in the sample —~—_- % Ca in the sample 6.00 5.00 4.00 3.00 O.O 0.001 0.01 O.1 1.0 Ammonium sulfate concentration (M) Fic. 1. Percent magnesium and calcium removed in 60 minutes of leaching for FIPR -230 mesh phosphate pebble at 24°C No. 3, 1993] PAN ET AL.—DOLOMITE EXTRACTION TABLE 2. Weight ratio of calcium/magnesium in the leaching liquor at 24°C Leaching system IMC 35-60 mesh pebbles CO,/H,0 CO,ANH,),SO, CO,ANH,),SO, 0.01 M 10M FIPR 35-60 mesh pebbles CO,/H,O CO,ANH,),SO, CO,ANH,),SO, CO,ANH,),SO, CO,ANH,),SO, 0.001 M 0.01 M 0.1M 10M 10 12.7 13.7 14.2 3.90 3.51 4.30 3.94 5.96 Leaching time (minute) 20 30 40 50 16.5 18.7 17.3 19.0 19.3 19.6 2.98 2.42 2.50 2.36 2.95 2.28 2.23 2.24 4.16 3.56 3.98 3.73 3.34 3.20 3.09 5.30 4.55 4.25 4.02 60 18.2 20.1 20.0 2.82 2.44 3.40 3.06 3.86 189 TABLE 3. Magnesium concentration and calcium concentration (mg/L) in liquor for FIPR 35-60 mesh phosphate pebbles leached by carbon dioxide-ammonium sulfate (0.1 M) buffer Leaching temperature,°C 0 A. Magnesium ge 15°C 24°C 30°C 45°C B. Calcium rC 15°C 24°C 35°C 45°C 10 15.16 17.74 19.36 34.92 36.34 76.70 82.10 76.31 138.8 133.4 Leaching time (minute) 20 30 40 50 16.33 22.45 26.92 30.00 23.16 27.40 34.46 37.52 19.36 33.74 40.35 45.00 48.11 63.40 59.88 62.00 44.81 59.88 63.40 71.18 84.8 103.7 114.5 114.5 111.8 122.6 136.1 136.1 72.24 113.0 129.2 141.5 158.7 179.3 173.9 163.1 136.1 182.0 155.9 206.3 60 36.3 39.6 52.8 67.4 91.1 141 149 161 179 237 190 FLORIDA SCIENTIST [VOL 56 leach time were the major factors in the amount of magnesium leached. These workers also found that some magnesium was so intimately associated with the phosphate mineral that it could not be removed prior to acidulation with sulfuric acid. Although there may be encapsulation, we believe that much of the magnesium is associated with distinct dolomitic phases that are not encapsulated. Crushing (as opposed to grinding) the pebbles causes the weakest mineral component to break up preferentially, and this is reflected in the distribution of magnesium and phosphate. Specifically, the data (Table 4) show that magnesium is concentrated in the smallest- mesh fraction and phosphorus is concentrated in the larger fraction. Benson (1982) quantitatively leached magnesium in brucite from serpentines by means of aqueous ammonium chloride and carbonic acid. Magnesia was selectively leached from low-grade magnesites by aqueous carbon dioxide (Sheila et al., 1991). Magnesium dissolution was fast, reached a maximum in 60 min. or less, and over 99% of the magnesium was leached under optimum conditions. Finally, Evans and St. Clair (1949) found that solutions of magnesium bicarbonate having metastable concentrations more than twice equilib- rium concentrations could be consistently prepared by leaching magnesium hydrox- ide, magnesium oxide, and calcined magnesite or dolomite with aqueous carbon dioxide. Metastable solutions could not be prepared from magnesium carbonate or when the leaching temperature was much greater than 30°C. Benson (1982) noted that formation of a metastable solution is necessary to be able to achieve desirable levels of aqueous magnesium. The possibility of leaching magnesium by aqueous carbon dioxide is attractive because carbon dioxide is a relatively cheap commodity, and easy to obtain. Given the concern with the so-called greenhouse effect, we may become increasingly concerned with useful disposal of carbon dioxide from stack gas (cf. Nair and Somasundaran, 1987). This study focused on the use of aqueous carbon dioxide and ammonium sulfate buffer mixture as a means of leaching magnesium from phosphate pebble. The pH TABLE 4. Percentage (wt/wt) distribution of crushed IMC pebbles Screen size Weight, g Wt % Median particle Meshes diameter, (mm) -18/+35 702.8 30.1 0.750 -3'7+60 618.5 26.5 0.375 -60/+ 120 369.9 15.7 0.187 -120/+230 LOIS 9.2 0.094 -230 453.4 19.4 <0.063 No. 3, 1993] PAN ET AL.—DOLOMITE EXTRACTION 19] of the buffer mixture is stabilized at 4.5. This pH is suitable for the formation of magnesium bicarbonate, favorable for leaching (Eqn. 1). (NH,),SO, + CO, + MgCO,*CaCO/Ca,(PO,), > NH, Tt + Mg(HCO,), + MgSO, + CaSO, 1 + CaCO, 1 +Ca,(PO,),1 (1) Ammonium sulfate in solution can be recovered by precipitation of the magne- sium salts as the carbonate and hydroxide with the addition of aqueous ammonia (Eqn. 2). MgSO, + Mg(HCO,), + NH, + H,O > MgCoO, J + Mg(OH), J + CO, T + H,O + (NH,),SO, (2) The results of this study indicated that the ammonium sulfate-carbonic acid buffer system was effective in leaching magnesium and calcium from solution. The leaching system appeared more favorable for FIPR pebbles than IMC pebbles, possibly because of the size distribution, since smaller sizes (FIPR -230 mesh) of pebbles showed greater leaching, especially removal of calcium. No detrimental effects of the leaching system on the pebbles was observed. For example, dilute aqueous acid (hydrochloric or sulfuric) can effect leaching of magnesium, but it will also cause a loss of phosphate. And the thought of having a source of hydrogen chloride or hydrochloric acid in the presence of stainless steel is stunning. In contrast, the buffer system we used did not reduce the phosphate level of the pebbles, nor would it attack stainless steel. On the other hand, our studies indicated that there was room for improvement in leaching effectiveness. About 2% of the magnesium was removed from IMC 35- 60 mesh pebbles, about 15% of the magnesium was removed from FIPR 35-60 mesh pebbles, and about 8% of the magnesium from FIPR -230 mesh. Thus while the buffer system is attractive, additional modifications and studies seem appropriate, including for example, pressure leaching, leaching calcined pebbles, and multiple-stage leaching. The need for an effective leaching system is evident, and the goal of finding an optimum system is a worthy one. Accordingly, an alternative approach that used CO,-based leaching of dolomitic phosphate pebble was developed. The approach consists of a three-loop process with a leaching loop, a liquor recovery regeneration loop, and a product loop resulting in removal of magnesium ion (and other nuisance ions) and production of magnesium carbonate, calcium carbonate and other useful products (Benson and Martin, 1993). ACKNOWLEDGMENTS—We gratefully acknowledge the financial support of the Florida Institute of Phosphate Research through grant No. 90-01-092. We are grateful for the helpful comments of Mr. G. Michael Lloyd, Jr., Project Coordinator, FIPR. We appreciated the phosphate samples provided by FIPR and IMC Corporation. We gratefully acknowledge technical assistance provided by Mr. Charles D. Norris with the atomic absorption spectrometry measurements. Dr. P. M. Dooris served as Consulting Editor. 192 FLORIDA SCIENTIST [VOL 56 LITERATURE CITED BECKER, P. 1989. Phosphates and Phosphoric Acid. Raw Materials, Technology, and Economics of the Wet Process. Marcel Dekker, Inc., New York, N.Y. BENSON, R. F. 1982. Method for leaching magnesium hydroxide from magnesium hydroxide containing composition. U. S. Patent 4,335,083. AND C. MARCOZZI. 1993. Thermogravimetric analyses of several carbonate minerals of magnesium and calcium—magnesite, dolomite, calcite, and hydromagnesite. Paper presented at the annual meeting of the Florida Sections of the American Chemical Society, May 6; Abstract in FLACS, 46(2):12. AND D. F. MARTIN. 1993. Patent pending. EVANS, R. L. AND H. W. ST. CLAIR. 1949. Carbonation of aqueous suspensions containing magnesium oxides or hydroxides. Ind. Engn. Chem. 41:2814-2817. HANSEN, J. P., B. E. DAVIS, AND T. O. LLEWELLYN. 1985. Removal of Magnesia from Dolomitic Southern Florida Phosphate Concentrates by SO, Leaching. Rpt. Investig. 895. Bureau of Mines, Washington, D.C. HARGIS, L. G. 1988. Laboratory Manual—Analytical Chemistry. Prentice-Hall, Englewood Cliffs, N.J. LLEWELLYN, T. O., B. E. DAvISs, G. V. SULLIVAN, AND J. P. HANSEN. 1981. Beneficiation of high- magnesium phosphate from southern Florida. Rpt. of Investig. 8609. Bureau of Mines, Wash- ington, D.C. Morais, S. B. 1989. Removal of cation impurities from inorganic solutions. U. S. Patent 4,861,490. NAIR, K. V. AND P. SOMASUNDARAN. 1987. Beneficiation of dolomite apatite. A new hydrometallurgical method. Proc.-Int. Symp. Benefic. Agglom. 2:45-55. PAN, Y. 1992. Dolomite extraction by phosphate pebble from carbon dioxide-ammonium sulfate buffer. Master thesis. Univ. South Florida, Tampa, FL. RUSHTON, W. E. AND W. R. ERICKSON. 1982. Magnesium oxide removal from phosphoric acid. Proc. Ann. Meet. Fertilizer Ind. Round Table 32:202-214. SHEILA, D., C. SANKARAN, AND P. R. KHANGAONKAR. 1991. Studies on the extraction of magnesia from low grade magnesites by carbon dioxide pressure leaching of hydrated magnesia. Minerals Eng. 4:79-88. WHITNEY, S. G. AND W. R. ERICKSON. 1982. Method of selectively removing adsorbed calcium and magnesium from cation exchange resins. U. S. Patent 4,363,880. AND H. J. KLEIss. 1991. Method of removing and crystallizing cation resin absorbed calcium and magnesium. U. S. Patent 5,021,216. Florida Scient. 56(3):185-192.1993. Accepted: May 18, 1993. INSTRUCTIONS TO AUTHORS Individuals who publish in the Florida Scientist must be active members in the Florida Academy of Sciences. Submit a typewritten original and two copies of the text, illustrations, and tables. All typewrit- ten material—including the abstract, literature citations, footnotes, tables, and figure legends— shall be double-spaced. Use one side of 81/2 x 11 inch (211/2 cm X 28 cm) good quality bond paper for the original; the copy may be xeroxed. Margins should be at least 3 cm all around. Number the pages through the Literature Cited section. Avoid footnotes and do not use mimeo, slick, erasable, or ruled paper. Use metric units for all measurements. Assistance with production costs will be negotiated directly with authors of pages which exceed 10 printed pages. Excess over 10 pages by an author ina single volume will result in charges for additional pages. AppkRESS follows the author’s address. ABSTRACT—AII manuscripts shall have a short, concise, single-paragraphed abstract. The abstract follows immediately the author’s address. ACKNOWLEDGMENTS are given in the body of the text preceding immediately the Literature Cited section. LITERATURE CITED section follows the text. Double-space every line and follow the format in the current issue. MANUSCRIPTS of 5 or less, double-spaced typewritten pages should conform to the short arti- cle protocol. See recent issue for proper format of both long and short articles. TABLES shall be typed on separate sheets of paper, double-spaced throughout. Each table shall contain a short heading. Do not use vertical rulings. Maximum character width, including spaces, of tables is 98.0. This includes a minimum space of 5 characters between columns. Tables are charged to authors at $45.00 per page or fraction. ILLUSTRATIONS—AIl drawings shall be done in good quality India ink, on good board or draft- ing paper. Letter by using a lettering guide or equivalent. Typewritten letters on illustrations are unacceptable. Drawings and photographs should be large enough to allow 1/3 to 1/2 reduction in size. Photographs shall be glossy prints of good contrast. Whenever possible, mount photographs in lots size. The author’s name and figure number should be penciled lightly on the back of each figure. Figure legends must be listed on a separate page and not on the drawing or photograph. The leg- end must be double-spaced. Illustrations are charged to authors at $25.00 per page or fraction. We ask authors to prepare accepted manuscripts in personal computer formats. Please do not submit a manuscript on disk. Once reviewers have recommended a manuscript be accept- ed, you will be given directions for submitting the final manuscript on disk. PROOF must be returned promptly. Notification of address change and proofreading are the author’s responsibility. Alterations after the type has been set will be charged to the author. REPRINTS may be ordered from the printer on forms provided when the corrected proofs are returned to the Editor. FLORIDA ACADEMY OF SCIENCES INSTITUTIONAL MEMBERS FoR 1993 Archbold Biological Station St. John’s River Community College, Palatka Brevard Community College, Florida Advanced St. Petersburg (Florida) Public Library Technology Center Library, Palm Bay U.S. Department of Commerce, National The British Library, Boston Spa, Weatherby, UK Marine Fisheries (Panama City) Charles Courtney / EPC Hillsborough County U.S. Environmental Protection Agency Library, Florida Institute of Technology Gulf Breeze, FL Harbor Branch Oceanographic Institution, Inc. University of Central Florida, Division of Lake-Sumter Community College Sponsored Research Newfound Harbor Marine Institute University of South Florida CORPORATE MEMBERS FOR 1993 Mote Marine Laboratory, Sarasota Membership applications, subscriptions, renewals, changes of address, and orders for back numbers should be addressed to the Executive Secretary, Florida Academy of Sciences, P.O. Box 033012, Indialantic, FL 32903-0012 PUBLICATIONS FOR SALE by the Florida Academy of Sciences Complete sets. Broken sets. Individual numbers. Immediate delivery. A few numbers reprinted by photo-offset. All prices strictly net. Special bargain: $500 for nearly complete sets. Prices quoted include domestic postage. PROCEEDINGS OF THE FLORIDA ACADEMY OF SCIENCES (1936-1944) Volumes 1-7—$10.00 per volume; single numbers $3.50 QUARTERLY JOURNAL OF THE FLORIDA ACADEMIC OF SCIENCE (1945-1972) Volumes 8-35—$10.00 per volume; single numbers $3.50 FLORIDA SCIENTIST (1973-75) Volumes 36-38—$10.00 per volume; single issues $3.50 except for symposium numbers priced separately. Volumes 39-47 —$13.00 per volume; single numbers $4.25 except for symposium numbers priced separately. Volumes 48 onward—$20.00/volume; single numbers $5.00 except for symposium numbers priced separately. Florida’s Estuaries—Management or Mismanagement?—Academy Symposium FLORIDA SCIENTIST 37(4)—$5.00 Land Spreading of Secondary Effluent—Academy Symposium FLORIDA SCIENTIST 38(4)—$5.00 Solar Energy—Academy Symposium FLORIDA SCIENTIST 39(3)—$5.00 (includes do-it-yourself instructions) Anthropology—Academy Symposium FLORIDA SCIENTIST 43(3)—$7.50 Shark Biology—Academy Symposium FLORIDA SCIENTIST 45(1)—$8.00 Future of the Indian River System—Academy Symposium FLORIDA SCIENTIST 46(3/4)—$ 15.00 Second Indian River Research Symposium FLORIDA SCIENTIST 53(3)—$15.00 Individual orders should be sent with payment. A statement will be sent in re- sponse to a bona fide purchase order over $10.00 from a recognized institution. Address all orders to: The Florida Academy of Sciences, Inc. P.O. Box 033012 Indialantic, Florida 32903-0012 Telephone: (407) 723-6835 ISSN: 0098-4590 / -Florida Scientist ~~ Volume 56 Fall, 1993 Num er 4. CONTENTS Hurricane Andrew and the Colonization of Five Invading Species in South TIPDTTIG'S, a soe tesSisatia eae sett ee arene TNE Ea Ose ONE AT 0) O eeRae Walter E. Meshaka, Jr. 193 Synthesis of a Sterically Challenging Bis-(B - diketone) .........0.0.c eee Venkatraj V. Narayanan 202 Geographic Distribution of the Striped Mullet (Mugil cephalus Linnaeus) muthne Atlantic and Eastern Pacific OCEANS «..........:..06.s.sseeeccsesesecovotvoteeee Carter R. Gilbert 204 Field and Laboratory Experiments on the Consumption of Mangrove Leaf Litter by the Macrodetritivore Melampus coffeus L. (Gastropoda:pulmonata) .............+. WRN ee Wines RIN kh a aa to C. Edward Proffitt, Kevin M.Johns, C. Bruce Cochrane, Donna J. Devlin, Theresa A. Reynolds, Deborah L. Payne, Sean Jeppesen, David W. Peel, and Dwane D. Linden 21 LL PST ee Cr ee ee Charles D. Norris 222 Fruit Removal and Interplant Distance in the Persimmon: Diospyros BIEN ere sa 1s dese soa gh ada con chloe Ne Sseusarn gules ven nal uSyhetoucwdanindowunstiacovveets Cris Cristoffer 223 Dimethyl Sulfoxide Catalyzed Racemization of Aspartic Acid .........0...0 Gregory P. Cusano, Mirtha Chavez, Diego Torres, and George H. Fisher 226 Additions to the Herpetofauna of Egmont Key, Hillsborough County, pels a EN he se llc Jeu ddboliouuluanenmieuubaa idsdaacis Lora L. Smith, Richard Franz, and C. Kenneth Dodd, Jr. 231 Tidal and Wind-driven Transport Between Indian River and Mosquito Pe RISEN SoA GL eL Pe 98 as ere ee eA heey ey ag baica 2Abe sesnaatsraucndand so cined Ned P. Smith 235 Partial Degradation of Arochlor 1242 PCBs by Alcaligenes Bacteria............ Stojan Stojkovski, Bruce D. James, and Robert J.Magee 247 ee er AE Et Me Oe le cn iow tidy Soames detoane William H. Taft 250 ete hI ee ls Ges eaisceioden dnabievnenssiniateentiersnonscnstoes 251 MMSE CAI TIE OL OVI WETS oe ca0182 2n5yc2 on ninsnedencctorencnnorésrseenensushdnoeusetsnevansnssee 951 VOLES hc ee eee 253 QUARTERLY JOURNAL OF THE FLORIDA ACADEMY OF SCIENCES FLORIDA SCIENTIST QUARTERLY JOURNAL OF THE FLORIDA ACADEMY OF SCIENCES Copyright© by the Florida Academy of Sciences, Inc. 1993 Editor Dr. DEAN MARTIN Co-Editor: Mrs. BARBARA B. MARTIN Institute for Environment Studies Department of Chemistry University of South Florida 4202 East Fowler Avenue Tampa, Florida 33620-5250 THE FLORIDA SCIENTIST is published quarterly by the Florida Academy of Sciences, Inc., a non-profit scientific and educational association. Membership is open to indi- viduals or institutions interested in supporting science in its broadest sense. Applica- tions may be obtained from the Executive Secretary. Both individual and institutional members receive a subscription to the FLORIDA SCIENTIST. Direct subscription is avail- able at $40.00 per calendar year. Original articles containing new knowledge, or new interpretation of knowledge, are welcomed in any field of Science as represented by the sections of the Academy, viz., Biological Sciences, Conservation, Earth and Planetary Sciences, Medical Sciences, Physical Sciences, Science Teaching, and Social Sciences. Also, contributions will be considered which present new applications of scientific knowledge to practical problems within fields of interest to the Academy. Articles must not duplicate in any substantial way material that is published elsewhere. Contributions are accepted only from members of the Academy and so papers submitted by non-members will be accepted only after the authors join the Academy. Instructions for preparation of manuscripts are inside the back cover. Officers for 1992-93 FLORIDA ACADEMY OF SCIENCES Founded 1936 President: Dr. Patrick J. GLEASON Treasurer: Dr. MARVIN IvEY 1131 North Parkway 14452 Hillview Drive Lake Worth, Florida 33460 Largo, Florida 34644 President-Elect: Dr. Patricia M. Dooris Executive Secretary: Mrs. BETTY PREECE Institute for Environmental Studies P.O. Box 033012 Department of Chemistry Indialantic, Florida 32903-0012 University of South Florida Tel: (407)723-6835 Tampa, FL 33620-5250 Program Chair: Dr. DEL DELUMYEA Secretary: Ms. MARCELA GUITERREZ-MAYKA Millar Wilson Laboratory for Chemical Research 701 E. River Dr. Jacksonville University Temple Terrace, FL 33617 Jacksonville, Florida 32211 Published by the Florida Academy of Sciences, Inc. P.O. Box 033012 Indialantic, Florida 32903-0012 Printing by C & D Printing Company, St. Petersburg, FL Florida Scientist QUARTERLY JOURNAL OF THE FLORIDA ACADEMY OF SCIENCES DEAN F. Martin, Editor BARBARA B. Martin, Co-Editor ‘ Volume 56 Fall, 1993 Number 4 B iological Sciences HURRICANE ANDREW AND THE COLONIZATION OF FIVE INVADING SPECIES IN SOUTH FLORIDA WALTER E. MEsuaka, JR.’ Department of Biological Sciences, Florida International University, University Park, Miami, Florida 33199 ABSTRACT: The responses of three amphibian and two reptilian species to Hurricane Andrew were recorded in an urban setting. Immediate responses to the hurricane boded well for species with a high colonizing potential. Short-term effects of the hurricane were beneficial for all but Anolis equestris. Anurans especially benefitted by stimulation to breed and by the creation of suitable egglaying sites. Cleanup activities by humans have altered long-term effects of the hurricane and minimized short-term benefits associated with the newly created habitats and/or breeding sites made available to most of these colonizing species following the hurricane. Cleanup activities notwithstanding, responses confirmed expectations of successful colonizers historically tied to hurricanes. IN THE early morning hours of 24 August 1992, Hurricane Andrew crossed southeast Florida, bringing with it tornados, sustained winds exceeding 280 kph, and gusts of 344 kph. Moving rapidly at 27 kph, only 7.6-15.3 cm of water was recorded in Miami (National Weather Service). The time from 0300-0630 hrs was for most life forms in south Dade County a life-threatening experience. By 0700 hrs there were still strong wind gusts and a constant drizzle. It was at this time that most humans could step outside and look upon the now unrecognizable subdivision. By noon the sun was out, and the air was breezy. Electricity was restored 7 September 1992. From the morning following the hurricane, I began recording post-hurricane responses of Bufo marinus, Eleutherodactylus planirostris, Osteopilus septentrionalis, Anolis equestris, and A. sagrei, all species known to inhabit the subdivision. The purpose of recording these observations was to answer three questions. Firstly, what were the direct effects of the hurricane on these species? Secondly, how did their responses reflect colonization ability? Thirdly, what effect would cleanup activities have on the responses of these species to the hurricane? Present address: Archbold Biological Station P.O. Box 2057 Lake Placid, Florida 33852 USA 194 FLORIDA SCIENTIST [VOL 56 METHODS—The Kendall Point subdivision, built in 1968, is located on Kendall Drive (S.W. 88 St.) between S.W. 82 and 87 Ave. in Miami, Dade Co., Florida. Snapper Creek canal borders the north side of most of the subdivision. The remainder of the subdivision is walled-in. The northeast end of Kendall Point is bordered by a new housing development on what was previously the grounds of Arvida Nurseries Inc. The King’s Creek development lies across the canal, to the north of Kendall Point and the new subdivision. There is an artificial lily pond at King’s Creek (2.5 X 24 m). The water depth fluctuates between a maximum of 20 cm to a minimum of 5 cm. A swimming pool at a residence near Cutler Bay served as a collection site for tadpoles 17 Sept. 1992. Activity or abundance was recorded for Bufo marinus, Eleutherodactylus planirostris, and Osteopilus septentrionalis, Anolis equestris, and A. sagrei. Data were collected 24 August- 4 September 1992, 14-18 September 1992, and 26 & 27 September 1992. A census consisted of a 0.5 hr walk through Kendall Point beginning at 1500 hrs. A yard count was an instant count of all lizards observed at once on the east side of my house at 1100 hrs. Ambient temperature (C) was recorded in the shade prior to each census. No lizards were captured for this study. Collections of tadpoles were made with a dipnet. Specimens were fixed and stored in formalin. Mean values of snout-vent lengths (SVL) are in mm and are followed by 2 standard deviations. Tadpoles and metamorphic individuals are deposited in the United States National Museum (USNM). RESULTS AND DiscussION—The three amphibian and two reptile species are treated in the following species accounts. Bufo marinus—The marine toad is a neotropical species intentionally intro- duced into south Florida (Neill, 1957). In the Kendall area they were very abundant at least throughout the 1970s and very early 1980s (pers. obs.). Their numbers have since declined, but they are still commonly seen from late May through October. Bufo marinus is known to spawn in Snapper Creek canal and in the artificial lily pond at the King’s Creek development. The night following the hurricane, 24 August 1992, B. marinus responded by calling vociferously from the canal two blocks from my house. Open windows and no traffic may have aided in hearing this toad, which in the past could not be heard from my house. The ambient temp. was 29 C. Calling was noticeably weaker the following night. No calling has been heard since 25 August 1992. Because of the existing curfew, I could not determine the extent of egglaying that occurred during the dates recorded for calling, and daytime search was too hazardous. Subsequent search indicated that egglaying had indeed taken place soon after the hurricane in puddles and in Snapper Creek canal. On 18 September 1992, I collected 159 tadpoles from puddles in a housing development between Kendall Point and Snapper Creek canal. Water temperature was 35 C at 1400 hrs, and most tadpoles showed varying degrees of hindlimb development. I visited the puddles again 1330 hrs on 26 September 1992. I found some puddles had evaporated, and water temperature of remaining puddles was 37 C (water depth was 15 cm). No tadpoles could be found, but four recent metamorphic individuals (X= 9.75 mm SVL + 0.191) were collected on wet sand in partial shade within 0.5 m of these puddles. Because larval transformation may occur within four weeks (Pemberton, 1934), evidence suggests that an immediate and successful breeding response had taken place in these newly created, ephemeral breeding sites. Snapper Creek canal is a traditional spawning site for this species and was visited only once, 18 September 1992. Tadpoles, some of similar size to those in the puddles, were observed, but not collected. At this site, as with the puddles, egglaying probably took place during the nights I heard calling after the hurricane. / No. 4, 1993] MESHAKA—HURRICANE ANDREW 195 Activity at the lily pond could not be monitored until after it had been cleared of debris 8-11 September 1992. Courtship activities soon followed the cleanup at this site. At 2230 hrs, following a heavy rain on the night of 14 September 1992, three pairs of amplecting B. marinus were observed in the lily pond. Ambient temperature was 26 C and water temperature was 27 C. On the following day, long strings of eggs were found in less than 15 cm of water. Free-swimming tadpoles in the pond were found during my next visit on 21 September 1992. An immediate courtship response in B. marinus followed Hurricane Andrew. The use of even a few newly created spawning sites following a tropical storm could prove beneficial in the expansion and establishment ofa highly fecund species. In the case of B. marinus, such a response may have enhanced rapid establishment and expansion following their deliberate introduction into Florida. In this urban environ- ment, debris was quickly removed. Recruitment of individuals could have been supressed by the loss of new microhabitat afforded by the hurricane. Eleutherodactylus planirostris—The greenhouse frog is a small, semi-fossorial leptodactylid first recorded in Florida by Cope (1875). Specimens may be found in natural habitats (Duellman and Schwartz, 1958) and under trash (Wilson and Porras, 1983). The tadpole stage of this species is passed in eggs laid in moist leaf litter. Eleutherodactylus planirostris was stimulated to courtship activities before and after the storm, and hurricane-related mortality was observed. At 2400 hrs, three hours prior to the storm, E. planirostris was heard calling from potted plants stored in my house. During the hurricane, two windows were broken on the windward side of my house. The wind blew under a closed door and down a hall. There, I collected a freshly killed specimen blown in from the storm at 0500 hrs. Immediately following the storm at 0700 hrs, individuals were heard calling from drains in the patio, where they have been commonly encountered in the past. That evening, accompanying the din of B. marinus and Osteopilus septentrionalis, E. planirostris was again heard calling from the patio drains. Owing to its semi-fossorial nature, many individuals survived the hurricane and immediately responded with courtship activities. Since Hurricane Andrew, I have found individuals under stacks of plywood and have heard males calling from underneath brushpiles. A combination of courtship activities and creation of habitat and egglaying sites could serve as ideal conditions for population increase and expansion. Following Hurricane Hugo, survivorship of adult E. coqui in Puerto Rico was within the normal range, but was low in the juvenile class (Woolbright, 1991). Rapid increase in frog density following the hurricane was explained by reduction in the density of invertebrate predators and the creation of new retreat sites (Woolbright, 1991). Left unaltered, a similar effect could take place with E. planirostris in south Florida. In an urban setting, however, long-term population increase associated with the creation of habitat has been minimized as debris is removed. Osteopilus septentrionalis—The Cuban treefrog is a large hylid first reported in Key West by Barbour (1931) and later reported from mainland Florida by Schwartz (1952). They are ephemeral site breeders most commonly encountered in mesophytic forests and many kinds of disturbed habitats (Carr, 1940; Duellman and 196 FLORIDA SCIENTIST [VOL 56 Schwartz, 1958; Wilson and Porras, 1983). In Everglades National Park, I have found tadpoles in pineland depressions, solution holes in hammocks, and prairie fringes. Flooded fields, puddles, cisterns, and birdbaths also are used successfully. Courtship activities accompanied Hurricane Andrew and continued for several nights afterwards. Further north of the eyewall of the hurricane in the vicinity of Miami International Airport, O. septentrionalis was heard calling prior to the hurricane beginning at 1100 hrs. Calling was explosive in roughly 10-30 minute intervals and lasted 3-4 minutes at a time. Frogs could be heard in the high winds during the storm until 0600 hrs (Gregory Small, 1992). At King’s Creek, O. septentrionalis is known to lay its eggs in the lily pond throughout the year, but individuals are only occasionally seen in Kendall Point. During the nights of 24-26 August 1992, O. septentrionalis called loudly in Kendall Point including my backyard. Calling intensity steadily decreased and by 4 Septem- ber 1992 no calling was heard. As with B. marinus, I was unable to search »°: cr day for them, but it seemed likely that egglaying occurred in traditional and in many newly created breeding sites, natural or otherwise, following the hurricane. On the rainy evening of 14 September 1992, calling males were observed at the King’s Creek lily pond. No eggs were found the following day, but the presence of newly hatched tadpoles suggested egglaying had taken place two to three days prior. Following the hurricane, swimming pools were used as breeding sites. From a sample of 16 pool owners in my neighborhood on 18 September 1992, 31% (5) of the pools contained tadpoles and/or adults since Hurricane Andrew. All five houses bordered Snapper Creek canal. Description of tadpoles and adults confirmed the _ species identity. On that day, no pools examined had tadpoles. I was informed by residents that their pools were drained or treated with chlorine shortly after the hurricane. From one swimming pool near Cutler Bay, I documented successful reproduc- tion following the hurricane. On 17 September 1992, tadpoles were collected from a pool unattended since Hurricane Andrew struck 24 days earlier. One scoop of a dipnet captured 75 tadpoles of which 47% possessed four legs and varying degrees of tail resorption, signifying metamorphosis (X= 14.1 mm SVL + 0.502). Tadpoles raised at 25 C and fed Tetra Reptomin sticks ad libitum metamorphosed in 27 days (X= 15.5 mm SVL + 0.781: N= 24). Two-legged individuals appeared in 18 days. Tadpoles raised outside in water that ranged from 24.5 C at night to 34 C during the day and fed Tetra Reptomin sticks ad libitum metamorphosed in 27 days (X= 14.8 mm SVL + 0.872: N= 91). Thus, it was likely that the tadpoles I collected came from eggs laid the night of the hurricane or a day or two thereafter. The Cutler Bay swimming pool was treated with chlorine before the hurricane, but it was quickly diluted. Because of storm surge, the pool filled to the top, the water was murky, and dead fish littered the lawn. Egglaying apparently took place in water containing very diluted chlorine. Egglaying also took place in large puddles soon after the hurricane. Twenty-five days later, on 18 September 1992, tadpoles were collected from puddles in the new but incomplete housing development behind Kendall Point. One of the 91 tadpoles / No. 4, 1993] MESHAKA—HURRICANE ANDREW 197 collected possessed erupted forelegs and many showed varying degrees of hindlimb development, indicating reproduction following the hurricane. I visited these puddles again at 1330 hrs, 26 September 1992. In 15 cm water, the temperature was 35 C. No living tadpoles were found, and eggs laid within a week after the storm could have metamorphosed in this time. The presence of a few rotting carcasses, however, suggested that at least some tadpoles perished possibly from heat stress. Osteopilus septentrionalis most benefitted from this disturbance. Their re- sponse was, like B. marinus, immediate and profound. Unlike B. marinus, the active response of O. septentrionalis was longer lasting and took place in a greater diversity of newly created egglaying sites. Darlington (1938), Simpson (1966), and Pregill (1981) argue that over-water transport in vegetation, not land bridges as argued by Barbour (1914), is the mechanism of dispersal in West Indies fauna. Creation of suitable egglaying sites (natural or otherwise) and stimulation to breed in O. septentrionalis were associated with Hurricane Andrew. Over-water transport by hurricanes during the breeding season could aid O. septentrionalis, not only in their dispersal, but in their establish- ment as well. This scenario would be most pronounced in areas of Dade County where cleanup activity by humans has been minimal. Anolis equestris—The Cuban knight anole is a large tree trunk-canopy dwelling omnivore (Williams, 1969) known to south Florida since 1955 where it was purposely introduced (Neill, 1957). Wilson and Porras (1983) stated that they are most commonly observed during the hottest days of the year. In south Miami, I have observed copulations in the wild from June through August. The response of A. equestris to Hurricane Andrew was evident in the behavior exhibited by individuals eight hours prior to when the storm struck land. At 1900 hrs Martin Tracey (1992) observed three adult specimens on the pavement near the biology building of Florida International University (F.I.U.). The lizards were crawling about slowly and were unresponsive to his presence. With no difficulty one specimen was captured by hand and held captive until the hurricane passed. When released the following morning, the lizard ran quickly to a nearby tree. Anolis equestris is strongly arboreal; when startled it retreats to the canopy (Williams, 1969). I have only twice seen A. equestris voluntarily come to the ground. On both occasions, it was to dash quickly to an adjacent tree. A search by these lizards for better refuge at F’.I.U. was plausible even if identifying the stimulus was difficult. Trees on the side of the biology building where the lizards were found were not large and lacked the cover necessary to protect lizards from strong winds. Stimulation to seek refuge occurred at late in the day when lizards are seldom active, but as the barometric pressure was decreasing. Ihave found no dead specimens and believe few, if any, were killed by the storm itself. Most large trees were toppled, but remained intact. If a search for cover occurred before the advent of this hurricane, A. equestris, in palm axils and cavities, could better survive the high winds of a tropical storm. Indeed, refuge in vegetation has been implicated as a vehicle for over-water transport in the colonization of initial West Indies stock groups (Darlington, 1938; Pregill, 1981), and possibly in the 198 FLORIDA SCIENTIST [VOL 56 TABLE 1. Post-hurricane yard counts and censuses of Anolis equestris in urban south Florida. Date | YardCount Census +» Aur Dempa yu June-Aug. 1992 4 5 = 1 4 pe 1 3 a 0 ca fs 24 Aug. 1992 = i 28 25 Aug. 1992 = 0 29 26 Aug. 1992 0 0 29 13 Sept. 1992 0 2 31 16 Sept. 1992 0 3 31 26 Sept. 1992 0 3 32 27 Sept. 1992 = 1 31 geographic expansion of some extant forms. In Kendall Point, the immediate post-hurricane response of adult A. equestris was to bask. No immature lizards were ever seen in the censuses, and so the effect of the hurricane on this size-class remained unknown. A census on 24 August 1992, just hours after the hurricane, revealed more lizards than prior and subsequent census values (Table 1.). All seven individuals were found basking in full sun on the many large, fallen, and partially defoliated trees. Most allowed approach within one meter. The absence of lizards on 25 and 26 August 1992 was probably in response to disturbance caused by cleanup activities. The immediate basking response of many A. equestris following the hurricane underscored their ability to survive and to respond quickly after the partial devastation of their habitat. A search for cover before the hurricane, little mortality, and an immediate basking response within hours after the hurricane inferred adaptations to a phenom- enon to which A. equestris has been subjected during their evolution. This scenario contrasted dramatically with their response to frost in south Florida. Unaccustomed to such cold conditions, many dead specimens fell from trees following a 1982 frost in south Florida (Wilson and Porras, 1983; pers. obs.). Although fewer lizards were found in post-hurricane censuses, too few data were collected (Table 1.) to warrant statistical comparisons of pre and post-hurricane numbers. The same may be said for yard counts of A. equestris, however, the reduction in number of lizards in the yard following the hurricane was more striking (Table 1.). Prior to the hurricane, when lizards were routinely seen, three Ficus, one No. 4, 1993] MESHAKA—HURRICANE ANDREW 199 coconut (Cocos nucifera), and one Queensland umbrella tree (Brassaia actinophylla) served as perching sites. Only the coconut tree was undamaged. The largest Ficus toppled, and the other two smaller Ficus suffered branch fall. The umbrella tree, site of both single sightings, was defoliated. No lizards were observed in the yard following the hurricane until December when daily sightings of an adult A. equestris resumed from the refoliated umbrella tree. In my yard, tree loss or damage was natural. At other residences, some remaining large trees have been pruned or removed. The result, natural or otherwise, was temporary and permanent decrease in available cover for this species. Measuring population changes of unmarked lizards following the hurricane may be difficult. Lizards were conspicuous on defoliated trees making visible individuals that otherwise might not have been seen. The result could have been an overestimation of lizards following Hurricane Andrew. For example, Betty Robertson (1992) observed three different individuals on fallen trees at her Homestead residence within three weeks following Hurricane Andrew, and none have since been seen. However, only one young specimen was observed within four weeks prior to the hurricane. Loss of trees preclude the return of A. equestris at that site, even though the visibility of exposed lizards temporarily underestimated the detrimental effects of habitat loss. Lizards were also more conspicuous on structures not used prior to the hurricane. For example, I found lizards perching on palm trees and defoliated hibiscus which were less than one meter in height. Lizards were also found perching on black olive trees defoliated by the hurricane and subsequently pruned. As more striking examples, an adult pair of A. equestris was routinely seen basking on a mailbox at a home in Cutler Ridge following the hurricane (Galleno, 1992), and an adult was found basking ona pine stump immediately after the hurricane (Robertson, 1992). Reagan (1991) also witnessed a shift in structural niche use in canopy-dwelling anoles following Hurricane Hugo. For some A. equestris, secondary reliance on alternative structures may slow or alleviate the decrease in abundance following the partial loss of their habitat. For a heliotherm such as A. equestris, the effects of a hurricane can be conflicting. The destruction and defoliation of trees could serve to maintain partially open habitat, as well as to open previously unsuitable habitat. In an urban setting, the short-term effect was different. Yard counts (Table 1) showed that tree loss, in what was mostly open landscape prior to the hurricane, decreased some of the habitat available to the many resident A. equestris that survived the hurricane. The long- term effect in this urban setting, however, is not likely to be detrimental. Although some large trees have been removed, most of the habitat loss and degradation was temporary. Anolis equestris is able to use alternative structures, and this ability could minimize the effects of temporary habitat loss following Hurricane Andrew. Anolis sagrei—The brown anole is a small tree trunk-ground dweller of open clearings and its populations may become very dense (Williams, 1969). Anolis sagrei was first reported in Florida by Garman (1887). The geographic expansion of this species through the state has been rapid and human-assisted (Godley et al., 1981). 200 FLORIDA SCIENTIST [VOL 56 The penchant of A. sagrei for disturbed sites, often provided by human activities, was witnessed in its natural form following Hurricane Andrew. Under dark skies and in cold rain at 0700 hrs, two adults were observed climbing on debris near the side of my house. Within hours of the hurricane, lizards of all sizes were basking in full sun on the many hot sections of roofing and clumps of vegetation scattered on the ground. The many hatchlings that appeared in early August were actively utilizing newly created habitat. By noon, there appeared to be no fewer lizards in the yard than before the hurricane. Within a few days, debris was put in a pile, and lizards followed. The benefits of the hurricane to A. sagrei were two-fold. Left unaltered, it would seem that more habitat for A. sagrei had been created by the hurricane, thus creating conditions conducive to colonization and establishment of this ecologically versatile species. Secondly, the timing of the hurricane with the presence of newly hatched anoles could have accelerated the recruitment process into the newly created habitat. In the urban setting, however, the hurricane-related benefit of habitat creation was short-lived as debris was quickly removed. CONCLUSIONS—Four inferences may be drawn from the combined observa- tions. First, despite high winds and tornadic conditions that accompanied Hurricane Andrew, five species of tropical amphibians and reptiles responded positively to the natural phenomenon to which four are historically tied. In the case of A. equestris, low barometric pressure preceding the hurricane may have stimulated a pre- hurricane search for acceptable cover. Second, anurans were strongly stimulated to courtship activities following the storm. Responses boded well for these potential colonizers, especially because favorable spawning sites accompanied newly disturbed habitats. This was docu- mented for O. septentrionalis and B. marinus, and seemed likely for E. planirostris. Third, Anolis equestris and A. sagrei responded by basking immediately after the hurricane. The effects of the hurricane were conflicting for both species. Anolis equestris suffered the partial loss of its habitat, but may have compensated by the use of secondary structures, whereas A. sagrei quickly took advantage of newly created habitat afforded by the hurricane. Fourth, the nature of long-term effects of Hurricane Andrew on the five species in an urban setting was altered by subsequent cleanup activities. The negative effect was minimal for A. equestris because most habitat loss was not permanent. For the remaining species, exploitation of hurricane-related benefits has proved to be short- lived. ACKNOWLEDGMENTS—I thank Mark Deyrup, Henry Mushinsky, and Martin Tracey for reading earlier drafts of this paper. Their comments and discussions regarding this subject, and those of William B. Robertson, Jr. and Glen Woolfenden, are greatly appreciated. No. 4, 1993] MESHAKA—HURRICANE ANDREW 20] LITERATURE CITED BARBOUR, T. 1914. A contribution to the zoogeography of the West Indies, with special reference to amphibians and reptiles. Mem. Museum Comp. Zool. 44:205-346. ___—, 1931. Another introduced frog in North America. Copeia. 1931(3):140. Carb, A. F., JR. 1940. A contribution to the herpetology of Florida. Univ. Florida Publ. Biol. Sci. Ser. $1) silel ales CopE, E. D. 1875. Check-list of North american Batrachia and Reptilia. Bull. U.S. Nat. Mus. 1:1-104. DARLINGTON, P. J., JR. 1938. The origin of the Greater Antilles, with discussion of dispersal of animals over-water and through air. Quart. Rev. Biol. 13:274-300. DUELLMAN, W. E. AND A. SCHWARTZ. 1958. Amphibians and reptiles of southern Florida. Bull. Florida St. Mus. 3(5):181-324. GALLENO, R. 1992. Staff, Florida International University, Miami, Florida, Pers. Commun. GARMAN, S. 1887. On West Indian reptiles. Iguanidae. Bull. Essex Inst. 19:1-26. GODLEY, J. S., F. E. LOHRER, J. N. LAYNE, AND J. RossI. 1981. Distributional records of an introduced lizard in Florida. Anolis sagrei. Herpetol. Rev. 12(3): 84-86. NEILL, W. T. 1957. Historical biogeography of present-day Florida. Bull. Florida St. Mus. 2(7):176-220. PEMPERTON, C. E. 1934. Local investigations on the introduced tropical toad Bufo marinus. Hawaiian planters’ Record. 38(3):186-192. PREGILL, G. 1981. Late Pleistocene herpetofauna from Puerto Rico. Univ. Kansas Mus. Nat. Hist. Misc. Publ. No. 7, 72 pp. REAGAN, D. P. 1991. The response of Anolis lizards to hurricane-induced habitat changes in a Puerto Rican rain forest. Biotropica. 23 (4a): 468-474. ROBERTSON, B. 1992. Naturalist, Homestead, Florida, Pers. Commun. ROBERTSON, W. B., JR. 1992. Research biologist, Everglades National Park, Homestead, Florida, Pers. Commun. uh SCHWARTZ, A. 1952. Hyla septentrionalis Dumeril and Bibron on the Florida mainland. Copeia 2:117- 118. SIMPSON, G. G. 1966. Zoogeography of West Indian land mammals. Amer. Mus. Novitates. 1759:1-28. SMALL, G. 1992. Resident, Miami, Florida, Pers. Commun. TRACEY, M. L. 1992. Scientist, Florida International University, Miami, Florida, Pers. Commun. WILLIAMS, E. 1969. Ecology of colonization as seen in the zoogeography of anoline lizards in small islands. Quart. Rev. Biol. 44(4a):345-389. WILSON, L. D. AND L. PoRRAS. 1983. The ecological impact of man on the south Florida herpetofauna. Univ. Kansas Nat. Hist. Bull. No. 9, 89 pp. WOOLBRIGHT, L. L. 1991. The impact of hurricane Hugo on forest frogs in Puerto Rico. Biotropica. 23(4a):462-467. Florida Scient. 56(4):193-201.1993. Accepted: June 9, 1993. 202 FLORIDA SCIENTIST [VOL 56 SYNTHESIS OF A STERICALLY CHALLENGING BIS-(B-DIKETONE) VENKATRA] V. NARAYANAN Institute for Environmental Studies, Department of Chemistry, University of South Florida, 4202, E. Fowler Avenue, Tampa, FL 33520 ABSTRACT: Bis-($-diketones) are quadridentate ligands and have a broad spectrum of uses, including uses as biologically active compounds and liquid crystals. Compounds of the type [(RCO),CH],CHR’ have been synthesized by an aldol condensation followed by Michael addition. This work reports the synthesis of 1,1,3,3-tetra-(2-naphthoyl)-propane by this procedure. Bis-(B-DIKETONES) form unique metal complexes that may be polymers, dimers or unique monomers (Martin etal., 1958; Holst, 1955; Wilkins and Wittbecker, 1953). They have been complexed with metals like beryllium to form coordination polymers of high molecular weight and extraordinary properties (Kluiber and Lewis, 1960). Some rhodium and iridium metal complexes of bis-(B-diketones) are found to act as catalyst precursors for propylene hydrogenation (Whitmore and Eisenberg, 1984). A bulky bis-(B-diketone), which can find potential uses in the above areas, has been made, and the synthesis is described here (Fig. 1). MATERIALS AND METHODS—Synthesis of 1, 1, 3, 3-tetra-(2-naphtholyl)-propane—To a solution of 1,3-di-(2-naphthyl)-1,3-propanedione (0.5 g, 0.0015 mole, Aldrich Chemical Co.) in 50 mL ethanol and 10 mL toluene was added ten drops of aqueous formaldehyde (37 %, 0.0023 mole). Upon standing for three days, crystals precipitated that were collected by filtration. Recrystallization (once) from chloroform, followed by drying in the oven at 100°C for 15 minutes, gave 0.16 g of white crystals, (yield: 32%): M.P., 198°C; Anal: Caled. forC, -H,,O,: C, 85.45 H, 4.84 %; Found for C _H,,O,: C, 85.22; H, 4.78 %. (Desert Analytics); IR (CHCI,), 3057, 9985, 1692 (carbonyl), 1630 (enol- chelete) 1268, 738 cm!; UV (CHCI,), 297.6 nm (sharp peak, log = 2.03), 339.2 nm (medium peak, log = 1.44); 'H NMR (CDCL,, 0.01 % TMS), 1.6 (broad OH peak), 3.0 (t, - CH, -, 2H), 6.1 (t, -CH ,CH-, 2H), 7.5-8.9:Gn; 28 El) Se NMR (CDCI,), 29.5 (-CH), 54.5 (-CH,), 124.1- 135. 9 (9 non- equivalent aromatic carbons), 196.7 (C = O). The EG of the product showed one spot (eluant, 1:4 ethylacetate-hexane). The compound was also characterized by an enol test which was done as follows: 20 mg of the solid bis-(8-diketone) in 1 mL of methylene chloride was taken in a test-tube and 3-5 drops of a 1 % ( weight/volume) solution of ferric chloride in methylene chloride was added to it. A reddish-brown color was seen on warming the test- tube in a hot water bath for five minutes. RESULTS AND DISCUSSION—The compound (1,1,3,3-tetra-(2-naphthoyl)-pro- pane) is made of bulky terminal groups and can act as a precursor for many coordination metal complexes which should have interesting properties. The possi- bility of forming coordination polymers with unique properties, following the example of Wilkins and Wittbecker (1950) should not be overlooked. The bulky terminal groups provide a steric challenge to approaching metal ions, and may lead No. 4, 1993] NARAYANAN—SYNTHESIS OF A BIS-B-DIKETONE 203 pipendine : OO QO Hes EtoH/ OyMe o 0 HO FIG. 1 Synthesis of 1,1,3,3-tetra-(2-naphthoyl)-propane to stereospecific coordination polymers. Also, the study of interactions between “host” and “guest” molecules is an area of continuing activity. Previous investigations of this type have dealt with organic hosts, including crown ethers, cyclophanes, and calixarenes, as well as with cyclic nitrogen donors that can accommodate both transition-metal ions and small molecule guests. Presumably 1,1,3,3-tetra-(2- naphthoyl)-propane and other similar sterically hindered bis-(B-diketones) would be good host molecules chelating with such metal ions as copper(II), following the example of Maverick and coworkers (1986), and also metal ions with higher coordination numbers zirconium and hafnium, following the example of Whitmore (1984). ACKNOWLEDGMENTS—The author wishes to gratefully acknowledge the helpful discussion with Dean F. Martin and Chuhua Wang in the conduct of this research. LITERATURE CITED HOLST, E. H. 1955. Doctoral Dissert. The Pennsylvania State University, University Park, PA. KLUIBER, R. W. AND J. W. LEwIs. 1960. Inner complexes. II. Macrocyclic beryllium chelates and their polymers. J. Am. Chem. Soc. 82: 5777-5779. MARTIN, D. F., M. SHAMMA, AND C. W. FERNELIUS, 1958. Bis-(8-diketones). II. The synthesis and spectra of compounds of the type [(RCO)R’COCH],CHR. J. Am. Chem. Soc. 80: 5851-5856. MAVERICK, A.W., S.C. BUCKINGHAM, Q. YAO, J.R. BRADBURY, AND G.S. STANLEY. 1986. Intramolecu- lar coordination of bidentate lewis bases to a cofacial binuclear copper(II) complex. J. Am. Chem. Soc. 108: 7430-7431. WHITMORE, B.C. AND R. EISENBERG. 1984. Flexibly bridged binuclear rhodium and iridium complexes of p-xylelenebis(3-(2,4-pentanedione)). Inorg. Chem. 23: 1697-1703. WILKINS, J. P. AND E. L. WITTBECKER. 1953. U. S. Patent 2,659,711. Florida Scient. 56(4):202-203.1993. Accepted: July 16, 1993. 204 FLORIDA SCIENTIST [VOL 56 Biological Sciences GEOGRAPHIC DISTRIBUTION OF THE STRIPED MULLET (MUGIL CEPHALUS LINNAEUS) IN THE ATLANTIC AND EASTERN PACIFIC OCEANS CARTER R. GILBERT Florida Museum of Natural History, University of Florida, Gainesville, Florida 32611 ABSTRACT: The striped mullet (Mugil cephalus Linnaeus), an inshore fish species, is distributed worldwide and forms an important part of the commercial catch in many areas, including Florida. Its geographic range in the eastern Pacific Ocean (northern Chile to central California) has been known for some time, but in the Atlantic Ocean its distribution has been a source of confusion. Evidence is presented to show that in the eastern Atlantic M. cephalus occurs as far south as Ghana (where it evidently is extremely rare) and does not range down to the tip of Africa, as some literature references indicate. In the western Atlantic it is found from New England south as far as the Gulf of Campeche, Mexico. Previous reports of its occurrence throughout the West Indies and along the Atlantic coasts of Central and South America are erroneous. THE STRIPED, or grey mullet (Mugil cephalus Linnaeus), animportant commer- cial fish in many areas (Rivas, 1980; Collins, 1985), is one of only a tiny number of inshore marine fish species to exhibit a worldwide distribution (14 according to Briggs [1960]). Itis also catadromous, with spawning occurring in salt water but with many adults spending most of their lives in fresh water, where they may move hundreds of miles upstream (Burgess, 1980; Robison and Buchanan, 1988). Briggs (1960) and others have suggested that studies using modern analytical techniques will show M. cephalus to be a complex of cryptic species. Until recently, the only published analysis of variation among populations appears to have been that by Ebeling (1957, 1961), who (employing standard morphological characters) compared specimens from the eastern Pacific region with those from Hawaii, Florida, Italy (Mediterranean Sea), and Romania (Black Sea). He could detect no discernible differences among these populations, and in fact found the greatest amount of morphological variation (relating to details of dentition) to be among eastern Pacific populations. Based on its title, Thomson’s (198la) paper on the taxonomy of grey mullets would seem to promise answers to this problem, but unfortunately this proves not to be the case. Detailed systematic studies, involving several teams of scientists, have recently been initiated on striped mullet populations throughout the world. Although morphological data are not yet available, analyses of the genetic structure of mitochondrial DNA (Crosetti and others, 1993; Crosetti and others, in press), as well as electrophoretic analyses of allozyme frequencies (Rossi and others, unpubl. ms), indicate considerable variation among populations. Surprisingly, none of these / No. 4, 1993] GILBERT—DISTRIBUTION OF THE STRIPED MULLET 905 studies has revealed a level of differentiation sufficient to justify the recognition of distinct species. Collins (1985) indicated Mugil cephalus to be generally distributed in coastal regions around the world between 42°N and 42°S, and this statement probably accurately reflects the general perception of its range among most ichthyologists and fisheries biologists. In the western Atlantic region its range has regularly been indicated as extending from Cape Cod to southeastern Brazil, including the West Indies (e.g., Jordan and Evermann, 1896; Hildebrand and Schroeder, 1928; Leim and Scott, 1966; Duarte-Bello and Buesa, 1973; Thomson, 1963, 1977; Burgess, 1980; Crosetti and others, 1993), and Miller (1966) included it in his list of species occurring in fresh waters of Central America. In the eastern Atlantic it has usually been defined as the entire Mediterranean region and the Atlantic Ocean from southern France (Bay of Biscay) south along the west African coast at least to 23°S latitude (e.g., Burgess, 1980; Thomson, 1981b). In the eastern Pacific it has been recorded as far north as San Francisco Bay (Reilly and Sakanari, 1982), but its normal range extends from extreme southern California south to northern Chile (Ebeling, 1961: Rosenblatt, 1993). The range limits for the striped mullet indicated above for the eastern Pacific region seem to be accurate, although (as discussed subsequently) details of its distribution throughout much of this region have heretofore been uncertain. Closer inspection of the literature, together with a survey of mugilid material in museum collections, indicate that the overall range of Mugil cephalus in the Atlantic region is much more restricted than indicated above. (Analysis of its distribution elsewhere is beyond the scope of this paper). Northern range limits indicated above for the Atlantic are essentially correct, although in the western Atlantic individuals may occasionally wander north seasonally to Maine and Nova Scotia (Leim and Scott, 1966). In the eastern Atlantic, the species is known to be widely distributed throughout the Mediterranean and Black seas, and it is also said to be present around the islands of Madeira and the Azores (Ben-Tuvia, 1986). If its reported occurrence around Madeira and the Azores is correct, it signifies a surprising difference compared to the situation on the other side of the ocean, since it is absent from islands in the western Atlantic region. Although said to be distributed along the entire coast of west Africa, Sanches (1966) did not include it in his work on Angolan fishes, nor did Daget and Iltis (1965) record it from the Ivory Coast, which is located along the northern coast of the Gulf of Guinea at about 5°N latitude. Cadenat (1951) and Massuti (1967) both listed it from Senegal (about 15°N), where it evidently is rare (Cadenat made no specific mention of level of abundance, but Massuti’s report was based on only a single specimen). Cadenat (1951) also indicated M. cephalus to range south and east along the African coast as far as Togo, but because he presented no substantiating evidence and because Togo is situated east of the Ivory Coast, this record was considered suspect. Recently, however, Jeffrey T. Williams informed me of an individual measuring about 140 mm standard length (SL), collected at Tema, Ghana, on 16 April 1962. This specimen (USNM 305487), which was only recently discov- ered among previously uncatalogued material, results in a significant range extension 206 FLORIDA SCIENTIST [VOL 56 for this species in the eastern Atlantic region, and lends credence to Cadenat’s earlier report from Togo. The striped mullet appears to be absent from the rest of the west African coast, but reappears in the southwestern Indian Ocean, all of the localities listed by Smith (1935) for South Africa falling east of the Cape of Good Hope. In the western Atlantic region, Menezes (1983) did not include Mugil cephalus in his guide to the mugilids of Brazil, Schultz (1949) and Cervigon (1966) did not list it from Venezuela, and Meek and Hildebrand (1923) did not find it in Caribbean waters of Panama. Gilbert and Kelso (1971) did not collect it at Tortuguero, Costa Rica, and Bussing (1976, 1987) did not include it in his studies on Costa Rican freshwater fishes. David Greenfield (1993) did not encounter striped mullet in the numerous collections made by him and colleagues in Belize. Lee and others (1983) indicated it as one of the species to be expected from the Greater Antilles, although they provided no substantiating records. Evermann and Marsh (1900), Beebe and Tee-Van (1928), Caldwell (1966), Bohlke and Chaplin (1968), and Burgess and Franz (1989) did not list it from Puerto Rico, Haiti, Jamaica, the Bahamas, or Hispaniola, respectively, although five other mullets (including four species of Mugil) known from Florida occur in one or more of these areas. Duarte-Bello and Buesa (1973) considered it to be of questionable occurrence in Cuba, but Jordan and Evermann (1896) had earlier stated that the species was not known from that island. Duarte-Bello and Buesa (1973) indicated it to be present in Bermuda, but neither Bean (1906) nor Beebe and Tee-Van (1933) found it there. Finally, no specimens of M. cephalus are present in the Florida Museum of Natural History, the University of Michigan Museum of Zoology (Miller, 1993), National Museum of Natural History (Williams, 1993), and the University of Puerto Rico ichthyological collection (Hensley, 1993; Dennis, 1993) from any West Indian island, although collections of other species of Mugil are available from throughout this region. The distribution of Mugil cephalus in the western Atlantic region thus appears to be restricted to continental waters of the United States and Mexico, with the southernmost confirmed place of occurrence apparently being Tampico, Mexico (Jordan and Dickerson, 1908). In the eastern Atlantic its range can be defined as the Mediterranean and Black seas, from the Bay of Biscay (France) south to Ghana and possibly Togo (ca. 5°N latitude), and presumably including the islands of Madeira and the Azores. In contrast to its restricted distribution in the Atlantic Ocean, Mugil cephalus exhibits a much broader latitudinal distribution in the eastern Pacific. The southern limits of its distribution in the latter region (i.e., northern Chile, at about 20°S latitude) have been documented by Ebeling’s (1957, 1961) studies, and its range is well documented along most of the western coast of South America (between Chile and Ecuador). To the north its distribution is also well documented. Reilly and Sakanari (1982) recorded it as a stray from San Francisco Bay, and it regularly ranges southward from the San Diego area along the coasts of California and western Baja California into the Gulf of California, where it is common (Thomson and others, 1979; Lea and others, 1988). Hildebrand’s (1925) record of the species from El Salvador is valid, as indicated by reexamination (by J. T. Williams) of a 320 mm SL specimen (USNM 87279) collected during the Hildebrand survey. Between El No. 4, 1993] GILBERT—DISTRIBUTION OF THE STRIPED MULLET 207 Salvador and Ecuador, however, the eastern Pacific distribution of Mugil cephalus becomes less clear. Villa (1982) reported it from fresh waters of Nicaragua, but provided no specific records. Bussing (1987) did not include it in his book on Costa Rican freshwater fishes. Jordan (1885) reported it from the Pacific coast of Panama, but since this was included in an unannotated checklist, it may have been listed solely on the basis ofits presumed occurrence there. This would seem to be verified by both Meek and Hildebrand (1923) and Loftin (1965), neither of whom encountered M. cephalus during their extensive surveys of Panama. The lack of records of striped mullet in the literature from much of Central America, together with the absence of voucher specimens in museums with good holdings of eastern Pacific shore fishes (National Museum of Natural History, Los Angeles County Museum of Natural History, Scripps Institution of Oceanography), initially seems to suggest that the species might be absent throughout much of this region, thus displaying an anti-tropical distribution. However, Lavenberg (1993) informed me that he has examined unquestioned specimens of Mugil cephalus during visits (in December 1990) to local fish markets along the Pacific coasts of Nicaragua, Panama, and Colombia, but did not find it in markets in El Salvador, Honduras, or Costa Rica. Also, Bussing (1993) told me that he has recently seen individuals in Pacific Costa Rican fish markets as well. Although voucher specimens unfortunately were not saved from any of these places, I consider this information to be reliable. Based on these observations, both Drs. Lavenberg and Bussing believe the absence of museum specimens from the Pacific coasts of Central and northern South America to be a collecting artifact, but they also agree that the species is not common throughout this area. In the western Atlantic region, at least, Mugil cephalus fulfills the definition of a continental species. As Robins (1971) and Gilbert (1972) have pointed out, such species are not confined to continental waters by their physical inability to reach insular areas, but rather by different ecological parameters associated with the two places (i.e., variable levels of turbidity, salinity, and temperature in continental waters versus normally greater stability of these conditions in insular waters). In the southeastern Pacific, however, Mugil cephalus is known to occur around offshore islands along the coasts of Peru and Chile (Ebeling, 1957; 1961), and reference was made above to its presumed occurrence around the Azores and Madeira, in the eastern Atlantic. Ebeling (1961) also examined specimens from Hawaii. No expla- nation is presently available to account for these apparent differences in habitat preference. Should there be subsequent literature reports of Mugil cephalus from areas outside those defined in this paper, it is important that they be verified by voucher specimens. In the western Atlantic region, at least, such records likely would be based on Mugil liza Valenciennes (vernacular name “liza” or “lebrancho”), which, among the western Atlantic species of Mugil, is clearly the one to which M. cephalus is most closely related (Rivas, 1980). Rivas (1980) pointed out that these species are not always easy to distinguish, but that M. liza differs from M. cephalus in reaching a greater length (specimens over two feet long are common), in having 31-36 scales in the lateral series (vs. 38-42 in M. cephalus), in having a less evenly concave caudal 208 FLORIDA SCIENTIST [VOL 56 fin, and in several mensural characters. In the eastern Pacific, by contrast, there is no other species of Mugil with which M. cephalus is likely to be confused (Lavenberg, 1993; Bussing, 1993). ACKNOWLEDGMENTS—I wish to express my appreciation to Robert R. Miller, University of Michigan; and Richard H. Rosenblatt, Scripps Institution of Oceanography, University of California at San Diego, both of whom reviewed a preliminary version of this paper and provided a number of useful comments and observations. I also thank Jeffrey A. Seigel and Camm C. Swift, Los Angeles County Museum of Natural History, for information on eastern Pacific mugilid distribution; David Greenfield, now of the University of Hawaii, who supplied information regarding mullet species collected by him during his ichthyological exploration of Belize; Scott A. Schaefer, Academy of Natural Sciences of Philadelphia; Dannie A. Hensley, University of Puerto Rico; and George D. Dennis, formerly of the University of Puerto Rico. I especially wish to thank the following individuals, each of who provided information resulting in significant changes in the original manuscript: Jeffrey T. Williams, National Museum of Natural History, who in addition to serving as a reviewer of this paper, supplied information on specimens of Mugil cephalus in the USNM collection that resulted in expansions of the range of this species in both the eastern Pacific and eastern Atlantic regions; Robert J. Lavenberg, Los Angeles County Museum of Natural History, and William A. Bussing, University of Costa Rica, for information resulting from visits to Pacific fish markets between E] Salvador and Colombia; and Donald G. Campton, University of Florida, and Anna Rita Rossi, University of Rome, who generously shared information emanating from their unpublished studies on genetic relationships of striped mullet populations from throughout the world and who also provided me with an unpublished manuscript dealing with DNA relationships of this species. LITERATURE CITED BEAN, T. H. 1906. A catalogue of the fishes of Bermuda, with notes on a collection made in 1905 for the Field Museum. Field Columbian Mus. Publ. 108 (Zool. Ser.) 7(2):; 21-89. BEEBE, W., AND J. TEE-VAN. 1928. The fishes of Port-au-Prince Bay, Haiti. Zoologica 10(1): 1-276. . 1933. Field book of the Shore Fishes of Bermuda and the West Indies. Dover Publ. Co., New York. xiii + 337 pp. BEN-TuviA, A. 1986. Mugilidae. Pp. 1197-1204. In WHITEHEAD, P. J. P., M. L. BAUCHOT, J. C. HUREAU, J. NIELSEN, AND E. TORTONESE (eds.), Fishes of the North-eastern Atlantic and the Mediterranean 3: 1015-1473. BOHLKE, J. E. AND C. C. G. CHAPLIN. 1968. Fishes of the Bahamas and adjacent tropical waters. Livingston Publ. Co., Narbeth, Pa. xxiii + 771 pp. BRIGGS, J. C. 1960. Fishes of worldwide (circumtropical) distribution. Copeia 1960(3): 171-180. BURGESS, G. H. 1980. Mugil cephalus Linnaeus, Striped mullet. P. 779 in D. S. LEE and others (eds.), Atlas of North American Freshwater Fishes. North Carolina St. Mus. Nat. Hist., Raleigh. i-x + 854 pp. Ee AND R. FRANZ. 1989. Zoogeography of the Antillean freshwater fish fauna. Pp. 263-304. In C. H. Woops, (Ed.), Biogeography of the West Indies. Sandhill Crane Publ. Co., Gainesville, FL xvii + 878 pp. BUSSING, W. A. 1976. Geographic distribution of the San Juan ichthyofauna of Central America with remarks on its origin and ecology. Pp. 157-175. In THORSON, T. B. (ed.), Investigations of the Ichthyofauna of Nicaraguan Lakes. Univ. of Nebraska, Lincoln, NE. x + 663 pp. . 1987. Peces de las aguas continentales de Costa Rica. Editorial de la Universidad de Costa Rica. 271 pp. . 1993. University of Costa Rica, San Jose, Costa Rica, Pers. Comm. CADENAT, J. 1951. Poissons de Mer du Senegal. Initiations Africaines 3, Inst. Franc. d’Afrique Noire, vol. 3. 345 pp., 241 figs. (dated “1950”) CALDWELL, D. K. 1966. Marine and freshwater fishes of Jamaica. Bull. Inst. Jamaica 17: 7-120. CERVIGON-M., F. 1966. Los peces marinos de Venezuela. Tomo 1. Estacion de Investigaciones Marinas de Margarita, Fundacion La Salle de Ciencias Naturales, Monogr. 11. 436 pp. COLLINS, M. R. 1985. Species profiles: life histories and environmental requirements of coastal fishes and invertebrates (South Florida)—Striped mullet. U.S. Fish & Wild. Serv. Biol. Rept. 82(11.34). U.S. Army Corps of Engineers, TR EL-82-4. 11 pp. CROSETTI, D., J. C. AVISE, F. PLACIDI, A. R. ROSSI, AND L. SOLA. 1993. Geographic variability in the No. 4, 1993] GILBERT—DISTRIBUTION OF THE STRIPED MULLET 209 grey mullet Mugil cephalus: preliminary results of mt DNA and chromosome analyses. Aquacul- ture 111: 95-101. , W.S. NELSON, AND J.C. AVISE. In press. Pronounced genetic structure of mitochondrial DNA among populations of the circum-globally distributed grey mullet (Mugil cephalus Linnaeus). J. Fish Biol. DaGET, J. AND A. ILTIS. 1965. Poissons de Cote d'Ivoire (eaux douces et saumatres). Mémoires de l'Institut Francais d’Afrique Noire 74: v-xi, 1-385. DENNIS, G. D. 1993. University of Puerto Rico (formerly), Mayaguez, PR. Pers. Comm. DUARTE-BELLO, P. P. ANDR. J. BUESA. 1973. Catalogo de Peces Cubanos (Primera revision). I. Indice Taxonomico. Ciencias, Serie 8 (Investigaciones Marinas no. 3). 255 pp. EBELING, A. W. 1957. The dentition of eastern Pacific mullets, with reference to adaptation and taxonomy. Copeia 1957(3): 173-185. . 1961. Mugil galapagensis, a new mullet from the Galapagos Islands, with notes on related species and a key to the Mugilidae of the eastern Pacific. Copeia 1961(3): 295-305. EVERMANN, B. W. AND M. C. MARSH. 1900. The fishes of Porto Rico. Bull. U. S. Fish Comm. 20(1): 49-350, col. pls. 1-49. GILBERT, C. R. 1972. Characteristics of the western Atlantic reef-fish fauna Quart. J. Florida Acad. Sci. 35(2-3): 130-144. AND D. P. KELSO. 1971. Fishes of the Tortuguero area, Caribbean Costa Rica. Bull. Florida St. Mus., Biol. Sci. 16(1): 1-54. GREENFIELD, D. 1993. University of Hawaii, Honolulu, HI, Pers. Comm. HENSLEY, D. A. 1993. University of Puerto Rico, Mayaguez, PR, Pers. Comm. HILDEBRAND, S. F. 1925. Fishes of the republic of El Salvador, Central America. Bull. U.S. Bur. Fish. Al: 238-287. AND W. C. SCHROEDER. 1928. Fishes of Chesapeake Bay. Bull. U. S. Bur. Fish. (1927) 43(1): 1-366. JORDAN, D. S. 1885. A list of the fishes known from the Pacific coast of tropical America, from the Tropic of Cancer to Panama. Proc. U. S. Nat. Mus. 8(526): 361-394. AND M. C. DICKERSON. 1908. Notes on a collection of fishes from the Gulf of Mexico at Vera Cruz and Tampico. Proc. U.S. Nat. Mus. 34(1592): 11-22. AND B. W. EVERMANN. 1896. Fishes of North and Middle America. Bull. U.S. Nat. Mus. 47(1): iii-lx, 1-1240. LAVENBERG, R. J. 1993. Los Angeles County Museum of Natural History, Los Angeles, CA, Pers. Comm. LEA, R.N.,C.C. SWIFT, AND R. J. LAVENBERG. 1988. Records of Mugil curema Valenciennes, the white mullet, from southern California. Bull. Southern Calif. Acad. Sci. 87(1): 31-34. LEE, D. S., S. P. PLATANIA, AND G. H. BURGESS. 1983. Atlas of North American Fishes. (1983 supplement). North Carolina St. Mus., Raleigh, NC. 67 pp. LEIM, A. H. AND W. B. SCOTT. 1966. Fishes of the Atlantic coast of Canada. Bull. Fish. Res. Bd. Canada i5a2 1-485: LOFTIN, H. G. 1965. The geographical distribution of freshwater fishes in Panama. Unpubl. Ph.D. dissert., Florida State University. MASSUTI, M. 1967. Los gondos de pesca en la plataforma continental de Marruecos, Sahara espaiiol. Mauritania, Senegal, Republic de Guinea, Ensenada de Biafra y Archipiélago de Cabo Verde. Trabajos del Instituto Espafiol de Oceanografia 34: 1-99. MEEK, S. E. AND S. F. HILDEBRAND. 1923. The marine fishes of Panama. Part 1. Field Mus. Nat. Hist., Publ. 215 (Zool. Ser.) 15: v-xl, 1-330. MENEZES, N. A. 1983. Guia pratico para conhecimento e identifiaco de tainhas e paratis (Pisces, Mugilidae) do litoral brasileiro. Revista Brasileira de Zoologia 2(1): 1-12. MILLER, R. R. 1966. Geographical distribution of Central American freshwater fishes. Copeia 1966(4): 773-802. (reprinted with revised pagination [pp. 125-156] In: THORSON, T. B. [ed.], 1976. Investigations of the Ichthyofauna of Nicaraguan Lakes). . 1993. University of Michigan, Ann Arbor, MI, Pers. Comm. REILLY, C. A. AND J. SAKANARI. 1982. Records of the striped mullet, Mugil cephalus, in San Francisco Bay, California. California Fish and Game 68(4): 190. RIVvAS, L. R. 1980. Synopsis of knowledge on the taxonomy, biology, distribution, and fishery of the Gulf of Mexico mullets (Pisces: Mugilidae). Pp. 34-53. In: FLANDORFER, M. AND SKUPIEN, L. (eds.), Proceedings of a workshop for potential fishery resources of the northern Gulf of Mexico. Miss.- Ala. Sea Grant Consort. Publ. MASGP-80-012. ROBINS, C. R. 1971. Distributional patterns of fishes from coastal and shelf waters of the tropical western 210 FLORIDA SCIENTIST [VOL 56 Atlantic. Pp. 249-255. In: Symposium on investigations and resources of the Caribbean Sea and adjacent regions. FAO, Fish. Rept. 71-2. ROBISON, H. W. AND T. M. BUCHANAN. 1988. Fishes of Arkansas. Univ. of Arkansas Press, Fayetteville, AK xvi + 535 pp. ROSENBLATT, R. H. 1993. Scripps Institution of Oceanography, La Jolla, CA, Pers. Comm. Rossi, A. R., D. CROSETTI, L. SOLA, and D. E. CAMPTON. unpubl. ms. Allozyme evidence for subspecies, genetic differentiation among global populations of Mugil cephalus (Pisces: Mugilidae). SANCHES, J. G. 1966. Peixes de Angola (Teledsteos). Notas Mimeografidas do Centro de Biologia Piscat6ria (Lisboa) 46: 1-227. SCHULTZ, L. P. 1949. A further contribution to the ichthyology of Venezuela. Proc. U. S. Nat. Mus. 99(3235): 1-211. SMITH, J. L. B. 1935. The fishes of the family Mugilidae in South Africa. Ann. South African Mus. 30(5): 587-644. THOMSON, D. A., L. T. FINDLEY, AND A. N. KERSTITCH. 1979. Reef fishes of the Sea of Cortez. John Wiley & Sons, New York. xvii + 302 pp. THOMSON, J. M. 1963. Synopsis of biological data on the grey mullet (Mugil cephalus Linnaeus, 1758). Fish. Synop. Div. Fish. Oceanogr., SCIRO, Australia. . 1977. Mugilidae. In: FAO Species Identification Sheets for Fisheries Purposes, Western Central Atlantic. FISCHER, W. (ed.). Vol. 3. (unnumbered pages). . 1981a. The taxonomy of grey mullets. pp. 1-15. In: OREN, O. H. (ed.), Aquaculture of Grey Mullets. Cambridge Univ. Press, IBP handbook, 26. xxi + 507 pp. . 1981b. Mugilidae. In: FAO Species Identification Sheets for Fishery Purposes, Eastern Central Atlantic. W., FISCHER, G., GIANCHI, AND W. B., ScoTrT (eds.). Vol. 3 (unnumbered ages). VILLA, I 1989. Peces Nicaraguenses de Agua Dulce. Coleccion Cultural, Serie Geografia y Naturaleza, no. 3. Managua, Nicaragua. 253 pp. WILLIAMS, J. T. 1993. National Museum of Natural History, Washington, DC, Pers. Comm. Florida Scient. 56(4):204-210.1993. Accepted: July 27, 1993. No. 4, 1993] | PROFFITT ET AL—EXPERIMENTS ON THE MANGROVE LEAF LITTER 21] Biological Sciences FIELD AND LABORATORY EXPERIMENTS ON THE CONSUMPTION OF MANGROVE LEAF LITTER BY THE MACRODETRITIVORE MELAMPUS COFFEUS L. (GASTROPODA: PULMONATA) C. Epwarb Prorritt', KEvIN M. Jouns, C. BRUCE COCHRANE, Donna J. DEVLIN, THERESA A. REYNOLDS, DEBORAH L. PAYNE, SEAN JEPPESEN, Davip W. PEEL, AND Dwane D. LINDEN Natural Sciences Program, St. Petersburg Junior College, St. Petersburg/Gibbs Campus, St. Petersburg, FL ABSTRACT: The gastropod Melampus coffeus rapidly consumes mangrove leaf detritus and shows preferences for leaves of certain species and degree of leaf senescence. In the field, most grazing on leaves occurred over the first 24 days with percent total leaf area affected by grazing being 84.6 (Rhizophora mangle), 16.6 (Laguncularia racemosa), and 98.8 (Avicennia germinans). By the end of the 42 day laboratory experiment, about 70 % of R. mangle green leaf area and 80 % of R. mangle brown leaf area had been completely consumed. Also, 70 % of A. germinans brown leaf area had been consumed, while for L. racemosa about 33 % of green and 60 % brown leaf area had been consumed. The high density of snails (99.4 - 130.2/m? ) at the field study site, coupled with rapid consumption of leaf detritus, suggests that M. coffeus is an important component of the overall breakdown and export of mangrove leaf organic matter. THE decomposition of detritus from coastal wetlands is an integral component of the food webs of coastal wetland, estuarine, and near-shore marine ecosystems (Heald, 1969; MacNae, 1974; Odum and Heald, 1972, 1975; Turner, 1977; Twilley et al., 1986). In south Florida and the tropics mangrove leaves comprise a substantial proportion of this detritus. Decomposition of mangrove leaf litter involves leaching of certain chemicals, microbial and fungal activities, and physical breakdown of particle size and consti- tution as a result of abiotic and macrodetritivore action. Some studies have focused on chemical and mass changes in detritus during decomposition (Fell and Masters, 1980; Flores-Verdugo et al., 1987; Heald, 1969; Twilley et al., 1986) and the contributions of microbes to decay of mangrove, salt marsh and sea grass detritus (Tenore, 1977). Various studies have acknowledged the contribution of macrodetritivores to mangrove leaf decomposition (Boonruang, 1978; Heald, 1969; Twilley et al., 1986), but attempts to quantify this component of leaf degradation have seldom been made in this hemisphere. Golley and co-workers (1962) suggested that because of low metabolic rates and densities, macrodetritivores were not of great importance in leaf decomposition in a Puerto Rican red mangrove forest. ' Current Address: Center for Environmental Research, P.O. Box 90655, McNeese State University, Lake Charles, La. 212 FLORIDA SCIENTIST [VOL 56 However, In Australia Robertson (1986) and Robertson and Daniel (1989) found that a suite of crabs (Family Grapsiidae) devour much of the leaf litter. Consequently, much of the mangrove production is transferred to the nearshore marine food webs as crab larvae and fecal pellets. Similarly, in Malaysia, Malley (1978) reported that the extremely abundant Sesarmid crab Chiromanthes onychophorum consumed mangrove leaf litter and converted leaf production into fecal particles. Our work focuses on the consumption of mangrove leaf litter by the macrodetritivore gastropod Melampus coffeus L. (Pulmonata: Ellobiidae). M. coffeus is abundant in intertidal regions of many Florida and Caribbean mangrove ecosys- tems (Golley, 1960, Heard, 1982). These snails have generally been listed as detritivores (Golley et al. 1962; Heard, 1982; Richie and Johnson, 1991), but also have been noted to eat fresh mangrove leaves in a laboratory setting (Mook, 1986). M. coffeus is reported to be attracted to both Rhizophora mangle L. and Avicennia germinans L. leaves, but to consume the A. germinans leaves more rapidly (Richie and Johnson, 1991). In addition, M. coffeus grazing of mangrove leaves may adversely affect mosquito populations that lay their eggs on leaf detritus (Richie and Johnson, 1991). M. coffeus and a closely related species Melampus bidentatus (Say) sometimes co-occur in Florida mangrove ecosystems, although M. bidentatus is more com- monly found in salt marsh habitats throughout much of the Atlantic and Gulf coasts of North America (Holle and Dineen 1957). In Tampa Bay, when the two species co- occur in mangrove forests, M. coffeus typically inhabits in the actual forested mangrove ecosystem and M. bidentatus dominates the highest tidal zone, often vegetated by Salicornia spp. (Mook, 1973). The animals forage on mangrove leaf litter when low tide exposes the forest floor (Richie and Johnson, 1991; Proffitt and Devlin, in prep.). At high tide, individuals larger than about 4 mm in shell length climb tree trunks, prop roots, seedlings and protruding sticks and debris to escape the rising water. Smaller juveniles stay in driftwood crevices or under leaves while covered by the tide (CEP and DJD Personal Field Observations). Densities of this animal can reach over 500 / m? (Heard, 1982). The hermaph- roditic snails M. bidentatus and possibly M. coffeus may produce as many as 30,000 eggs per individual per year (Apley, 1968). Similar numbers of eggs have been reported for M. coffeus (Marcus and Marcus, 1965) but detailed studies on reproduc- tive behavior and output have not been performed. Egg masses are attached to leaves and driftwood. Larvae hatch as veligers and spend several weeks in the plankton community before returning to the mangrove ecosystem on a high tide for metamor- phosis into mature form (Apley, 1968). The large populations of this snail in many mangrove forests suggest that a substantial proportion of mangrove leaf litter in many mangrove forests may enter the food web after being completely or partially consumed by M. coffeus. However, little quantitative information exists as to rates of consumption, preferences of M. coffeus for various types of mangrove leaves, or whether the snail feeds primarily on microbial components of detritus or directly on the vascular plant material. No. 4, 1993] | PROFFITT ET AL—EXPERIMENTS ON THE MANGROVE LEAF LITTER 213 METHODS—In laboratory and field experiments, we measured rates of M. coffeus grazing on leaves of three species of mangroves. In the laboratory, we assessed preferences of the snail for leaves of different mangrove species and various stages of senescence and decomposition (green, yellow, and brown leaves). In the field, we determined the rate of consumption of leaves of the various species of mangroves by a population of the snails. All field studies of M. coffeus populations and mangrove leaf degradation were carried out at Veteran’s War Memorial Park in northern Boca Ciega Bay, Pinellas County, Florida. In order to characterize the forest, densities of mangroves were assessed during low tide on 1/14/ 91 at various points along a 130 m long transect that encompassed the intertidal zone. At selected sites along the transect numbers of saplings, seedlings, and unrooted propagules were recorded for triplicate 1 x 1 m quadrats adjacent to each of larger, single (5 x 5 m) plots used to characterize density patterns of mature trees. Densities of M. coffeus were sampled using five replicate 30 x 30 cm quadrats at various stations along the same transect at approximately bimonthly intervals from 2/15/91 to 2/13/92. Quadrats were located in a stratified haphazard manner that ensured that at least one of the five replicates contained a tree trunk ora prop root. For the field experiment, mangrove leaves of all three species were collected 1/7/91 and length, width, thickness, cleaned wet weight, and surface area measured. Only intact leaves, with no holes, were used. A tracing of each leaf was made on graph paper having squares of 1 x 1 mm. On 1/25/91, 15 leaves of R. mangle and A. germinans and 10 of Laguncularia racemosa Gaertn.f were tethered to individual stakes by tying 30 cm long strands of thin monofilament line to their petioles. These leaves are referred to hereafter as “Exposed” leaves, because they were available for macrodetritivore grazing. As a control, an additional 15 leaves of each species were enclosed within individual polypropylene screen bags having a mesh size of 1 2mm and tethered to individual stakes. Exposed and Control leaves were placed 50 - 55 m from the shoreline, within the region of the intertidal zone dominated by A. germinans, and secured by driving the stakes into the substratum. ae Control and Exposed leaves were collected (typically n=5, max n=10, min n=1 per treatment), at 10, 24, and 53 days. The leaves were gently rinsed to remove as much muck, salt, and sand as possible and blotted dry. Leaf area loss was assessed in two ways. First, the Area of Holes produced by macrodetritivore grazing was determined by placing the leaf over its tracing and counting graph paper squares where leaf material was missing. Counts were made while viewing leaves through a dissecting microscope. However, in many feeding instances, snails tended to graze some layers of cells off large sections of leaves which produced easily visible, translucent patches on the leaves. Thus, this necessi- tated use of asecond measure of consumption, the Total Leaf Area Affected which includes both the area of holes in a leaf and the area of translucent patches of leaf where layers of cells were shaved off by macrodetritivore grazing activity. The area of translucent patches was measured in similar fashion to that of holes. After area measurements were completed, leaves were dried in an oven at 70 degrees C. until constant oven dry weight was obtained. In the laboratory experiment, we assessed M. coffeus feeding preferences and rates of leaf consumption. Snails were held under room temperature conditions, with normal light-dark periods, and natural sediment in their aquarium that was kept moist but without standing water. In this experiment, 150 mature M. coffeus were placed in an aquarium (61 x 31.5 cm) and presented choices of open Petri dishes containing green (pulled from trees), yellow (taken from trees, nearly ready to abscise and fall as litter), and brown (from the ground) leaves of R. mangle, A. germinans, and L. racemosa. Three replicate leaves of each type were used. As acontrol, three other green, yellow, and brown leaves of each mangrove species were placed in Petri dishes that were covered to exclude snails. Selected leaves were represen- tative of mature, undamaged leaves of their species and were generally similar in size. The Area of Holes im in leaves and the Total Leaf Area Affected by snail feeding were assessed at 8, 14, 21, 28, and 42 ays. For both field and laboratory experiments, statistical analyses of leaf consumption measurements are presented as treatment means (with one standard deviation in parentheses) in units of mm of leaf area eaten. RESULTS—Natural mangrove and M. coffeus density patterns—Along our 130 m transect from shoreline to upland in the War Veteran’s Park, there are 4063 (2937.8) mangrove trees per ha. R. mangle is the dominant canopy species within 20 m of the shoreline, while A. germinans dominates the canopy from 20 - 130 m. Densities of M. coffeus over 5 mm in shell length during the year 21391-21542 214 FLORIDA SCIENTIST [VOL 56 ranged from 99.4 (45.8) to 2130.2 (128.5) / m’. Since the snails tend to aggregate, numbers in individual sample quadrats were highly variable ranging from 0-511/ m’. Observations on snail foraging behavior—in the field and in the laboratory experiments M. coffeus quickly aggregated to leaves presented in experimental trials (unpublished data). During field experiment set-up, we observed up to 7 snails occurring on some experimental leaves within a few minutes of the leaves being staked out. Over the course of the field experiment, we frequently found snails on the outside of the bags being used for snail exclusion. M. coffeus grazed on some leaves within the bags by inserting its proboscis through the openings in the mesh which gave the leaves a “screen door” appearance. Field experiment on M. Coffeus grazing rates—There was virtually no area loss of bagged Control R. mangle and L. racemosa leaves over the 53 days of the study (Fig. 1). However, there was significant loss of Control A. germinans leaf area because of through-the-mesh grazing by M. coffeus (Fig. 1 and Table 1) and possibly other decomposition mechanisms. After day 10, area loss of Exposed R. mangle leaves always exceeded that of Control leaves (Fig. 1). Exposed R. mangle leaves had as much as 84% Total Area Affected by grazing at the 24 day collection. However, apparently because of snail TABLE 1. Field Experiment: One-way ANOVA and multiple comparision test summaries of among-species differences for the field experiment on detrital grazing by M. coffeus. Leaves Leaves Exposed Enclosed To M. Coffeus In Mesh Bags (Controls) 10 DAYS Area of Holes NS NS Tot. Area Affected NS Ag>Rm=Lr 24 DAYS Area of Holes Ag>Rm=Lr NS Tot. Area Affected Ag=Rm>Lr Ag>Rm=Lr 53 DAYS Area of Holes Ag>Rm NS Tot. Area Affected Ag>Rm NS NS = no significant differences at P=0.05. Among-species comparisons are read as in the following example: Rm=Ag>Lr (R. mangle and A. germinans values were statistically the same, and both were greater than the mean for L. racemosa) No. 4, 1993] 1000 500 * CE NS cE il Le 1000 “ E < E é < CE q tT 500 te q ve NS q Cc CE prey | an 1500 < NS Ce NS CE 1000 NY ae TOT. AREA Holes AFF, 10 DAYS Cc CE l * c CE fl NS E NS : Cc C aed | | = | ay * a 4 NS AREA TOT. OF AREA HOLES AFF. 24 DAYS PROFFITT ET AL._—_EXPERIMENTS ON THE MANGROVE LEAF LITTER 915 R. mangle L. racemosa A. a cS cr > NN \ AEEA aS HOLES Arr. 53 DAYS Fic. 1. Field Experiment: The Area of Holes grazed in leaves (H) and the Total Leaf Area affected by some level of grazing (T) are presented for leaves Exposed (“ ”) leaves enclosed in mesh bags to prevent macrodetritivore grazing. Values are floor and Control (“C E”) to M. coffeus grazing on the forest means and one standard error. The dark triangle next to the ordinate axis indicates the mean area of intact (ungrazed) leaves used in the treatment. “NS” indicates no significant difference between C and E, while * denotes a difference at P=0.05. 216 FLORIDA SCIENTIST [VOL 56 patchiness, the collection at day 53 had leaf area losses that were similar to the lower levels seen at day 10 (Fig. 1). Area loss of L. racemosa leaves was typically low (maximum over 53 days was 16.6 %) and highly variable among replicates (Fig. 1). Only for Total Area Affected at day 10 was there a significant difference in area loss between Exposed and Control leaves (Fig. 1). A. germinans leaf area loss was rapid but also variable (Fig. 1). Some exposed leaves had been nearly entirely consumed by day 10, while others showed no loss of area at all. The coupling of high variability in grazing and significant through-the- mesh grazin zon Control leaves, resulted in several Control vs. Exposed comparisons not being significantly different despite considerable grazing by M. coffeus (Fig. 1). Among-mangrove species differences in the rate and amount of area loss of Exposed leaves occurred by day 24 (Table 1). Area loss of Exposed A. germinans leaves was typically greater than that of R. mangle after day 10. Through-the-mesh grazing produced greater leaf area loss in Control A. germinans leaves than in other species at days 10 and 24. Area loss of Exposed L. racemosa leaves was lower than other mangrove species at day 24 (Table 1). Comparisons of L. racemosa at day 53 were not possible due to a lack of leaves initially placed in the field. Percent leaf weight loss (WL), calculated as the difference between initial laboratory air dry weights and final oven dry weights, was significantly and strongly dependent on the percent total leaf area affected by grazing (TAA) for R. mangle and A. germinans leaves. Regression equations are: R. mangle: % WL = 50.656 + 0.53 x [% TAA], R? = 0.709 A. germinans: % WL = 51.112 + 0.48 x [% TAA], R? = 0.745 L. racemosa: % WL = 66.928 + 0.53 x [TAA], R? = 0.255 All of the above regression equations were significant at the P=0.01 level. Laboratory leaf choice experiment—Leaves kept in covered Petri dishes which excluded grazing by M. coffeus had no loss of area over the course of the study. Grazing on R. mangle yellow and brown leaves occurred very rapidly (Figs. 2 and 3). Consumption of green leaves was slower initially, but by day 28, most of the leaf surface area of all types of R. mangle leaves had been grazed upon (Figs. 2 and 3). The only significant grazing on L. racemosa leaves over the first 14 days of the study was on yellow leaves (Figs. 2 and 3). Grazing on the yellow leaves proceeded rapidly and were nearly entirely consumed by day 21. L. racemosa brown leaves finally began to show substantial area loss at day 21, and green leaves at day 28. By the conclusion of the experiment at day 42, neither brown nor green leaves were as heavily grazed as yellow L. racemosa leaves, or most leaves of the other mangrove species. Very little grazing occurred on A. germinans leaves until the day 14 observation (Figs. 2 and 3). Then, grazing proceeded rapidly on all green and yellow A. germinans leaves and 2 of 3 replicate brown leaves. By day 21, over 95 % of green and about 86 % of yellow leaf area had been grazed upon, and 2 of the 3 replicate brown leaves had over 90 % of their area grazed. The third brown leaf replicate was unexplainably ignored or avoided by the snails and had very little area loss over the entire study (Figs. 2 and 3). No. 4, 1993] | PROFFITT ET AL._EXPERIMENTS ON THE MANGROVE LEAF LITTER 917 NS 1000 | ‘ie m R. mangle NS b ba 500 b b a ill tt al il l 1000 L. racemosa Geen Ek Sakae be ane N E < £ “ ad > < NS “5 500 < a Aad N ba b 7 uw 500 ) per Le TS 4 = NS | | & a md | EE =. .- § u 1500 < A. germinans NS NS NS aad] z | Ag=Lr Tot. Area Affected NS Rm>Ag=Lr Rm>Ag=Lr Day 14 Area of Holes NS NS Rm=Ag>Lr Tot. Area Affected Ag>Rm=Lr NS Rm=Ag>Lr Day 21 Area of Holes Ag>Rm=Lr NS Rm=Ag>Lr Tot. Area Affected Ag=Rm>Lr NS NS Day 28 Area of Holes Ag>Rm>Lr NS NS Tot. Area Affected Ag=Rm>Lr Ag>Rm>Lr NS Day 42 Area of Holes Ag=Rm, Rm=Lr, Ag>Lr NS NS Tot. Area Affected Rm=Ag>Lr Ag>Rm>Lr NS NS=no Sea difference in ANOVA at P=0.05. Multiple comparisons are read as follows: Rm>Ag=Lr means that area loss of R. mangle was significantly greater than that of A. germinans and L. racemosa, which were not different from one another. In among-species comparisons, by day 14 for green leaves and day 28 for yellow leaves, area loss of A. germinans was greater than or equal to that of R. mangle, and both generally exceeded that of L. racemosa (Table 2). For brown leaves, R. mangle area loss was initially (day 8) greater than for brown leaves of other mangrove species, but there were no significant differences among species at days 28 and 42 (Table 2). DIsCussION—Our work indicates that in some mangrove forests, a substantial portion of mangrove leaf organic matter may rapidly enter forest floor and estuarine food webs as M. coffeus feces and larvae, instead of being exported as leaves or leaf parts coated with bacteria, fungi, and other aufwuchs. This leaf decomposition 220 FLORIDA SCIENTIST [VOL 56 pathway, and perhaps the effects of other abundant macrodetritivores, will need to be further explored and quantified if rates of coupling of mangrove forests and estuarine food webs are to be completely understood. Historically, studies of mangrove litter decomposition in Florida systems have not quantified the effects of macrodetritivores on the rates of breakdown because workers have typically used closed litter-filled mesh bags that exclude most detritivores (Heald, 1969; Twilley et al.; 1986). When M. coffeus is abundant in mangrove basin forests, the closed mesh bag method may substantially underestimate the total rates of litter decomposition in ecosystems. The snail is frequently found in high densities in many Florida mangrove forests. We have noted large numbers of this snail in certain forests of upper Boca Ciega Bay and Southern Tampa Bay (Ft. Desoto Park). In addition, two of us (CEP and DJD) have recorded dense patches of M. coffeus in mangrove forests of Naples in southwestern Florida. The snail also is abundant in the extensive basin mangrove forests landward of the Ten Thousand Islands in southwestern Florida (Snedaker, pers. comm.) and has been observed by one of us (DJD) in large numbers in mangroves of the Indian River Lagoon near Vero Beach. Thus, we expect M. coffeus is a macrodetritivore of major importance to carbon and energy balance in Florida mangrove ecosystems. The percentage of total mangrove leaf litter that is consumed by M. coffeus is as yet unknown. However, in other parts of the world, macrodetritivores are known to be of considerable importance in leaf degredation. In some Australian mangrove ecosystems, macrodetritivores such as Sesarmid crabs consumed 71 - 79 % of the» total annual litterfall. The abundance of standing crop leaf litter on the floor of our study forest makes it seem unlikely that the overall influence of M. coffeus is as great as that of the Australian detritivores. However, our data and field observations suggest that the snail’s effects are considerable. The leaf decomposition picture is made more complex by the fact that M. coffeus shows a preference leaves of some mangrove species and certain leaf types. This also influences rates and pathways through which mangrove detritus enters forest and estuarine food webs. These observations suggests that further studies on other detritivores in the mangrove community are needed to determine if the preference for A. germinans and R. mangle leaves over those of most leaf types of L. racemosa is acommon characteristic of many mangrove detritivores or is only specific to M. coffeus. Since the laboratory leaf grazing experiment was without replacement of consumed leaves, it is not clear if later differences in grazing reflect: a) switching to available alternative food sources, b) certain leaves becoming increasingly more attractive to the snails as chemical, physical, and microbial action changed over time, or c) the eventual loss of a cuticle or certain chemical components that might make a leaf more easy to consume. As a consequence of grazing on mangrove leaf detritus, M. coffeus produce a substantial amount of fecal matter (unpublished data). Some of this is undoubtably exported to the estuarine communities, while a certain percentage probably contrib- utes to the build-up of organic matter in forest sediment. Further studies are needed No. 4, 1993] | PROFFITT ET AL—EXPERIMENTS ON THE MANGROVE LEAF LITTER 22] to assess the link between M. coffeus fecal production and the food supply and population densities of microbes and detritivores such as the fiddler crab Uca spp. that prefer smaller particles. ACKNOWLEDGMENTS—We thank the management of Veteran’s War Memorial Pinellas County Park for allowing access to the study area. We are grateful to Drs. E. Heald, S. Richie, S. Snedaker, T.]. Smith and R. Twilley for useful discussions on mangrove leaf litter decomposition and Melampus coffeus distribution around the state and the Caribbean. This study was initiated as part of a course in marine biology at St. Petersburg Junior College (St. Petersburg/Gibbs Campus) and then continued and expanded by CEP and DJD. LITERATURE CITED APLEY, M.L. 1968. Field and experimental studies on pattern and control of reproduction in Melampus bidentatus (Say). Ph.D. dissert., Syracuse University, Syracuse, NY. 154 pp. BOONRUANG, P. 1978. The degradation rates of mangrove leaves of Rhizophora apiculata (Bl.) and Avicennia marina (Forsk.) Vierh. at Phuket Island, Thailand. Phuket Mar. Biol. Cent. Res. Bull. 26: 7 pp. FELL, J.W. ae I.M. MASTERS. 1980. The associated and potential role of fungi in mangrove detrital systems. Bot. Mar. 23: 257-263. FLORES-VERDUGO, F.]., J.W. DAY, AND R. BRISENO-DUENAS. 1987. Structure, litterfall, decomposi- tion and detritus dynamics of mangroves in a Mexican coastal lagoon with an ephemeral inlet. Mar. Ecol. Prog. Ser. 35: 83-90. GOLLEY, F.B. 1960. Ecologic notes on Puerto Rican mollusca. Nautilus 73(4): 152-155. , H.T. ODUM, AND R.F. WILSON. 1962. The Structure and Metabolims of a Puerto Rican Red Mangrove Forest in May. Ecol. 43: 9-19. HEALD, E.J. 1969. The Production of Organic Detritus in a South Florida Estuary. Ph.D. Thesis, Univ. of Miami, Coral Gables, FI. HEARD, R.W. 1982. Guide to Common Tidal Marsh Invertebrates of the Northeastern Gulf of Mexico. Mississippi Alabama Sea Grant Consortium publ. No. MASGP-79-004. 82 pp. HOLLE, P.A. AND C.F. DINEEN. 1957. Life history of the salt marsh snail, Melampus bidentatus Say. Nautilus 70(3): 90-95. MALLEY, D.F. 1978. Degredation of leaf litter by the tropical sesarmid crab Chiromanthes onychophorum. Mar. Bio. 49: 377-386. MARCUS, E. AND E. MARCUS. 1965. On Brazilian supratidal and estuarine snails. Bol. Fac. Fil., Cien. Letr. Univ. S. Paulo. No. 287, Zoologia. No. 25: pp 19-82. MACNAE, W. 1974. Mangrove Forests and Fisheries. FAO/UNDP Indian Ocean Programme. IOFC/ DEV/7434. MOOK, D. 1986. Absorption efficiencies of the intertidal mangrove dwelling mollusk Melampus coffeus Linne and the rocky intertidal mollusk Acanthopleura granulata Gemlin. Mar. Ecol. 7: 105-113. Mook, M.S. 1973. Intertidal zonation of Melampus bidentatus say and Melampus coffeus L. (Gastropoda: Pulmonata). M.S. thesis, University of South Florida, Tampa, FL 65 pp. ODUM, W.E. AND E.J. HEALD. 1972. Trophic analysis of an estuarine mangrove community. Bull. Mar. Sci. Gulf. Caribb. 22: 671-738. . 1975. Mangrove forests and aquatic productivity. Pp. 129-136. In: HASLER, A.D. (ed.). Coupling of Land and Water Systems. Springer-Verlag Publ. New York, NY. PROFFITT, C.E. AND D.J. DEVLIN. 1992. Foraging Behavior and Consumption of mangrove leaf detritus by Melampus coffeus L. (Gastropod: Pulmonata). Submitted to the Proceeding of the Society of Wetland Scientists, 1992 Symposium. RICHIE, S.A. AND E.S. JOHNSON. 1991. Aedes taeniorhynchus (Diptera: Culicidae) oviposition patterns in a Florida Mangrove Forest. J. Med. Entomol. 28: 496-500. ROBERTSON, A. 1986. Leaf-burying crabs: their influence on energy flow and export from mixed mangrove forests (Rhizophora spp.) in northeastern Australia. J. Exp. Mar. Biol. Ecol. 102:237- 248. AND P.A. DANIEL. 1989. The influence of crabs on litter processing in high intertidal mangrove forests in tropical Australia. Oecologia 78: 191-198. SNEDAKER, S.A. 1992. Univ. of Miami, Miami, FL. Personal Conversation. TENORE, K.R. 1977. Food Chain Pathways in Detrital Feeding Benthic Communities: A Review, with 229 FLORIDA SCIENTIST [VOL 56 new observations on sediment resuspension and detrital recycling. pp. 37-54 in, Coull, B.C. ed., Biology of Marine Benthos, U. of South Carolina Press, Columbia, S.C. TWILLEY, R.R., A.E. LUGO, and C. PATTERSON-ZUCCA. 1986. Litter production and turnover in basin mangrove forests in southwestern Florida. Ecol. 67: 670-683. TURNER, R.E. 1977. Intertidal Vegetation and Commercial Yields of Penaeid Shrimp. Trans. Am. Fish. Soc. 106: 411-416. Florida Scient. 56(4)211-222.1993. Accepted: July 27, 1993. REVIEW U.S. Department of the Navy, Poisonous Snakes of the World, Dover Publications, Inc., New York, 1991. Pp. iii + 203. Price: $14.95 (Paperbd.). THE first edition of this work was published in 1962 as a training manual for U.S. armed forces who might encounter venomous snakes while in service overseas. The book was republished with 174 photographs (120 B/W and 54 color) to provide us with a plethora of useful information, with references, on the identification and geographical distributions of several common (and some rare) species of venomous snakes, including various cobras, rattlesnakes, pit vipers, mambas, kraits, and others. Many helpful recommendations on prevention, recognition, and treatment of snakebites from a wide selection of snake families are suggested. Descriptions of the various symptoms, and estimations of their extents (including lethality), are provided in both textual and tabular form. The medical information provided on snakebite treat- ment is useful to the layman and to qualified medical personnel; preliminary bite treatment as well as appropriate medical follow-up treatment are discussed. A table of worldwide sources of antivenin is provided for the physician’s reference. This book is a handy, helpful reference source, even to the Florida resident. Florida’s habitat is home for six common species of venomous snakes, at least four of which are encountered statewide. The material presented here is vital to one interested in the outdoors. It permits one to accurately identify a sighted specimen, using the “field guide”-style information at the beginning of each chapter. It also familiarizes one with what to expect (symptoms, etc.) if bitten, and what to do about it. Only one type of rattlesnake indigenous to northern Florida, the Canebrake rattlesnake, is not mentioned here by name. The Canebrake rattlesnake (Crotalus horridus atricaudatus) is a southern variety of the Timber rattlesnake (Crotalus horridus horridus), which is usually found in the northeastern and mid-western US., and is discussed separately. The geographic distribution provided here for the Timber rattlesnake includes northern Florida; this may account for the Canebrake variety. Anyone with an interest in herpetology, from the backyard enthusiast to the professional, would certainly benefit from the variety and extent of information provided here. It ranges from interesting to vital, depending on how immediate one’s need is for antivenin. —Charles D. Norris, Department of Chemistry, University of South Florida, Tampa, FL. No. 4, 1993] CRISTOFFER—FRUIT REMOVAL 293 Biological Sciences FRUIT REMOVAL AND INTERPLANT DISTANCE IN THE PERSIMMON, DIOSPYROS VIRGINIANA Cris CRISTOFFER 255 Osceola Street, Clermont, Florida 34711 ABSTRACT: Removal of fruits from persimmon (Diospyros virginiana), presumably by Virginia opossums (Didelphis virginiana) and raccoons (Procyon lotor), was monitored for six weeks in 1992. Trees which were distant (>3 m) from other individuals of the same species exhibited no associations among the number originally present and percent removed. By contrast, for trees less than 3 m apart, there was a significant negative correlation between percent removed and distance to nearest persim- mon. It is suggested that fruit removal by mammals differs in several respects from that by birds. NORTH America (and by implication Florida) lacks nonhuman primates, fruit bats, and other tropical mammals highly specialized for frugivory, i.e., the eating of fruit (Eisenberg, 1981). Hence, latitude notwithstanding, frugivory by Florida mammals can only occur in species with temperate zone affinities. There has, however, been little study of fruit patch selection by temperate zone mammals. A study by Bodmer (1990) of a lowland tapir (Tapirus terrestris) in the Amazon basin revealed that the animal exhibited a high rate of turning every 10 m while in forests of fruiting palms, suggesting that it was reluctant to leave patches in which it was foraging successfully. By contrast, most studies investigating frugivory in temperate zones have involved observations of birds, which are often quite mobile and move long distances between fruiting trees (Gorchov, 1988; Malmborg and Willson, 1988). Foraging strategies of nonflying mammals in temperate zones could therefore resemble those of temperate zone birds, nonflying tropical mammals, or neither. The intent of this research was to provide evidence for the hypothesis that the percent of fruit removed from a persimmon tree (Diospyros virginiana) would correlate negatively with distance to the nearest fruiting tree of the same species, because either nearby trees would be easier to locate and/or energy-conservative animals would prefer to move a short rather than a long distance. METHODS—The study site was Lake Louisa State Park in Lake County, Florida. Several natural communities are found in the park, as well as plant associations characteristic of human disturbance. Although there are many species of potential frugivores present in the park, observations of animals, tracks and scats near the study trees suggest raccoons (Procyon lotor) and Virginia opossums (Didelphis virginiana) as primary candidates. Small birds (species not recorded), an eastern cottontail (Sylvilagus floridanus), and a gray squirrel (Sciurus carolinensis) were observed one time each among the persimmon trees, but did not feed during the period of observation. Persimmon is considered especially suitable for study of fruit removal because it is favored by mammals and because it is easy to count fruits on these typically small (1-5 m) trees. Some 3] trees were monitored. The trees were numbered and tagged with labeled flagging tape, and the distance to the nearest fruiting member of the same species was measured with a tape measure. Direct observation of the frugivores would have disrupted their normal foraging behavior. 294 FLORIDA SCIENTIST [VOL 56 Fortunately fruit removal was monitored during the day rather than at night, when most frugivory probably occurred. Although removal of all the fruit by frugivorous mammals could not be ascertained, this seemed the most reasonable explanation for the loss. Fruits were never observed on the ground, hence even if fruits fell from the trees they were probably eaten by frugivores. The number of fruits was counted on each tree anda similar census was made weekly from October 11 to November 22, 1992. The study was terminated when too few fruits remained on the trees for statistical analysis (some trees had lost all fruit by the final week). RESULTS AND DISCUSSION—Kendall’s Tau-beta correlation (Agresti and Agresti, 1979) was used to assess the associations among the following variables: number of fruits originally present, number removed, percent removed, and distance to nearest fruiting tree of the same species. A scatter diagram indicated a negative trend up to about 3 m, such that an increase in percent removed was associated with a decrease in interplant distance. Beyond 3 m, however, the scatter increased to such an extent that no pattern was discernible to the eye. Since this was a scatter diagram and not a plot depicting a statistical procedure, the scatter above 3 m could not be explained as a statistical artifact. The analysis was therefore repeated, excluding trees more than 2.8 m from the nearest fruiting tree of the same species. The number of fruits present at the beginning of the study varied from 6-143/ per tree. As expected, the number removed was strongly and positively correlated (tau-beta = 0.92, p < 0.001) with number originally present; however, the correlation between the percent removed and the number present was weak and nonsignificant (tau-beta = 0.00, p = 0.96). This suggested that animals were removing approxi- mately the same number of fruit from trees with large crops as from trees with small crops. Percent loss was negatively and significantly correlated with distance to nearest fruiting tree of the same species (tau-beta = -0.31, p = 0.04). As noted above, this association was nonsignificant until the most distant trees were removed from analysis. The seven most isolated trees were more than 2.8 m from any fruiting tree of the same species and displayed no apparent patterns of association among variables. Apparently foraging animals preferentially foraged in the closest trees when the distance between trees was less than about 3 m but showed no preference among the more distant trees. Several explanations for this foraging pattern are possible. One explanation is that, at least among the trees less than 2.8 m apart, those farthest apart (i.e., closest to 2.8 m) had smaller crops of fruit than trees that were very close (i.e., much less than 2.8 m) to other fruiting trees and thus were less attractive to frugivores. The evidence does not support this explanation, however, because the association between crop size and distance to nearest fruiting tree of the same species, was positive and nonsignificant (tau-beta = 0.22, p = 0.14). Hence, if crop size was the most important variable, percent loss would have remained the same or even increased with distance. This is contrary to what was found. Another explanation is, however, more likely. The discrepancy in use between distant and nearby trees is more reasonable if we assume that the animals had more difficulty in detecting trees distant from other fruiting trees and, therefore, encoun- tered them only opportunistically. The mammals would then encounter trees distant from other fruiting trees randomly, thereby displaying no apparent pattern of use No. 4, 1993] CRISTOFFER—FRUIT REMOVAL 225 with distance. Among the trees growing closer together, the animals probably foraged in an energy-conservative fashion, feeding from the trees nearest to those they had just fed from. This would produce the negative association that was found between distance and percent removed among the closer trees. In conclusion, the removal of persimmon fruits was consistent with the hypoth- esis that frugivorous mammals tend to forage in the closest trees to conserve energy. However, it is rather surprising that a preference for the closest trees was detectable even among trees less than 3 m from the nearest fruiting tree of the same species. By contrast, the animals may have difficulty detecting or assessing more distant crops. Foraging mammals at an isolated tree might be unable to select (or perhaps even detect) the most abundant fruit crop at a distance. Thus fruit removal from distant trees could be opportunistic, unpredictable, and appear random to an investigator. The spatial patterns of fruit removal by such mammals as raccoons and opossums may be quite different from that of flying birds. ACKNOWLEDGMENTS—I thank Park Manager John Lynch for permission to do field research at Lake Louisa State Park. Several anonymous reviewers provided useful comments on the preparation of this manuscript. LITERATURE CITED AGRESTI, A. AND B.F. AGRESTI. 1979. Statistical Methods for the Social Sciences. Dellen, San Francisco. BODMER, R.E. 1990. Fruit patch size and frugivory in the lowland tapir (Tapirus terrestris). J. Zool. Lond. 222:121-128. EISENBERG, J.F. 1981. The mammalian radiations: an analysis of trends in evolution, adaptation, and behavior. The University of Chicago Press, Chicago. GORCHOV, D.L. 1988. Does asynchronous fruit ripening avoid satiation of seed dispersers?: a field test. Ecology 69(5):1545-1551. MALMBORG, P.K. AND M.F. WILLSON. 1988. Foraging ecology of avian frugivores and some consequences for seed dispersal in an Illinois woodlot. Condor 90:173-186. Florida Scient. 56(4):223-225 Accepted: August 10, 1993. 996 FLORIDA SCIENTIST [VOL 56 Chemical Sciences DIMETHYL SULFOXIDE CATALYZED RACEMIZATION OF ASPARTIC ACID Grecory P. Cusano, MIRTHA CHAVEZ, Dieco TorRES, AND GEORGE H. FISHER* Department of Chemistry, Barry University, Miami Shores, FL 33161 ABSTRACT: Aspartic acid (Asp) is one of the fastest racemizing amino acids due to stabilization of the intermediate carbanion by the electron withdrawing beta carboxyl group. Aqueous solutions of free L- aspartic acid were heated at 100 for various lengths of time in the presence of dimethyl sulfoxide (DMSO) and without DMSO (controls). The amount of racemization increased as a function of time in both sets of samples, but to a greater extent in the presence of DMSO. Racemization also increased as the percent of DMSO in solution increased. It is postulated that the increased racemization in DMSO is due to acombination of the higher dielectric constant of DMSO and symmetrical solvation of the intermediate carbanion by DMSO, thus further stabilizing the carbanion. Loss of the alpha proton of an amino acid gives a carbanion intermediate which can be reprotonated from either side, thus resulting in racemization (change in configuration from L to D or vice versa). Aspartic acid (Asp) is one of the fastest racemizing amino acids due to stabilization of the intermediate carbanion by the electron withdrawing effect of the side-chain beta carboxyl group. For several years we have investigated the phenomenon of in vivo racemization of aspartic acid in human brain (Man et al., 1983, 1987), in myelin basic protein (Fisher et al., 1986), and neuroproteins of Alzheimer brains (Fisher et al., 1991, 1992; Payan et al., 1992). Racemization can be catalyzed by pH, temperature, metal chelation, and solvation (Bada, 1982). In our current research we are investigating the effect of solvation on racemization using dimethyl sulfoxide (DMSO). DMSO is a polar, aprotic solvent capable of solvating both cations and anions. Cram and coworkers (1961) previously showed the catalytic effect of DMSO on the base promoted racemization of 2-methyl-3-phenylpropionitrile. We were interested in seeing if DMSO has a similar effect on the racemization of aspartic acid. MATERIALS AND METHODS—Pure L-aspartic acid was obtained from Sigma Chemicals. Reagent grade (Mallinckrodt) DMSO was used as received. Ten tubes were set up containing 100 WL of an aqueous solution of 0.1 mg/mL L-Asp (70 nmole, pH 3.5). To five of the tubes were added 5 pL (75 nmole) of DMSO. The tubes were sealed under vacuum. Two tubes (one with DMSO and one without DMSO) were left unheated. The remaining sets of two tubes (with and without DMSO) were heated at 100°C for 1, 2, 5, and 7 days, respectively. In a second experiment, the catalytic effect of increasing percentages of DMSO was studied. Six tubes were made up of aqueous L-Asp(0.1 mg/mL)-DMSO solutions, containing 0, 25, 50, 75, and 90% (v/v) DMSO, respectively. These tubes were sealed under vacuum and heated for 7 days at 100°C. After heating, all the tubes were opened, evaporated by a combination of rotary evaporation and lyophilization, and the resulting residue was dissolved in 1 mL of water. Since an automatic polarimeter *To Whom reprint requests should be sent No. 4, 1993] CUSANO ET AL_—DIMETHYL SULFOXIDE 227 was not available, we used a direct method of measuring the percent D-Asp by high performance liquid chromatography (HPLC). An aliquot of each sample was derivatized by treating the D/L-Asp enanti- omers with o-phthaldialdehyde (OPA) and enantiomerically pure N-acetyl-L-cysteine to form a pair of fluorescent diastereomeric (L-L and D-L) N-alkyl-2-thioalkyl isoindole derivatives, which are separable by HPLC on a reversed-phase C-18 column using a methanol-sodium acetate gradient with detection bya fluorometer (Aswad, 1984). Quantitative determination of the percent D-Asp was accomplished by area measurement of the respective peaks using a computing integrator. Solutions of known D/L-Asp ratios were run as standards. RESULTS AND DISCUSSION— Results of the first experiment are shown in Table 1 where the percent D-Asp formed is reported as a function of the number of days of heating at 100° for controls (no DMSO) and the samples with DMSO. Two effects are noted from these data. First, the amount of racemization increases in both sets, presumably as a result of the temperature and time of heating. Secondly, the amount of racemization is greater in the samples containing DMSO, showing the catalytic effect of DMSO. Figure 1 shows a computer generated least-squares fit of percent D-Asp plotted versus days of heating. From the slopes of these lines the rate constants for racemization are calculated to be 0.64% D/day for the control L-Asp/ H,O samples and 0.70% D/day for the L-Asp/DMSO-H,O samples. The ratio of the rate constants of DMSO to control is 1.1, signifying that racemization is 1.1 times faster in the DMSO-water solutions than in water alone. This rate difference increases significantly with increasing percent DMSO, as shown by the data in the second experiment below. The effect of varying concentrations of DMSO on racemization is shown in Figure 2 where the percent D-Asp is plotted versus the mole percent DMSO, using the data from Table 2. For a given amount of time that the samples were heated, the amount of racemization increases as the percentage of DMSO the increases. At 70 mole percent DMSO racemization is eight times greater than in water alone. It is not totally surprising that aspartic acid racemizes faster in DMSO. Cram and Nielsen (1961) have pointed out that dimethyl sulfoxide, with its much higher TABLE 1. Racemization of aspartic acid as a function of time heated at 100°C with and without DMSO %D-Asp + S.D. Number of days heated at 100°C Controls (L-Asp) L-Asp + DMSO 0 0.0 0.0 1 0.13 £0.01 1.6 +£0.15 Z 2.6 £0.38 2.8 £0.32 5 3.1 +0.001 3.3 £0.001 i 4.6 1.03 5.8 £0.33 228 FLORIDA SCIENTIST [VOL 56 L-Asp/DMSO —~ ToD OX ) 1 2 3 4 5 6 7 8 Days of Heating FIG. 1. Racemization of aspartic acid as a function of time of heating at 100°C with (A) and without (g) DMSO. 25 20 % 15 D A 10 Ss p 5 0 0 10 20 30 40 50 60 70 80 Mole % DMSO FIG. 2. Racemization of aspartic acid as a function of the mole percent of DMSO. No. 4, 1993] CUSANO ET AL.—DIMETHYL SULFOXIDE 299 TABLE 2. Racemization of aspartic acid as a function of percent DMSO after 7 days at 100°C Volume % Mole % Relative DMSO DMSO %D-Asp Racemization 0 0 2.83 £0.04 1.00 25 7.8 4.74 +0.15 1.67 50 20.2 8.24 +0.13 2.91 (&) 43.2 21.04 +1.30 7.43 90 69.5 23.75 £4.37 8.39 dielectric constant, is a better dissociating solvent. Dissociation reactions in general occur faster in solvents with high dielectric constants. The results seen here could merely be the effect of having a solvent with a higher dielectric constant. Alternatively, the results could reflect specific interactions between the DMSO and aspartic acid, as pointed out by Cram and Nielsen (1961). Dimethyl sulfoxide, although capable of solvating anions, is not a proton donor. Symmetrical solvation of an anion by DMSO would then lead to racemization. Therefore, we believe that the catalytic effect of DMSO on the racemization of aspartic acid may also be due to symmetrical solvation and stabilization of the resulting carbanion by the DMSO, as depicted in Figure 3. The partially positive sulfur coordinates with the negative carbanion, while the partially negative oxygen coordinates with the abstracted proton, holding it in such a way that reprotonation can occur from either side of the carbanion, thus resulting in racemization. Increased stabilization of the carbanion would then result in increased racemization. As mentioned previously, D-Asp has been found in metabolically stable living proteins. It has been hypothesized that DMSO will not symmetrically solvate any intermediate carbanion formed within the chiral environment of all the protein- bound L-amino acids. The significance of this is that in the presence of DMSO a D- Asp residue might be racemized back to the normal L-Asp as a way of repairing damaged proteins. Work is currently underway to test this hypothesis using proteins known to contain a high percentage of D-Asp. o) os é H te e = 50 0° A A st O + (CH3))S° © °S(CH3)2 ty Fig. 3. Symmetrical solvation of intermediate carbanion by DMSO. 230 FLORIDA SCIENTIST [VOL 56 ACKNOWLEDGMENTS—We gratefully acknowledge financial support from NIH Minority Biomedical Research Grant #GM 45455, from a Cottrell College Science Award of the Research Corporation, and from the Barry University Alzheimer’s Research Fund. LITERATURE CITED AswaD, D.E. 1984. Determination of D- and L-aspartate in amino acid mixtures by high performance liquid chromatography after derivatization with a chiral adduct of o-phthaldialdehyde. Anal. Biochem. 137:405-409. BADA, J.L. 1982. Racemization of amino acids in nature. Interdis. Sci. Rev. 7:30-46. CraAM, D.J. AND W.D. NIELSEN. 1961. Electrophilic substitution at saturated carbon. VIII. Mixed solvents and steric course. J. Amer. Chem. Soc. 83:2174-2178. , B. RICKBORN, C.A. KINGSBURY, AND P. HABERFIELD. 1961. Electrophilic substitution at saturated carbon. XII. Solvent control of rate of acid-base reactions that involve the carbon- hydrogen bond. J. Amer. Chem. Soc. 83:3678-3687. FISHER, G.H., N.M. GARCIA, I.L. PAYAN, R. CADILLA-PEREZRIOS, W.A. SHEREMATA, AND E..H. MAN. 1986. D-Aspartic acid in purified myelin and myelin basic protein. Biochem. Biophys. Res. Commun. 135:683-687. , A. D’ANIELLO, A. VETERE, L. PADULA, G.P. CUSANO, AND E.H. MAN. 1991. Free D- aspartate and D-alanine in normal and Alzheimer brain. Brain Res. Bull. 26:983-985. , I.L. PAYAN, S.-J. CHOU, E.H. MAN, S. CERWINSKI, T. MARTIN, C. EMORY AND W.H. FREY, II. 1992. Racemized D-aspartate in Alzheimer neurofibrillary tangles. Brain Res. Bull. 2812 7-13. MAN, E.H., M. SANDHOUSE, J. BURG, AND G.H. FISHER. 1983. Accumulation of D-aspartic acid with age in human brain. Science 220:1407-1408. , G.H. FISHER, I.L PAYAN, R. CADILLA-PEREZRIOS, N.M. GARCIA, R. CHEMBURKAR, ARENDS, AND W.H. FREY, II. 1987. D-Aspartate in human brain. J. Neurochem. 48:510-515. PAYAN, I.L., G.H. FISHER, S.-J. CHOU, E.H. MAN, C. EMORY, AND W.H. FREY, II. 1992. Altered aspartate in Alzheimer neurofibrillary tangles. Neurochem. Res. 17:187-191. Florida Scient. 56(4):226-230.1993. Accepted: August 24, 1993. No. 4, 1993] SMITH ET AL—ADDITIONS TO THE HERPETOFAUNA 231 Biological Sciences ADDITIONS TO THE HERPETOFAUNA OF EGMONT KEY, HILLSBOROUGH COUNTY, FLORIDA Lora L. SmitH”), RICHARD FRANZ”), AND C. KENNETH Dopp, JR.’ “Florida Museum of Natural History, University of Florida, Gainesville, FL 32611; ®)National Biological Survey, 412 N.E. 16th Avenue, Room 250, Gainesville, FL 32601 ABSTRACT: The green treefrog (Hyla cinerea) and mole skink (Eumeces egregius) are recorded for the first time, and the presence of black racers (Coluber constrictor) is confirmed on Egmont Key, Hillsborough County, Florida. The unique combinations of characters found in Eumeces egregius and Coluber constrictor suggest that these elements of the herpetofauna may have a longer association with this island than previously thought. We note biogeographical similarities with populations in the Florida Keys. FRANZ and co-workers (1992) listed 10 species of amphibians and reptiles from Egmont Key, located in Tampa Bay, Hillsborough County, Florida. Another four species, previously recorded by collections in 1869-1870 and 1904, apparently were extirpated in the interim. Our recent investigations substantiated reports of the black racer (Coluber constrictor) on Egmont and led to the discovery of the green treefrog (Hyla cinerea) and mole skink (Eumeces egregius) on the island. These records necessitate further biogeographic interpretation beyond that offered by Franz and co-workers (1992). Hyla cinerea—Franz and co-workers (1992) reported Hyla squirella from Egmont Key. In June 1993, we heard several groups of green treefrogs calling from the vicinity of the Big Swale following heavy rain showers. Both H. squirella and H. cinerea are abundant on the surrounding mainland and easily could have been transported to the island. Coluber constrictor—Although Franz and co-workers (1992) had not observed the black racer on Egmont Key, they listed it as a current resident based on sightings by state park rangers. The species also had been collected in 1869-1870 and in 1904. We encountered five black racers on the southern half of the island in October 1992, April, May, and June 1993. Two snakes from the April trip were captured, photo- graphed, and released. EGMONT SPECIMENS: Female: Pilot’s Compound, 24 April 1993, SVL. 685 mm, TL 965 mm, Wt 110 g. Male: Cross Island Road, 23 April 1993, SVL 710 mm, TL 1018 mm, Wt 137 g. Both individuals are black above and grey below; skin between the scales on the posterior half of the body has a tan or brownish appearance. Tan coloration occurs on the preocular, loreal, nasal, and rostral scales of the head. Tan also invades the upper portions of supralabials 1-3 and the lower surfaces of the prefrontals, but there is no tan color on the chin as in Coluber constrictor helvigularis 932, FLORIDA SCIENTIST [VOL 56 from the Apalachicola area in west Florida. The color arrangement is very similar to that of Coluber constrictor paludicola from the Everglades and Florida Keys. Supralabials are mostly white, as are the chin and throat areas. The white on the belly extends distally for four ventral scales before becoming clouded with grey; clouded appearance becomes uniformly grey by the 22nd ventral scale. Lower portions of the iris are a subdued dark orange to brown, whereas the top is brighter reddish-orange. The second and the anterior edge of the third supralabials in the Egmont snakes are in contact with the loreal scale; loreal scale is longer than wide; both of these features are more similar to northern peninsula specimens than to snakes from extreme south Florida. Both snakes were generally docile when they were handled during the photographic session, although the male attempted to bite when it was originally caught. Eumeces egregius—We collected one specimen of mole skink (UF84031) in coarse, white, beach sand under a large piece of wood at the south end of the Tampa Bay Pilot Association’s compound on 22 February 1993. The area north of the border fence was regularly mown lawn grasses, whereas the area immediately south of the fence where the specimen was collected was composed of sea oats (Uniola paniculata) and scattered Australian pine (Casuarina equisetifolia). EGMONT SPECIMEN: Male (Figure 1), SVL 54 mm, TL 115 mm (including regenerated tail). Head length, 8.3 mm. Axilla-groin length, 35.5 mm. Scale rows at midbody, 22. Supralabial scales, 6 on one side, 7 on the other side. Midventral scales, 63. Two white dorsolateral stripes narrowly edged with dark pigment extend from about 5 scales behind the head distally for a distance of 24 scales, becoming obscure near midbody; white stripes are narrow (about 1/3 scale wide), parallel, not diverg- ing; a dark stripe, bordered by light stripes, occurs on the dorsolateral aspects of the original tail (absent on the regenerated portions); tail stripes appear to be continu- ations of the anterior dorsolateral stripes. A very narrow, pale, lateral stripe extends from the supralabial area, through the ear, to the anterior insertion of the front leg. In life, this lizard was silvery grey with a slight pinkish cast to the skin, which was slightly more pronounced on the tail. There was no evidence of the orange, reddish, violet, or blue colorations found in other populations of E. egregius. The coloration of this lizard was very reminiscent of that of Neoseps reynoldsi. The Egmont specimen shows a unique combination of features that makes it difficult to assign to a subspecies. Its unusual coloration is unlike any other E. egregius we have seen. Based on the range map presented by Mount (1968), the peninsula mole skink (E. e. onocrepis) is the subspecies that occurs in the vicinity of Tampa Bay. Unlike the Egmont specimen, however, this subspecies has widening and diverging, rather than parallel, dorsolateral stripes. It also has 20 scale rows at midbody, instead of Egmont’s 22, and 53-62 (mean, 57.91) midventral scales, instead of 63 (see McConkey, 1957; Mount, 1965). The count of 63 midventral scales for the Egmont specimen is similar to the count found in the holotype of E. e. insularis from the Cedar Keys and falls within the range listed by McConkey (1957) for E. e. similis (56-65, mean 60.36); however, both of these populations have two fewer scale rows at midbody (20) (see Mount, 1965). The midbody scale row and midventral scale No. 4, 1993] SMITH ET AL.—ADDITIONS TO THE HERPETOFAUNA 933 FIG. 1. Male Eumeces egregius from Egmont Key, Hillsborough County, Florida (UF 84031). 934 FLORIDA SCIENTIST [VOL 56 counts are similar to those found in E. e. egregius of the Florida Keys (McConkey, 1957). This unique specimen suggests that the island’s mole skink population may represent an undescribed subspecies, but additional material is necessary before its taxonomic status can be ascertained. DIscussloN—Further interpretations on the origins of Egmont Key’s Herpetofauna—The presence of distinctive Eumeces egregius and Coluber constric- tor on Egmont Key suggests a long association with the island. This is in contrast to the predictions by Franz and co-workers (1992), who suggested that the fauna could have arrived more recently via human transport or crossed over seawater to reach the island. The presence of the fossorial E. egregius could equally suggest a residual herpetofaunal component on Egmont that persisted from a period when the island was connected to the mainland during lowered sea levels. The similarities in scale counts between E. egregius from Egmont Key and the Florida Keys suggest another example of Christman’s (1980, pp. 214-215) Lower Keys-North Florida Pattern. He listed Storeria dekayi, Coluber constrictor, and Opheodrys aestivus as demonstrating this pattern. The two specimens of C. constrictor that we photographed from Egmont Key showed the extensive tan facial coloration of the south Florida racers, but not the pale venter, which was mentioned by Christman (1980). It is unclear at this time whether the Egmont Coluber constrictor is part of the Lower Keys-North Florida Pattern, noted by Christman (1980), or represents a unique island pattern. More collections from the islands along the west coast of Florida are necessary before generalizations concerning these faunas should be attempted. ACKNOWLEDGMENTS—We wish to thank the Florida Park Service and Cameron Shaw (manager of Chassahowitzka National Wildlife Refuge) for permission to work on Egmont Key; Robert Baker (Park Manager, Egmont Key State Park) and the park rangers who gave of themselves and their time in support of our activities on the island; David Auth for verifying scale counts on the Eumeces egregius specimen from Egmont Key; Max Nickerson and Sam R. Telford Jr. for reviewing the manuscript. LITERATURE CITED CHRISTMAN, S. P. 1980. Patterns of geographic variation in Florida snakes. Bull. Florida State Mus. 25(3):157-256. FRANZ, R., C. K. DODD, Jr., AND A. M. BARD. 1992. The non-marine herpetofauna of Egmont Key, Hillsborough County, Florida. Florida Sci. 55(3):179-183. . MCCONKEY, E. H. 1957. The subspecies of Eumeces egregius, alizard in the southeastern United States. Bull. Florida State Mus. 2(1):13-23. MOUNT, R. 1965. Variation and systematics of the scincoid lizard, Eumeces egregius (Baird). Bull. Florida State Mus. 9(5):183-213. 1968. Eumeces egregius (Baird). Mole Skink. Cat. Amer. Amphib. Rept. 73.1-73.2. Florida Scient. 56(4):231-234.1993 Accepted: Sept. 21, 1993. No. 4, 1993] SMITH—TIDAL AND WIND DRIVEN TRANSPORT 935 Oceanographic Sciences TIDAL AND WIND-DRIVEN TRANSPORT BETWEEN INDIAN RIVER AND MOSQUITO LAGOON, FLORIDA NED P. SMITH Harbor Branch Oceanographic Institution 5600 North U.S. Highway 1, Fort Pierce, Florida 34946 ABSTRACT: Observations of current speed and direction are combined with wind measurements to describe flow through Haulover Canal, connecting Mosquito Lagoon with Indian River lagoon. During a 73-day study period in late summer and autumn, 1989, the instantaneous flow varied from about +30 to -30cms" and reversed in response to both tidal and wind forcing. The mean flow at the study site was about 4 cm s" into Indian River lagoon. Assuming logarithmic and parabolic current profiles in the vertical and across the canal, respectively, measurements are translated into a mean volume transport of about 20m? s". The tide-induced residual volume transport alone was 0.06 m’s"'. Tidal currents carried water the full length of the canal on each half tidal cycle, except under neap tide conditions. Wind forcing over time scales on the order of one week was the most effective inter-lagoon exchange mechanism. Spectral analysis suggests that under normal seasonal wind conditions the net transport is from Mosquito Lagoon to Indian River lagoon during spring and summer months and back into Mosquito Lagoon during fall and winter months. THE movement of water through a coastal lagoon and the exchange of lagoon water with ocean water through inlets are of primary importance for maintaining water quality in the interior of the lagoon. A long-term net movement of water through the interior of a lagoon can be a direct response to fresh water entering on the landward side, or an indirect response associated with horizontal density gradients. It can be a tide-induced residual transport (Robinson, 1983; Smith, 1990a), or it can occur as a wind-driven transport (Smith, 1990b; 1990c). Tide- induced residual transport also arises within a lagoon served by two or more inlets when the inlets are of different sizes or when tidal conditions differ at the mouths of the inlets (van de Kreeke, 1976). The exchange of water between two directly connected coastal lagoons is a special case not involving inlets directly. It serves more as a dilution mechanism within the lagoon system than as a flushing process. In some instances, however, such as in regions with poor natural flushing, this may constitute the only mechanism for exchange and renewal. Wong and DiLorenzo (1988) have described the tidal and subtidal exchanges of water between Indian River Bay and Rehoboth Bay (Dela- ware) as a response to coastal pumping. The response to local wind forcing was not considered. Indian River lagoon is a shallow bar-built estuary lying along the Atlantic coast of central Florida (Fig. 1). The lagoon is 196 km long and 2-4 km wide. Water depth is generally 1-3 m. The Atlantic Intracoastal Waterway, extending nearly the entire length of the lagoon, is dredged to a depth of 3.5 m. Three jettied inlets, all located 236 FLORIDA SCIENTIST [VOL 56 : ‘ PONCE DE LEON INLET SN : 80°30” 28°50’ CAPE CANAVERAL a a pa 0 10 20 30 40 50Km 80°45’ Z | SEBASTIAN INLET FIG. 1. Map showing the Haulover Canal study area between Mosquito Lagoon and Indian River lagoon. The solid circle shows the location of the weather station southeast of the study site. No. 4, 1993] SMITH—TIDAL AND WIND DRIVEN TRANSPORT 237 in the southern half of the lagoon, are commonly used to define three sub-basins. The lagoon is microtidal (Smith, 1987), with M, amplitudes 0-5 cm in the northern sub- basin (north of Sebastian Inlet), 5-10 cm in the central sub-basin (between Sebastian Inlet and Ft. Pierce Inlet) and 10-15 cm in the southern sub-basin (between Ft. Pierce Inlet and St. Lucie Inlet). In much of the lagoon, the nontidal rise and fall of water level is similar or greater in magnitude than the rise and fall of the tide (Smith, 1986). Tidal currents can be significant in the immediate vicinity of the inlets (Smith, 1990d), but amplitudes decrease markedly in the interior of the lagoon. In the northern half of the northern sub-basin, the lagoon is virtually tideless, and wind forcing is primarily responsible for currents and changes in water level. Even in the southern sub-basin, where tidal and nontidal water level variations are similar in magnitude, the wind-driven circulation dominates tide-induced residual flow over time scales in excess of a few days (Smith, 1990b). The Indian River lagoon system is defined to include Mosquito Lagoon and Banana River lagoon, as well as Indian River lagoon because all are connected directly by a system of dredged canals. These canals serve a useful purpose, partly because of the volume of water they transport, and partly because they permit an exchange of plant and animal life. Exchanges between Banana River and the northern sub-basin of Indian River lagoon have received some attention (Yusof, 1987), because Banana River is completely isolated from direct exchanges with Atlantic shelf waters. Exchanges between the northern sub-basin of Indian River lagoon and Mosquito Lagoon through Haulover Canal have not been investigated. Flushing studies of Indian River lagoon that have considered the lagoon as a whole have assumed that the northern sub-basin of the lagoon is closed (Williams, 1985; Sheng et al., 1990; Smith, 1993). The purpose of this study is to estimate volume transport through Haulover Canal using current speed and direction measurements from a study conducted in late summer and fall, 1989, to compare the relative magnitudes of tidal and wind-driven exchanges, and finally to examine the assumption that the northern sub-basin is closed at its northern end by comparing the mean volume transport through the canal with the volume of the receiving body. DaTA—Current speed and direction were recorded hourly during a 73-day time period at a study site near the midpoint of Haulover Canal. The canal connects the northeastern end of Indian River lagoon with the southwestern shore of Mosquito Lagoon. The study site is isolated from direct exchanges with Atlantic shelf waters. From Haulover Canal south-southeast to Sebastian Inlet is approximately 110 km; it is 43 km north-northwest to Ponce de Leon Inlet. The canal is 2 km long, and in places it is cut through rock. As a result, the cross-section is very nearly rectangular, with steeply-sloping sides that at some points form ledges. At the study site, the canal is 58 m wide. Water depth was measured at five locations to establish the local cross-sectional area. Away from the steeply sloping sides, water depth was within 10 cm of a mean of 5 m. Based on measurements made on two occasions, and given the seasonal deviation from the multi-annual mean (Smith, 1986), a depth of 5 m was used to calculate volume transport. Currents were recorded 1 m above the bottom from August 17 to October 30, 1989, using a General Oceanics Model 2010 film recording inclinometer fitted with oversized fins. Inclination angles were read to the nearest I’, and the precision of the resulting current speeds is approximately +1 cm s" for the range of inclination angles found for this study site. Current vectors were decomposed into along-channel (045- 225°) and across-channel components; only the along-channel components were retained for further analysis. Data were smoothed with a 3-weight “hanning” filter to reduce high-frequency variability without influencing tidal and lower frequencies, then sub-sampled to provide hourly values. 938 FLORIDA SCIENTIST [VOL 56 Wind data were obtained from the Kennedy Space Center Shuttle Airport, 15 km south-southeast of the study site. Hourly wind speeds and directions were read to the nearest 0.5 m s! and 10°, respectively. METHODS—The estimation of inter-lagoon volume transport began with the calculation of the depth-averaged along-channel current speed at the study site and the cross-sectionally averaged flow. Assuming that local wind forcing is minimal in the protected waters of the canal, and that the vertical current profile is logarithmic, the current at any depth, u,, can be given by Oe ee: ne a ae (1) “~o where u, is the friction velocity, z is height above the bottom, k is the Karman constant, and z, is the roughness length. When u, is the measured current at a height of 1 m above the bottom (u,,,), and using a roughness length of 0. 2-em for a mud bottom with shell fragments, one may solve for the faction velocity. The locally depth-averaged current speed, U, is then obtained from 12 tinge Jee maka 2 ac} (2) ~o where Az is the difference between the total water depth, Z, and the roughness length. This equation can be used in two ways. Water level measurements are not available from the time and place of the study, but predicted tidal currents and water levels can be used to evaluate a depth-averaged tidal current speed. Second, in view of the small tidal amplitudes in Haulover Canal (Smith, 1987), water level can be approximated by a constant value, at least over time scales on the order of a few days, and the depth-averaged current speed at the study site can be obtained as a function of the measured reference current: U = 1.099u,,, (3) Surface currents measured laterally across the channel at the study site for both flood and ebb conditions supported the selection of a parabolic relationship of the form 2y-W) u(y) =U] 1-( 24] I ; (see Kjerfve, 1976) where U_, , is the depth-averaged along-channel current speed in mid channel, and W is the width of the canal. It follows that the measured flow 5 m from the bank is about 32% of the mid-channel maximum speed. Conversely, the mid-channel maximum is 3.2 times the speed at the study site. Integrating from shore to shore, one finds that the laterally-averaged along-channel flow will be 2/ 3rds of the mid-channel maximum. Combining the above, the cross-sectionally averaged along-channel current speed can be expressed as function of either the measured current or the local depth-averaged value: U,, = 2.32u,,, = 2.11U. (5) Volume transport for each hour of the study, as well as the cumulative net value through the n" hour, was then obtained from the product of the cross-sectionally averaged current speed and the 290 m? cross-sectional area of the canal at the study site, resulting in V, =6.74u_,, =6.13U, (6) where V,, will be in m? s! when u_,, or U is in cm s" Oheancd wind vectors were converted to wind pane using the linear wind stress coefficient recommended by Wu (1980). Wind-driven nontidal exchanges were investigated by low-pass filtering the hourly along-channel current speeds, using the “D39” filter described by Groves (1955). The filter has a half-power point at a periodicity of about 55 hours. The filtered current was then decimated (retaining every 10th value) and compared to similarly treated components of the wind stress vectors. Filtering and sub-sampling time series restricted the analysis to periodicities longer than about 60 hours. Time series of wind stress components can be compared to time series of along-channel transport No. 4, 1993] SMITH—TIDAL AND WIND DRIVEN TRANSPORT 939 TABLE 1. Statistics associated with tidal exchanges through Haulover Canal. Amplitudes of measured reference speeds (A_,), the cross-sectionally averaged current speeds, A. and cross- sectionally averaged volume transport (V__) are incm s’, cm s! and m’® s", respectively; local phase angles (Kk) are in degrees. Tidal excursions for the cross-sectionally averaged current speed and volume transport are in km and thousands of m’, respectively. Constituent AG A.. V.. K jae Kies M, 8.1 18.8 54.6 235 2.68 ee S, 1.6 30 10.8 237 0.51 148.5 N, 2.1 4.9 14.2 229 0.71 206.0 K. 3.0 7.0 20.2 325 1.92 593.9 O, 1.9 4.4 12.8 071 1.30 378.7 Ps 1.0 2.3 6.7 325 0.63 184.8 to investigate wind forcing. Wind-driven transport can then be compared to the volume of adjacent portions of the lagoon to determine the significance of the resulting exchanges. Spectral analysis (Little and Schure, 1988) paired along-channel flow through Haulover Canal with four wind stress components. Wind stress vectors were decomposed into north-south, northeast-southwest, east-west and southeast- northwest components, then low-pass filtered and decimated as described above. Coherence spectra established the statistical significance of wind forcing, then phase spectra determined the in-phase or out-of-phase relationship of winds and currents. The transfer function quantified the relative magni- tudes of the coherent parts of the two time series at a given periodicity. Harmonic analysis of measured or calculated current speed or volume transport (Dennis and Long, 1971) provided the harmonic constants (amplitudes and local phase angles) of the principal tidal constituents (Table 1). A series of analyses of 29-day time series, offset in time by one week, produced a collection of harmonic constants for each tidal constituent that were vector-averaged, as suggested by Haurwitz and Cowley (1975). Given the amplitude, A, of a tidal constituent, the tidal excursion, E.,, was calculated by integrating over one-half tidal cycle: T/2 | hag AT Eek | inl 2 => (7) where T is the tidal period. The number of assumptions involved in the translation of the current speed measured at the study site to the volume transport through the local cross-section prompted a sensitivity analysis. The choice of z,, for example, influences both the friction velocity and the estimation of the vertically averaged current speed. Varying z, from 0.2 cm (Heathershaw and Langhorne, 1988) to either 0.1 or 0.3 cm results in an approximately 7% decrease or increase, respectively, in u,, but U varies by only 1%. Volume transport estimates are very sensitive to the assumed lateral distribution of along-channel flow. While drogue data collected at five locations across the canal at the study site supported the parabolic distribution given by (4) on one occasion, in another case the observed distribution was somewhat more “pointed” in mid channel. Using an alternate expression of the form W 3 U(y) = (Ue k= aS ie 2 240 | FLORIDA SCIENTIST [VOL 56 110 Nel 100 90 80 70 WATER LEVEL (cm) 100 90 80 70 S) 11 is at 4 (2 20 28 6 14 22 30 AUG SEF t OCT FIG. 2. Hourly water levels recorded in Haulover Canal by the National Ocean Service, August 3 through October 28, 1970. Hourly values are shown in the bottom part of the plot; low-pass filtered values are shown in the top part. results in along-channel flow increasing linearly from zero at the boundary to U__in mid channel. Integrating this across the width of the channel, one obtains U,, =3.19 U.., (9) which is 37% larger than the value obtained with the parabolic distribution. While a linear variation is neither supported by theory nor indicated by observation, it provides an upper limit that is useful for quantifying sensitivity. Historical water level records suggest that at this location, water depth, Z, varies little over time scales ranging from hours to days. The rise and fall in water level is primarily over seasonal time scales. Figure 2 shows an 87-day water-level record collected by the National Ocean Service in 1970, but during the same time of year as the present study. Their study site was also in Haulover Canal, about 600 m from where the current meter was moored. The most significant feature of the plot is the seasonal rise of about 30 cm that occurs between early September and mid October. Tidal fluctuations are on the order of a few centimeters at most, which is less than 0.5% of the water depth. Low-frequency nontidal variations in water level are on on the order of 5-10 cm, or 1-2% of the total water depth, but it is unlikely that nontidal water level fluctuations are consistently related to the direction of nontidal flow through the canal. At the study site in the middle of the canal, the rise in water level should be roughly the same for No. 4, 1993] SMITH—TIDAL AND WIND DRIVEN TRANSPORT 94] Current Speed (cm s”) Cumulative Net Displacement (Km) 40 20 100 0 -100 -200 -300 177 24 ~~ 331 7 14 21 28 5 12 uw) 245) AUG SEP OCT FIG. 3. Instantaneous along-channel flow (top plot) and cumulative net along-channel displacement, August 17 through October 30, 1989. Positive values indicate flow toward the northeast, from Indian River lagoon into Mosquito Lagoon. 242 FLORIDA SCIENTIST [VOL 56 a wind-induced set-up in Indian River (forcing flow toward the northeast) as for a set-up in Mosquito Lagoon (forcing flow toward the southwest). Varying water depth by +30 cm during the study changes the volume transport into Indian River by less than 1%. Thus, the error in the Haulover Canal volume transport calculations is due almost entirely to deviations from the assumed parabolic distribution of the flow across the channel. This has not been quantified, but it is felt that transport calculated from (6) is within 10% of the true value. This is smaller than the error estimated by Kjerfve, and co-workers (1981) for a single station. Bottom topography at their study site was substantially more complex, however, and they could not justify the assumption of a parabolic flow pattern across the channel, such as that used in (4). RESULTS—Results of the Haulover Canal study are subdivided into two parts. The first deals with measured currents and the relationship to local wind forcing. The second deals with the volume transport estimates and their physical significance. Current measurements—The upper part of Figure 3 is a plot of along-channel current speed vs. time for the 73-day study. Positive speeds indicate flow toward the northeast, into Mosquito Lagoon. Highest current speeds are about +30 cm s". Reversals in direction are frequent, and flow in either direction does not persist for more than 3-4 days in most cases. The mean flow recorded by the current meter during the study was -4.1 cm s’, however the flow into Indian River lagoon during the final week of the study averaged about -30 cm s". Tidal oscillations do not stand out distinctly or consistently. Harmonic analysis of overlapping 29-day segments produced the amplitudes (A__,) and local phase angles (x) for the six semidiurnal and diurnal constituents listed in Table 1. Other tidal constituent amplitudes were within the precision of the current meter and are considered unreliable as well as physically insignificant. Using (5), amplitudes given in Table 1 can be converted to values representing the mean cross-sectional flow. For example, the 8.1 cm s' M, amplitude at the study site becomes 18.8 cm s" for the cross-section as a whole, and this is an important difference. The tidal excursion associated with the flow at the study site is 1.2 km, or just over half of the length of the canal. The tidal excursion calculated from the cross- sectional mean flow is 2.7 km, indicating a tidal excursion about 35% greater than the length of the canal. As much as 40% more and 40% less moves through the canal under spring and neap tide conditions, respectively. The fraction of the water that is carried from one lagoon to the next, but that returns on the following half tidal cycle cannot be determined with the available data, and this remains an important follow- up to the work described here. The lower part of Figure 3 shows the cumulative net displacement of water past the study site calculated from u,,. Presented in this way, the net flow into Indian River lagoon stands out clearly. The resultant flow is largely due to a net southwest- ward flow during a two-week period in the middle of September, and a more persistent flow in the same direction during the last week of the study. The progressive vector diagram of hourly wind stress (Fig. 4) contains dots at the beginning of every week to indicate in a general way the timing of events of interest. For the first four weeks of the study, the net wind stress was in a southerly direction. Following three weeks of predominantly east-west wind stress, progressive vectors trace a zig-zag pattern toward the southwest and northwest. The spacing of the dots during the final week and a half of the study indicate that wind speeds increased to No. 4, 1993] SMITH—TIDAL AND WIND DRIVEN TRANSPORT 943 20 North-South Cumulative Windstress l 5 -100 -120 -100 -80 -60 -40 -20 0 20 East-West Cumulative Windstress Fic. 4. Progressive vector diagram of cumulative north-south and east-west windstress, August 17 through October 30, 1989. Winds were recorded hourly at the weather station shown in Figure 1. Dots have been entered every seven days along the plot. well over values recorded at the start of the study. Inspection of the hourly wind speeds reveals a mean value of 5.75 ms". Coherence spectra are not shown because along-channel flow was highly coherent with wind forcing at virtually every time scale and wind stress component considered. Of particular interest is the phase relationship between wind forcing and volume transport, because this will identify the direction of the inter-lagoon trans- port in response to specific wind conditions. Spectral analysis of these relatively short time series permits a few generalizations regarding cause-and-effect relationships. Comparison of along-channel flow and wind stress components (not shown) reveals an in-phase relationship for north-south and northeast-southwest wind stress com- ponents. Thus, for example, winds out of the northerly quadrant (northwest through northeast) will force water out of Mosquito Lagoon and into Indian River lagoon. Conversely, along-channel flow is generally out of phase with both east-west and southeast-northwest wind stress components. It follows that winds out of the southerly quadrant will force water out of Mosquito Lagoon and into Indian River 944 FLORIDA SCIENTIST [VOL 56 lagoon. Over this same range of time scales, the average magnitude of the transfer function is just below 100 cm s" per 1 dyne cm”. The average magnitude of the low- pass filtered wind stress used in the spectral analysis was 0.175 dyne cm”. Thus a representative wind-driven current through Haulover Canal would be about 17 cm s', in good agreement with recorded speeds (upper part Fig. 3). Volume transport—Using the parabolic distribution to describe the across- channel variation in along-channel flow, the -4.1 cm s mean flow into Indian River lagoon during the study is equivalent to -21.9 m® s’, and the cumulative volume transport is approximately 138 million m’. Tidal exchanges can be expressed similarly in terms of volume transport. Using (6), the 8.1 cm s M, amplitude translates into a maximum volume transport of 54.6 m® s', or about 777,000 m? per half M, tidal cycle. This is about 30% larger than the volume of the canal. Thus, on average, the canal is flushed completely by the M, tide. As noted above, under spring tide conditions, flushing increases by as much as 40% due to the added effects of the S, and N, constitutents. Results for all six tidal constituents investigated in this study are listed in Table 1. The long-term tide-induced residual volume transport is a small fraction of the total mean transport. Using (2) and (6), the average tide-induced transport is 0.06 m° s'. This is a transport from Mosquito Lagoon to Indian River lagoon, consistent with a progressive tidal wave moving in the same direction, but it is less than 0.3% of the total volume transport. DIscuss1ON—The most significant finding of the Haulover Canal study is the substantial, if irregular, transport of water from Mosquito Lagoon to the northern end of Indian River lagoon. Strictly speaking, results obtained from the current meter measurements relate only to the volume of water transported. The real significance, however, may lie in the associated transport of dissolved and suspended material, and even actively swimming animal species. Ehrhart (1983), for example, surveyed marine turtles in northern Indian River and southern Mosquito Lagoon. Results provide evidence of the physical linkage of these two bodies of water, and it is probable that currents of 30 cms" aid significantly the transport of turtles from one lagoon to the other. Similarly, Snelson (1983, 1993) studied the fishes of southern Mosquito Lagoon and northern Indian River and suggested that Haulover Canal may play an important role as a route of migration for fishes that spawn offshore and then move as juveniles into and through the Indian River system. The physical significance of the long-term net volume transport through Haulover Canal must be determined in terms of the volume of the receiving body of water. Smith (1993) has calculated a volume of about 92 million cubic meters for a segment of Indian River lagoon lying north of the northernmost causeway at 28°39'N. This is about 53% of the volume of water entering Indian River through the canal during the 73-day study. Little is known of the circulation of the northern end of Indian River lagoon, thus the flushing effect this might have cannot be estimated. Results obtained in this study, however, suggest that a substantial amount of water No. 4, 1993} SMITH—TIDAL AND WIND DRIVEN TRANSPORT 945 is at least passing through the northern end of the lagoon. Combining results of this study with climatological data from Cape Kennedy Air Force Station, it appears that under normal wind conditions, the wind-driven transport of volume will undergo a seasonal variation that includes transport from Mosquito Lagoon to Indian River lagoon from April through September, and transport from Indian River into Mosquito Lagoon from October through March. Incorporating the slightly stronger winds characteristic of fall and winter months, one gets the impression that the net flow might be toward the northeast, into Mosquito Lagoon. This tentative conclusion must be explored and verified with a longer data base, as well as a broader study including freshwater gains and losses. The inflow of water into the northern end of Indian River lagoon, and the general results of the present study, help explain a discrepancy that appeared in a modeling study of the flushing of the northern sub-basin of Indian River lagoon (Smith, 1990c). The model assumed a closed northern boundary and calculated the wind-driven set up and set down of water level in the northern sub-basin in response to local wind forcing. Model results tended to over-estimate slightly the observed slope in water level along the longitudinal axis of the lagoon. While the model has not been revised to include gains and losses through Haulover Canal, it appears that the volume transport quantified here is consistent with the errors that arose as a result of assuming that the northern end of the northern sub-basin of Indian River lagoon was closed. A more realistic model therefore has to take into account wind-driven exchanges with Mosquito Lagoon, and it may be necessary to model more of the Indian River lagoon system to reproduce observations to a sufficiently close approxi- mation. ACKNOWLEDGMENTS—Patrick Pitts and James Liu assisted in the installation and recovery of the current meter in Haulover Canal. Water level data (Fig. 2) were obtained from the National Ocean Service. Weather records for the Kennedy Space Center Shuttle Airport were obtained through the National Climatic Data Center. Support for this study was provided by the Florida Department of Natural Resources through Grant Agreement No. 6598. Harbor Branch Oceanographic Institution Contribution Number 997. LITERATURE CITED DENNIS, R.E. AND E.E. LONG. 1971. A user’s guide to a computer program for harmonic analysis of data at tidal frequencies. NOAA Tech. Rept. NOS 41, U.S. Dept. Comm., Rockville, MD, 31 pp EHRHART, L.M. 1983. Marine turtles of the Indian River lagoon system. Florida Scient. 46:337-346. GROVES, G. 1955. Numerical filters for discrimination against tidal periodicities. Trans. Amer. Geophys. Union 36:1073-1084. HAURWITZ, B. AND A.D. COWLEY. 1975. The barometric tides at Zurich and on the summit of Santis. Pure Appl. Geophysics 113:355-364. HEATHERSHAW, A.D. AND D.N. LANGHORNE. 1988. Observations of near-bed velocity profiles and seabed roughness in tidal currents flowing over sandy gravels. Estuar. Coastal Shelf Sci. 26:459- 482. KJERFVE, B.J. 1976. Circulation and salinity distribution in coastal Louisiana bayous. Contrib. Mar. Sci. 20:1-10. L.H. STEVENSON, J.A. PROEHL, T.H. CHRZANOWSKI, AND W.M. KITCHENS. 1981. Estimation of material fluxes in an estuarine cross section: a critical analysis of spatial measure- ment density and errors. Limnol. Oceanogr. 325-335. LITTLE, J. AND L. SCHURE. 1988. Signal processing toolbox user’s guide. The MathWorks, Inc., Nantic, Mass. 246 FLORIDA SCIENTIST [VOL 56 ROBINSON, I.S. 1983. Tidally induced residual flows. Pp. 321-356 In: JOHNS, B. (ed.) Physical Oceanogr. of Coastal and Shelf Seas, Elsevier Oceanogr. Series, vol. 35, (Johns, B. ed.), Elsevier, New York. SHENG, Y.P., S. PEENE, AND Y.M. Liu. 1990. Numerical modeling of tidal hydrodynamics and salinity transport in the Indian River lagoon. Florida Scient. 53:147-168. SMITH, N.P. 1986. The rise and fall of the estuarine intertidal zone. Estuaries 9:95-101. . 1987. An introduction to the tides of Florida’s Indian River lagoon. I, water levels. Florida Scient. 50:49-61. . 1990a. Computer simulation of tide-induced residual transport in a coastal lagoon. Journ. Geophys. Res. 95:18,205-18,211. . 1990b. Wind domination of residual tidal transport in a coastal lagoon. Pp. 123-133. In: (CHENG, R. T. ed.) Coastal and Estuarine Studies 38, Springer-Verlag, New York, NY. . 1990c. Longitudinal transport in a coastal lagoon. Estuar., Coastal Shelf Sci. 31:835-849. . 1990d. An introduction to the tides of Florida’s Indian River lagoon. II, currents. Florida Scient. 53:216-225. . 1993. Tidal and nontidal flushing of Florida’s Indian River lagoon. Estuaries 16 (in ress). eee F.F. 1983. Ichythyofauna of the northern part of the Indian River lagoon system, Florida Scient. 46:187-206. . 1993. Dept. of Biological Sciences, Univ. of Central Florida, Orlando, Florida, Pers. Commun. VAN DE KREEKE, J. 1976. Tide-induced mass transport: a flushing mechanism for shallow lagoons. Journ. of Hydraul. Research 14:61-67. WILLIAMS, J.L. 1985. Computer simulation of the hydrodynamics of the Indian River lagoon. M.S. thesis, Florida Institute of Technology, Melbourne, FL, 145 pp. WONG, K-C. AND J. DILORENZO. 1988. The response of Delaware’s inland bays to ocean forcing. Journ. Geophys. Res. 93:12,525-12,535. WU, J. 1980. Windstress coefficients over sea surface near neutral conditions—a revisit. J. Phys. Ocean. 10:727-740. YUSOF, M. R. 1987. Canaveral barge canal circulation study. M.S. thesis, Florida Inst. of Technology, Melbourne, FL. 142 pp. Florida Scient. 56(4):236-246.1993 Accepted: October 6, 1993. No. 4, 1993] STOJKOVSKI ET AL.—PARTIAL BIODEGRADATION 247 Environmental Chemistry PARTIAL BIODEGRADATION OF AROCLOR 1242 PCBs BY ALCALIGENES BACTERIA STOJAN STOJKOVSKI, BRUCE D. JAMES, AND RoBERT J. MAGEE Department of Chemistry, La Trobe University, Bundoora, Victoria 3083, Australia. ABSTRACT: The dechlorination potential of an aerobic bacterial species of Alcaligenes for Aroclor 1242 PCB has been investigated. GC-MS analysis of the dechlorinated products revealed that dechlo- rination was most effective for the tri- and tetrachloro derivatives and not significant for the penta- and hexachloro derivatives. MICROORGANISMS are known to play a role in the degradation of organochlo- rine compounds in the environment. The most extensive studies on the biodegrad- ability of chlorinated biphenyls involved the examination of aerobic degradation of 36 pure PCB isomers by Alcaligenes and Acinetobacter bacterial species (Furukawa et al., 1978 and 1979). Enzymatic activities were detected and a mechanism of degradation proposed, whereby chlorine substituents are replaced either by hy- droxyl groups or hydrogen atoms (Masse et al., 1984). More recently, the degrada- tion of a number of Aroclors by Alcaligenes eutrophus H850, also under aerobic experimental conditions has been reported (Bedard et al., 1987a,b). However, biological degradation of PCBs is also possible under anaerobic mediated dechlori- nation. This has been demonstrated with anaerobic sediments and slurries (Quensen et al., 1988; Brown et al., 1987; Sylvestre and Fauteux, 1982). We report here some preliminary studies on the degradative properties of two strains of Alcaligenes species towards PCBs under aerobic conditions in which DNA plasmids have been incorporated in the bacteria which are capable of degrading chlorinated compounds. MATERIALS AND METHODS—The microorganisms Alcaligenes eutrophus, strain JMP 134 p]P4 (2,4,D*, Hg") and Alcaligenes paradoxus, strain JMP 116 pJP1 (2,4,D*, Hg‘) were kindly donated by J.M. Pemberton, University of Queensland, St. Lucia, Queensland, Australia. These organisms contained a DNA plasmid capable of degrading chlorinated species. Strain JMP 116 is similar to JMP 134, except that itis mercury sensitive, whereas JMP 134 is mercury resistant. The bacteria were sub-cultured (3 days incubation) on medium agar plates, containing 1.5% (w/w) of Oxoid nutrient (consisting of “Lab-Lemco” powder 1, yeast extract 2, peptone 5, sodium chloride 5, and agar 15 g/L; pH 7.4) which was obtained from Difco Laboratories, Melbourne. Nutrient broths (1.5% Oxoid nutrient broth, with the constituents as before, but without the agar) were prepared for the degradative studies and the optimal temperatures for incubation of JMP 116 and JMP 134 were found to be 26°C and 35°C, respectively. A Hewlett-Packard Gas Chromatograph, model 5880, coupled to a Finnigan MAT-ion trap detector (ITD™) and a Hewlett-Packard 3880A integrator were used. The GC operating conditions were as follows: capillary column (SGE-BP1, 0.2mm x 25m) cold trapped at 65°C for 0.4 min; oven temperature ramped to 220°C at 20°C min’; injector port 250°C; helium gas carrier-linear velocity 30 cm’ min | The ITD scan mode conditions were: ionization 70 eV; transfer line at 250°C ; 45-500 a.m.u. scan mass /charge units and 1 scan s"; injection sample 1-5 uL, splitless. The dechlorination procedure was carried out as follows, using triplicate samples of Aroclor PCB 248 FLORIDA SCIENTIST [VOL 56 1242 (Foxboro Analytical, North Haven, CT 06473). 600 mg/L (or 600 ppm) was dissolved in chromatography grade n-hexane (Mallinckrodt Australia Pty. Ltd.) and placed in a 1-L erlenmeyer flask containing glass beads (500 x 5 mm diameter). The n-hexane solvent was evaporated under a stream of nitrogen without loss of Aroclor 1242 PCBs. Distilled water was then added to each flask with the appropriate nutrient broth (i-e., 1.5% nutrient broth) and all the flasks autoclaved at 110°C. The bacteria (in 1.5% nutrient broth) were then grown at the required temperature. Controls (without the bacteria) were also treated under identical conditions. The cultures with and without Aroclor 1242 were mechanically shaken and incubated for 2 weeks under their optimum temperatures. When incubation was completed, the degraded Aroclor 1242 PCBs were extracted with n-hexane for GC-MS analysis. RESULTS AND DISCUSSION—Alcaligenes eutrophus and Alcaligenes paradoxus were chosen because their incorporated DNA plasmids have the capability of degrading polychlorinated biphenyls. The investigations were carried out in vitro with the pure bacteria, under a wide range of aerobic conditions; for example, humidity, temperature and pH. Table 1 indicates the percentage degradation of homolog congeners of the Aroclor 1242 (which consists of predominantly tri- and tetrachlorobiphenyls) before and after aerobic degradation by Alcaligenes eutrophus (JMP 134) and Alcaligenes paradoxus (JMP 116) species. Although the analyses were performed by GC-MS, not all quantitations were established. Nevertheless, it was evident that peak profiles were altered owing to the activity of the bacteria. It was found that the microbial action was effective for dechlorination of the mono-, di-, tri- and tetra-chlorobiphenyls, but it was not obvious with higher chlorine-substituted compounds, such as penta- or hexachlorobiphenyls. From Table 1 it can be seen that there are lowered levels of trichlorobipheny] and tetrachlorobipheny] and increased levels of monochlorobiphenyl, dichlorobipheny! and biphenyl. This trend has been reported in the literature by others (Sylvestre and Fauteux, 1982; Sylvestre et al., 1985; Furukawa et al., 1983 and 1979). It should be noted that the percentage degradation is based on chromatographic peak areas of undegraded PCBs (ie., control) versus aerobically degraded PCBs. It would appear that degradation under aerobic conditions is generally limited to congeners with four or fewer chlorine atoms on the PCBs (Bedard et al., 1987a, b). It is now well established that aerobic degradation of PCBs is catalyzed by a dioxygenase; more specifically 2,3-dioxygenase with two adjacent chlorine atoms at ortho- and meta- positions or by 3,4-dioxygenase when two ortho- chlorines are on the same ring. In fact, chlorines are removed more efficiently with 2,3-dioxygenase than 3,4-dioxygenase (Bedard et al., 1987a, b; Quensen et al., 1988; Brown et al., 1987). The replacement of mono- and dichloro- substituents by hydroxy groups to form hydroxybiphenyls is well understood, but we were only able confirm this qualitatively and not consistently. Furthermore, the presence of biphenyl was clearly evident after the degradation procedure was carried out. It appears that the two bacterial species JMP 134 and JMP 116 could have some potential for the biodegradation of chlorinated compounds and thus may be useful in the area of waste treatment, except that JMP 116 could not be used when mercury is also present. The variations in the dechlorination values for the two strains are possibly due to the difference in their logarithmic growth rates. At their optimum growth temperatures, JMP 134 grows approximately three times faster than JMP 116. No. 4, 1993] STOJKOVSKI ET AL.—PARTIAL BIODEGRADATION 249 TABLE 1. Percentage distribution of PCB congeners of an Aroclor 1242 (at 600 ppm) before and after 2 weeks of dechlorination by aerobic Alcaligenes sp. Congeners Retention time % Distribution of Congeners (min) (in mole % of PCB recovered) Dechlorination before after 2 weeks JMP 134 ==JMP 116 (at 35°C) — (at 26°C) Biphenyl 2.3 0 95 18 Monochlorobiphenyl 3.2 0 20 15 Dichlorobiphenyl 4.0 9 18 15 Trichlorobiphenyl 4.4-5.2 49 40 45 Tetrachlorobipheny] 6.4 36 28 32 Pentachlorobiphenyl 7.0-8.1 5 5 5 Hexachlorobipheny! 9.0-10.8 1 1 1 ACKNOWLEDGMENTS—We wish to thank Prof. J.S. Waid, Department of Microbiology, La Trobe University for the use of their GC/MS system. One of us (S.S.) acknowledges receipt of aCommonwealth of Australia Postgraduate Research Scholarship. LITERATURE CITED BEDARD, D.L., R.E. WAGNER, M.J. BRENNAN M.L. HABERL, AND J.F. BROWN, Jr. 1987a. Extensive degradation of Aroclors and environmentally transformed polychlorinated biphenyls by Alcaligenes eutrophus, H850. Appl. Environ. Microbiol. 53: 1094-1102. , M.L. HABERL, R.J. MAY, AND M.J. BRENNAN. 1987b. Evidence for novel mechanisms of polychlorinated biphenyl] metabolism in Alcaligenes eutrophus, H850. Appl. Environ. Microbiol. B3-1103-1112: BROWN, J.F., JR., D.L. BEDARD, M.J. BRENNAN, J.C. CARNAHAN, H. FENG, AND R.E. WAGNER. 1987. Polychlorinated biphenyl dechlorination in aquatic sediments. Science 236:709-712. FURUKAWA K., N.TOMIZUKA, AND A. KAMIBAYASHI. 1979. Effects of chlorine substitution on the bacterial metabolism of various polychlorinated biphenyls. Appl. Environ. Microbiol. 38:301- 310. , K. TONOMURA, AND A. KAMIBAYASHI. 1978. Effects of chlorine substitution on the biodegradability of polychlorinated biphenyls. Appl. Environ. Microbiol. 35:223-227, , N. TOMIZUKA, AND A. KAMIBAYASHI. 1983. Metabolic breakdown of Kaneclors (PCBs) and their products by Acinetobacter sp. Appl. Environ. Microbiol. 46:140-145. , K. TONOMURA, AND A. KAMIBAYASHI. 1979. Metabolism of 2,4,4'-trichlorobiphenyl by Acinetobacter sp.P6 Agric. Biol. Chem. 43:1577-1583. MASSE, R., F. MESSIER, L. PELOQUIN, C. AYOTTE, AND M. SYLVESTRE. 1984. Microbial biodegradation of 4-chlorobiphenyl, a model compound of chlorinated biphenyls. Appl. Environ. Microbiol. 47:947-951. QUENSEN, J.F.,III, J.M. TIEDJE, AND S.A. BoyD. 1988. Reductive dechlorination of polychlorinated biphenyls by anaerobic microorganisms from sediments. Science 242:752-754. 250 FLORIDA SCIENTIST [VOL 56 SYLVESTRE, M., AND J. FAUTEUX. 1982. A new facultative anaerobic capable of growth on 4- chlorobiphenyl. J.Gen. Appl. Microbiol. 28:61-72. , R. MASSE, C. AYOTTE, F. MESSIER, and J. FAUTEUX. 1985. Total biodegradation of 4- chlorobiphenyl (4-CB) by a two-membered bacterial culture. J. Appl. Microbiol. Biotechnol. 21:192-195. Florida Scient. 56(4):247-250. Accepted: October 13, 1993. FLORIDA ACADEMY OF SCIENCES 1993 ANNUAL MEETING OUTSTANDING STUDENT PAPER AWARDS (con’t) BIOLOGICAL SciENcEs Kym Rouse Demora, Department of Biology, University of Central Florida, “Bioaccumulation of Heavy Metals in Fish Living in Stormwater Treatment Ponds.” Graduate Award. Penny Cople, Department of Biology, Stetson University, “The Effects of Satellite Nests on Predation in Artificial Nests of Pseudemys floridana peninsularis.” Undergraduate Award. REVIEW Salvadore, Amos (ed.), The Gulf of Mexico Basin, The Geological Society of America, P.O. Box 9140, Boulder, Colorado, 1992, pp. 568. Price $77.50 postpaid. Tuis book and accompanying slip case of six plates is truly a remarkable collection of papers that will stand for some time to come as the definitive reference material for anyone interested in the Gulf of Mexico and surrounding environs. The book is an assemblage of eighteen comprehensive chapters that are testimonials to the many authors who provided original information permitting the chapter's authors to synthesize and summarize what is known. The book begins with historical observations and speculations of the Gulf’s origin. The next five chapters provide physiographic and bathmetric divisions, structural framework, crust underlying the Basin, location and importance of salt tectonics and evidence of igneous activity beginning in Late Triassic and continuing until Mid-Tertiary. The next six chapters deal with sedimentation from Pre-Triassic to Quaternary. The remaining chapters describe the Basin’s origin and formation as well as oil and gas reserves, mineral resources and geo-pressured-geo-thermal energy. Clearly, our understanding of the Gulf Basin has come a long way, but as pointed out in Chapter eighteen, there remain “many questions to be answered and fundamental problems are yet to be solved.” : —William H. Taft, Boston University, Boston, MA No. 4, 1993] 251 OBITUARY JAMES G. POTTER, 1907-1993 Jim Potter, physicist, Professor Emeritus at Florida Institute of Technology, former president of the Florida Academy of Sciences (1971-72) died October 22 at the West Melbourne Health Care Center. He was a dedicated teacher, having covered the aspects of physics to a remarkable number of students during a career that spanned six decades. He received a bachelor’s from Princeton (1928), a masters from New York University (1931) and a doctorate from Yale (1939). He was a faculty member at Illinois Institute of Technology, head of physics at South Dakota School of Mines (1940-44), Texas Tech (Professor and head of physics, 1945-67). He became head of physics at Florida Tech in 1967 and formally retired after 20 years there. In addition he had worked for Bell Telephone Research Laboratories and Naval Ordinance Test Station at China Lake, California. All of us will remember his dedication, his high standards, his ability to motivate, but also his kindness. He is survived by Mrs. Dorothy Potter, his wife of 52 years, two sons, three grandchildren, and innumerable friends. DFM, BBM ACKNOWLEDGMENT OF REVIEWERS It is a pleasure to acknowledge the service, dedication and cooperation of the following persons, who generously contributed their time and expertise in reviewing manuscripts for Volume 56. Some kindly reviewed more than one manuscript. Daniel F. Austin David Burr Wiley Kitchens Jacobus van de Kreeke Clarence Collison John M. Lawrence Bruce C. Cowell James N. Layne Jim Cox G. Michael Lloyd, Jr. Franklin F. Snelson, Jr. William M Spikowski Kerry Steward I. Jack Stout William H. Taft Clinton J. Dawes Bijan Dehgan Steve Dicks C. Kenneth Dodd Ernest Esteves Richard Franz Allen Gettman Martin K. Huehner Steve Humphrey Dale Jackson Leon Mandell Dean F. Martin Frank Mazzotti Paul Moler Ralph E. Moon Henry Mushinsky Ron Myers John Osborne H. Shimawauki Donald A. Shroyer Walter K. Taylor Fred G. Thomson Walter Thomson Daniel Ward Jeftrey T. Williamson Larry David Wilson Charles Woods Richard P. Wunderlin Florida Scientist QUARTERLY JOURNAL of the FLORIDA ACADEMY OF SCIENCES VOLUME 56 DEAN F. MARTIN Editor BARBARA B. MARTIN Co-Editor Published by the FLORIDA ACADEMY OF SCIENCES, INC. Indialantic, Florida 1993 The Florida Scientist continues the series formerly issued as the Quarterly Journal of the Florida Academy of Sciences. The Annual Program Issue is published independently of the journal and is issued as a separately paged Supplement. Copyright © by the FLoripa ACADEMY OF SCIENCES, INc. 1993 954 FLORIDA SCIENTIST [VOL 56 CONTENTS OF FLORIDA SCIENTIST VOLUME 56 NUMBER ONE Movement Patterns of Translocated Big Cypress Fox Squirrels (Sciurus niger vicennid) ........csscssssnoersnnoonsrnsesccosetieysss eee Randy S. Kautz The Ecological Basis of the Kissimmee River Restoration Plan .................0. Louis A. Toth Low Clutch Viability of American Alligators on Lake Apopka ..............00 Alan R. Woodward, H. Franklin Percival, Michael L. Jennings, and Clinton T. Moore First Record of the Eastern Big-eared Bat (Plecotus rafinesquii) in Sovthernm Florida ..cc.iscccsssccsesveccsseeeesstveunscobacas soetosec te ean eee Larry N. Brown and Curtis K. Brown NUMBER Two Cattle Fatalities From Prolonged Exposure to Aedes Taeniorhynchus in Southwest Florida wicca. coe i eee David S. Addison and Scott A. Ritchie Habitat Structure and the Dispersion of Gopher Tortoises on a Nature BY CSETVE iiss age s'onidus tasadennnaddedeoondamamsctooete Om aie Naesese anise eee M.C. Stewart, D. F. Austin and G.R. Bourne Natural History of a Small Population of Leiocephalus Schreibersii (Sauria: Tropiduridae) from Altered Habitat in the Dominican Republic ........... Michael C. Schreiber, Robert Powell, John S. Parmerlee Jr., Amy Lathrop and Donald D. Smith BOOKIREVIEW.ccstosiahieeso loa. ee ee ee James N. Layne Trends in Numbers of Loggerhead Shrikes on Roadside Censuses in Peninsular Florida, 1974-1992 «..........0:::.0:e:ecroseeane Dene ats seedeecese eee Reuven Yosef, James N. Layne and Fred E. Lohrer Adsorption of Several Atmospheric Polluting Gases Upon Dehydrated GYPSUM 0 soeoieesiececeeseenesnde soeeaeZcsendoeteins secede basins don Spncenetan eee eee Robert F. Benson and George D. Blyholder Scaled Chrysophyceae and Synurophyceae From Florida: IV. The Flora of Lower Lake Myakka and Lake Tarpow...............0..00000:00000005: eo Peter A. Siver and Daniel E. Wujek Predation on Artificial Ground Nests in Southwest Florida ................00000eeeee Kimberly J. Babbitt and Jeffrey L. Lincer Book Review .......:s..cssensssesueonsosesoseestetseesnotsoeee. Wettceee seer 25 52 63 65 70 82 91 92 98 No. 4, 1993] Mee wameetia GH let dien)X|Y Complexes \.........5500rcncaresereresneronencnonsecsnsitosesores Jay W. Palmer, William E. Swartz Jr., David King and Joseph A. Stanko Iii edt cad Recs ve er er ar ete EN ue vlesticl snsinvanoulive hvmiiessea DAREN Coes NUMBER THREE Notes on the Impact of the December 1989 Freeze on Local Populations of Rivulus marmoratus in Florida, with Additional Distribution PyS S07 GUUS See A cach see a A D. Scott Taylor eee Ae hae I ec caiciin hook buch bad bomisseibibanventenoveeds Carl A. Luer Vegetative Cover in Florida Based on 1985-1989 Landsat Thematic AHO Tal IMM ACE ayaa sensed. Sohn ndsscsbed acd tesece Sent ocbhdedes Sadach sn cddl Nedaduedne eens web Randy S. Kautz, D. Terry Gilbert, and Gregory M. Maulden Tree Planting and Preservation Practices at Single-Family Residences: >t Heyy COONS CIE TOTON Sire eee cee ener ee PETE Lisa B. Beever, Tim Eckert, and Jeffrey S. Magnun Selection of Nest Cavity Volume and Entrance Size by Honey Bees in FF LOTTE 2) consoeedegn coho sae Rene Seno TAC Gee MERE ToT eT On a mn a a Roger A. Morse, James N. Layne, P. Kirk Visscher, and Francis Ratnieks Movement of Fluridone in the Upper St. Johns River, Florida ..............0..... Andrew J. Leslie, Don C. Schmitz, R. L. Kipker, and D. L. Giradin Chemical Differences Between Stressed and Unstressed Individuals of bealal Gress, laxoditsm Gistich Un wesssieres unescctosedeo. ween ctsadewewseneedbaeds s0e: Donna Hall Miller, Sydney T. Bacchus, and Harvey A. Miller Dolomite Extraction from Phosphate Pebble by Aqueous Carbonic Acid- Pomona Subbate: BULOr sc... sekeags desl ee hance shes eveedethcadevbstusseea desaadoces Svante. Yixue Pan, Robert F. Benson, and Dean F. Martin NUMBER FOouR Hurricane Andrew and the Colonization of Five Invading Species in South [21 CISC cccaboccoctosnen Rabbaneen: sckscasnscs pate iee coecen en che he ee Walter E. Meshaka, Jr. Synthesis of a Sterically Challenging Bis-(B - diketone) ..0....0..c cess Venkatraj V. Narayanan Geographic Distribution of the Striped Mullet (Mugil cephalus Linnaeus) me tie Atlantic and Eastern.Pacific OCEANS ¢.cc........3-.cceseccceesecceeeeeceseesceees Carter R. Gilbert Field and Laboratory Experiments on the Consumption of Mangrove Leaf Litter by the Macrodetritivore Melampus coffeus L. (SAS TG OO AOC TOO RIE pee anor accce es Sec ece EER eee ee C. Edward Proffitt, Kevin M.Johns, C. Bruce Cochrane, Donna J.Devlin, Theresa A. Reynolds, Deborah L. Payne, Sean Jeppesen, David W. Peel, and Dwane D. Linden 123 128 129 134 135 155 163 168 178 185 204 211 256 FLORIDA SCIENTIST [VOL 56 BOOK REVICW Links t.cccibodiccecobecocsecesmas seen. c te ee Charles D. Norris 222 Fruit Removal and Interplant Distance in the Persimmon: Diospyros virginiana saiudunbea gh’ dedewnbe obloduabewa bonadd obucdectad OASern eee oe eee eae ee Cris Cristoffer 223 Dimethy] Sulfoxide Catalyzed Racemization of Aspartic Acid .............e Gregory P. Cusano, Mirtha Chavez, Diego Torres, and George H. Fisher 226 Additions to the Herpetofauna of Egmont Key, Hillsborough County, Florida ....:s.ccccescsosceticdsstesibastdeiedhssnedisbesvaswnt eet Lora L. Smith, Richard Franz, and C. Kenneth Dodd, Jr. 231 Tidal and Wind-driven Transport Between Indian River and Mosquito Lagoon, Florida. ..csccc.cccccsnnsedbeecsd toads sae scens nando l Ned P. Smith 235 Partial Degradation of Arochlor 1242 PCBs by Alcaligenes Bacteria............ Stojan Stojkovski, Bruce D. James, and Robert J.Magee 247 BOO REVICW: Sh.0ceL eid eee Oe ee ee William H. Taft 250 ODI ALY: se.05sdovedncsnsntsscuinosnonosinedeonb’ hon esueotiovoulestl Selene ole (ease ge eee tetera het rn 251 Acknowledgment of Reviewers ::.............c5.000+sscusewevtsnoedeewe ong oot 951 Index, Vol 56 ccosdccsceseececcescteechcaen. aoe a I ee 953 INSTRUCTIONS TO AUTHORS Individuals who publish in the Florida Scientist must be active members in the Florida Academy of Sciences. Submit a typewritten original and two copies of the text, illustrations, and tables. All typewritten material—including the abstract, literature citations, footnotes, tables, and figure legends—shall be double-spaced. Use one side of 8'/2 x 11 inch (21'/2 cm x 28 cm) good quality bond paper for the original; the copy may be xeroxed. Margins should be at least 3 cm all around. Number the pages through the Literature Cited section. Avoid footnotes and do not use mimeo, slick, erasable, or ruled paper. Use metric units for all measurements. Assistance with production costs will be negotiated directly with authors of pages which exceed 10 printed pages. Excess over 10 pages by an author in a single volume will result in charges for additional pages. AppreEss follows the author’s address. ABsTRACT—AIl manuscripts shall have a short, concise, single-paragraphed abstract. The abstract follows immediately the author’s address. ACKNOWLEDGMENTS are given in the body of the text preceding immediately the Literature Cited section. LITERATURE CITED section follows the text. Double-space every line and follow the format in the current issue. Manuscrirts of 5 or less, double-spaced typewritten pages should conform to the short article protocol. See recent issue for proper format of both long and short articles. TaBLEs shall be typed on separate sheets of paper, double-spaced throughout. Each table shall contain a short heading. Do not use vertical rulings. Maximum character width, including spaces, of tables is 98.0. This includes a minimum space of 5 characters between columns. Tables are charged to authors at $13 per quarter page, $29 per half page, and $39 per full page. ILLusTRATIONS—All drawings shall be done in good quality India ink, on good board or drafting paper. Letter by using a lettering guide or equivalent. Typewritten letters on illustrations are unaccept- able. Drawings and photographs should be large enough to allow 1/3 to 1/2 reduction in size. Photographs shall be glossy prints of good contrast. Whenever possible, mount photographs i in lots size. The author’s name and figure number should be penciled lightly on the back of each figure. Figure legends must be listed on a separate page and not on the drawing or photograph. The legend must be double spaced. Illustrations are charged to authors at $6.00 per page or fraction. We ask authors to prepare accepted manuscripts in personal computer formats. Please do not submit a manuscript on disk. Once reviewers have recommended a manuscript be accepted, you will be given directions for submitting the final manuscript on disk. Proor must be returned promptly. Notification of address change and proofreading are the author’s responsibility. Alterations after the type has been set will be charged to the author. REPRINTS may be ordered through the FAS office from the printer on forms provided at the time that proofs are sent to authors for checking. FLORIDA ACADEMY OF SCIENCES INSTITUTIONAL MEMBERS FOR 1993 Archbold Biological Station St. John’s River Community College, Palatka Brevard Community College, Florida Advanced __ St. Petersburg (Florida) Public Library Technology Center Library, Palm Bay St. Thomas University, Miami, FL The British Library, Boston Spa, Weatherby, U.S. Department of Commerce, National UK Marine Fisheries (Panama City) Charles Courtney / Hillsborough County EPC U.S. Environmental Protection Agency Library, Florida Institute of Technology Gulf Breeze, FL Harbor Branch Oceanographic Institution, Inc. University of Central Florida, Division of Lake-Sumter Community College Sponsored Research Newfound Harbor Marine Institute University of South Florida CORPORATE MEMBERS FOR 1993 Mote Marine Laboratory, Sarasota Membership applications, subscriptions, renewals, changes of address, and orders for back numbers should be addressed to the Executive Secretary, Florida Academy of Sciences, P.O. Box 033012, Indialantic, FL 32903-0012 PUBLICATIONS FOR SALE by the Florida Academy of Sciences Complete sets. Broken sets. Individual numbers. Immediate delivery. A few numbers reprinted by photo-offset. All prices strictly net. Special bargain: $500 for nearly complete sets. Prices quoted include domestic postage. Some issues may not be available. All are $20 per volume or $5 per issue, except for symposium issues. PROCEEDINGS OF THE FLORIDA ACADEMY OF SCIENCES (1936-1944) Volumes 1-7 QUARTERLY JOURNAL OF THE FLORIDA ACADEMY OF SCIENCE (1945-1972) Volume 8-35 FLORIDA SCIENTIST (1973) Florida’s Estuaries—Management or Mismanagement?—Academy Symposium FLORIDA SCIENTIST 37(4)—$5.00 Land Spreading of Secondary Effluent—Academy Symposium FLORIDA SCIENTIST 38(4)—$5.00 Solar Energy—Academy Symposium FLORIDA SCIENTIST 39(3)—$5.00 (includes do-it-yourself instructions) Anthropology—Academy Symposium FLORIDA SCIENTIST 43(3)—$7.50 Shark Biology—Academy Symposium FLORIDA SCIENTIST 45(1)—$8.00 Future of the Indian River System—Academy Symposium FLORIDA SCIENTIST 46(3/4)—$15.00 Second Indian River Research Symposium—Academy Symposium FLORIDA SCIENTIST 53(3)—$15.00 Please send payment with order. If required, an invoice will be sent on purchases over $20 from a recognized institution. Please include Florida sales tax if shipped to a Florida address, unless a copy of a valid Florida Consumer’s Sales Tax Certificate is enclosed. The Florida Academy of Sciences, Inc. P.O. Box 033012 Indialantic, Florida 32903-0012 Telephone: (407) 723-6835 f . “ Ae - a ' a aa a * CORED t/a OS ee Furey gag i ge : mex mene = ISSN: 0098-4590 ‘Florida Scientisi Volume 57 Winter/Spring, 1994 Number 1, 2 CONTENTS Predation of Hatchling Diamondback Terrapin, Malaclemys terrapin (Schoepff), by the Ghost Crab, Ocypode quadrata (Fabricius). IL.......... Rudolf G. Arndt 1 Reproductive Cycle of the Indo-Pacific Gecko, Hemidactylus garnotii, in Sem oti EsR| ay tal Ol MN eR cco eh eh ce Pes ne clieed aesuwiets sanasouvledeaideadeustansoiierior’ Walter E. Meshaka, Jr. 6 Effect of Treated Kraft Black Liquor on Hydrilla verticillata (Royle)........... Andrew L. Hassell, Barbara B. Martin, Dean F. Martin, and Jess M. Van Dyke 10 Macroinvertebrates of the Northern Everglades: Species Composition and CL OUPDIPE® SRW Sao ee cee ee ee el Russell B. Rader 22 Silver Accumulation in Three Species of Fish (Family: Centrarchidae) in Seemminr ee Mreatiment, POWGS .2....s:.c2cs-1c-r0ssocencsnedecvasave sanecbeesdecseasevvevwesesos Kym Rouse Campbell 34 an ots ioe 0 loess bas bed vsnievgundshlwaebeleecne nts Joseph A. Stanko 42 The Use of Computer-Assisted Molecular Modeling in College General CL ELITIS TE 7 cccocbosnegoee BleetOn ACEO Cnet James R. Yount 43 A Mass Stranding of Leach’s Storm-Petrel in Georgia and Florida ............... Carol A. Ruckdeschel, C. Robert Shoop, and George W. Sciple 48 Yeast Interactions Inferred From Natural Distribution Patterns .................. Philip F. Ganter, Eduardo Bustillo, and Jennifer Pendola 50 MN ini ee esse Aa cine vs Pods Suags coi vaveeudendosiessdeateviean 00s 62 DOOR CVICW — ooo. coco oeccsaceendc coe celeceN rs: Se ae Robert F. Benson 63 LAs or Ty ETS TEMG ee ee een or 64 FLORIDA SCIENTIST QUARTERLY JOURNAL OF THE FLORIDA ACADEMY OF SCIENCES Copyright© by the Florida Academy of Sciences, Inc. 1994 Editor: Dr. DEAN MartTIN Co-Editor: Mrs. BARBARA B MartTIN Institute for Environmental Studies Department of Chemistry University of South Florida 4202 East Fowler Avenue Tampa, Florida 33620-5250 Phone: (813) 974-2374 E-Mail: Martin@chuma.cas.usf.edu THE FLORIDA SCIENTIST is published quarterly by the Florida Academy of Sciences, Inc., a non-profit scientific and educational association. Membership is open to indi- viduals or institutions interested in supporting science in its broadest sense. Applica- tions may be obtained from the Executive Secretary. Both individual and institutional members receive a subscription to the FLORIDA SCIENTIST. Direct subscription is avail- able at $40.00 per calendar year. Original articles containing new knowledge, or new interpretation of knowledge, are welcomed in any field of Science as represented by the sections of the Academy, viz., Biological Sciences, Conservation, Earth and Planetary Sciences, Medical Sciences, Physical Sciences, Science Teaching, and Social Sciences. Also, contributions will be considered which present new applications of scientific knowledge to practical problems within fields of interest to the Academy. Articles must not duplicate in any substantial way material that is published elsewhere. Contributions are accepted only from members of the Academy and so papers submitted by non-members will be accepted only after the authors join the Academy. Instructions for preparation of manuscripts are inside the back cover. Officers for 1993-1994 FLORIDA ACADEMY OF SCIENCES Founded 1936 President: Dr. Patrick J. GLEASON Treasurer: Dr. Marvin Ivey 1131 North Parkway 14452 Hillview Drive Lake Worth, Florida 33460 Largo, Florida 34644 President-Elect: Dr. Patricia M. Dooris Executive Secretary: Mrs. BETTY PREECE Institute for Environmental Studies P.O. Box 033012 Department of Chemistry Indialantic, Florida 32903-0012 University of South Florida Tel: (407)723-6835 Tampa, FL 33620-5250 Program Chair: Dr. DEL DELUMYEA Secretary: Ms. MARCELA GUITERREZ-MAYKA Millar Wilson Laboratory for Chemical Research 701 E. River Dr. Jacksonville University Temple Terrace, FL 33617 Jacksonville, Florida 32211 Published by the Florida Academy of Sciences, Inc. P.O. Box 033012 Indialantic, Florida 32903-0012 Printing by C & D Printing Company, St. Petersburg, FL Florida Scientist QUARTERLY JOURNAL OF THE FLORIDA ACADEMY OF SCIENCES DEAN F. Martin, Editor BARBARA B. Martin, Co-Editor Volume 57 Spring, 1994 Number | Biological Sciences PREDATION ON HATCHLING DIAMONDBACK TERRAPIN, MALACLEMYS TERRAPIN (SCHOEPFF), BY THE GHOST CRAB, OCYPODE QUADRATA (FABRICIUS): II. RUDOLF G. ARNDT Faculty of Natural Sciences, Stockton State College, Pomona, NJ 08240, U.S.A. Asstract: Actual and presumed predation by the ghost crab on a total of 16 hatchling diamondback terrapin was observed on a barrier island in southern New Jersey in October 1990, 1991 and 1992. Hatchlings are attacked and killed by bites to the head. Feeding starts on the head and typically continues posteriorad until all soft parts, as well as the anterior portions of the shell, are consumed. Predation is probably opportunistic. Hatchlings probably represent only a minor food of the abundant and highly- active crab, but the crab could be a significant source of mortality on the much less abundant hatchlings. OcyYPODE QuapDRATA and some congeners (O.ceratophthalmus, O. cordimanus, O. kuhlii) are rapacious predators on the eggs and hatchlings of almost all species of sea turtles as well as of several species of birds (least tern, sooty tern, white-tailed tropicbird, Cape rock thrush, black skimmer)in the southeastern United States, Australia and Africa (Beebe, 1904; Bustard, 1966, 1979: Carr, 1967: Comins, 1961: Dodd, 1988; Ernst and Barbour, 1972; Hendrickson, 1958; Phillips, 1987; Sprunt, 1948). Other reports of unspecified crab predators on turtles are probably also referable to O. quadrata, as well as to other species of Ocypode (see reviews in Dodd, 1988; and Ernst and Barbour, 1972). Arndt (1991) reported predation on hatchlings of Malaclemys terrapin by O. quadrata, the only species of Ocypode found in the United States, and this paper extends those observations. METHODS AND MATERIALS—Survey area—Field work was done on Little Beach Island (LBI), Edwin B. Forsythe National Wildlife Refuge, Atlantic County, New Jersey. LBI is an uninhabited y- shaped barrier island ca. 5.5 km long and a maximum of ca. 2 km wide. High to low dunes in all stages of succession occur along all of the eastern (ocean-facing) side of the island, on much of the western (saltmarsh-facing) side, as well as at the inlet-bordered southern and northem ends; much of the interior is low, and contains 2 large bays and tidal creeks and salt marsh pools. Vegetation in the survey area ranged from absent to heavily vegetated by 1 or more species of several grasses, forbs, and shrubs. 9 FLORIDA SCIENTIST [VOL 57 Washed-up debris was locally abundant. Additional description and a figure of the island are in Burger and Montevecchi (1975). Survey dates and hours—The survey area was visited on 31 October, 2, 4 and 7 November 1990; 18 and 23 October, 6 November 1991; and 10, 15 and 23 October 1992. All visits were between 1200 and 1700 hrs. Survey methodology—Areas were searched in 1990 for greatest sign of crab (individual alive or dead specimens, burrows, tracks) and terrapin (individual alive or dead specimens, carapaces, tracks, and dug-up nests), other terrapin predators, and of predator-terrapin interaction. The areas of highest crab and terrapin density, namely the southern tip of the island and the southern half of the eastern shore, were systematically visually surveyed in 1991 and 1992. I started from the SW edge of the island, then walked E for ca. 0.5 km, and then continued N on routes of an additional straight-line distance, depending on date, of about 2.2, 2.5, or 4.3 km; actual distances walked were about 10% longer. Each survey was conducted only while walking away from the SW portion of the island, and usually required 3 to 4 hours. On each visit I surveyed a swath of ca. 5-8 m, its width determined by the immediate topography; type and abundance of vegetation; frequency of sign of crab, terrapin and other possible turtle predators; and abundance of debris. Swath location varied slightly on each visit, but was always located between the upper beach and about 50 m inland. Field work in 1990 was perfunctory and preliminary. Al! crab and terrapin sign are distinctive and cannot be confused with those of other species. The tracks of Ocypode and of hatchlings are quickly distorted and then eliminated by wind and rain, and thus their relative ages can be determined. Direction of movement can also be determined from tracks. Likewise, an active crab burrow opening, be it open or closed, can be differentiated from an inactive opening (pers. obs., and see Cowles, 1908; Milne and Milne, 1946; and Frey and Mayou, 1971). Burrow excavation—Seven burrows of Ocypode found closest to attacked and to freshly killed hatchlings were excavated within 2 to 5 days after the actual and the presumed attacks to search for prey remains and crab feces. Temperature recording—Air temperature (in shade, at ground level, and out of the wind) and soil temperature (at 1-1.5 cm depth and in the sun) were usually recorded with a field thermometer where hatchling and crab activity were pronounced. ResuLts—Air temperature during the surveys ranged from 8.3-25.0 C and soil temperature from 21.0-37.6 C. Crabs and hatchlings were active on most survey dates, and crabs were abundant in those areas where terrapin sign was frequent. I found a maximum of ca. 23 ghost crab at their burrow entrances and/or prowling on the sand on each date as late as 4 November. On 15 October, on a survey route of about 3.3 km, I counted 951 large open, and 430 large plugged closed and presumed active, burrows, all with a minimum burrow opening dimension of about 2 cm, i.e. sufficient to accomodate an O. quadrata large enough to be a potential predator on hatchling terrapin. Fresh tracks of up to ca. 10 hatchling terrapin and generally heading away from the ocean were found on each date as late as 31 October. Three terrapin nests with still-buried hatchlings and/or fresh egg shells were found by following hatchling tracks backward. The first hatchling of an additional nest was observed to break through the sand above its nest on 23 October. Several nests freshly dug up by unknown predators were found on each of several dates. A total of 16 dead or wounded hatchling terrapin was found. One hatchling had apparently just been attacked and released by a crab as the latter seized another hatchling (Arndt 1991). Another 7 of the 16 were fresh carcasses, in some cases not yet dead, and with presumed bite marks on the head. Such marks were as those noted on the | hatchling above that had just been released. Five of the 9 hatchlings were found 3.5 cm to 6.4 cm from a crab burrow (Fig. 1), and 4 at 23 cm to 150 cm from a burrow. I observed a crab to pull 1 of these fresh carcasses into its burrow, and another crab to appear inside its burrow entrance where there was a wounded hatchling lying just outside. These 9 hatchlings were found on 10, 18 and 31 October. Air temperatures then ranged from 20.5 C to 25 C. No. 1, 2 1994] ARNDT—PREDATION ON DIAMONDBACK a Y/ Y Y 4, $5 Fic. 1. Attacked terrapin hatchling as found outside a ghost crab burrow opening. Of the 7 other terrapins, 6 lacked all, or almost all, soft parts, and 1 had considerable remains of soft parts. This last hatchling was strongly misshapen, but I still count it as prey of Ocypode. One of these hatchlings was 24 cm, and another was 91 cm, from an active burrow opening; the remainder were several meters from a burrow. These hatchlings were found on 10, 15 and 23 October. The condition of the 16 hatchlings when found was: 7 with sign of feeding only on the head, which in 1 included removal of the guts; 3 with sign of feeding on the head and with 1 or both front legs removed, which in 1 included removal of the guts; 1 cleaned of all flesh and with only the tail remaining; and 5 cleaned of all flesh (Fig. 2). Those hatchlings found wounded and alive showed no or few presumed bite marks on the shell, those found dead showed slightly more shell damage, and those found devoid of flesh showed the greatest amount of shell removal (e.g. Fig. 2, far right shell). The 7 burrows of adult Ocypode searched for food remains and feces ranged in length from 119m to 135m. One crab was found in each of 3 burrows, and 4 lacked a crab. No burrow yielded prey remains or feces. Discussion—An attack by O. quadrata on the head probably kills or wounds the hatchling and prevents its escape. All flesh is then typically consumed. A hatchling found with flesh remaining suggests that the crab had not yet finished feeding, was interrupted in feeding by the availability of other potential prey, or was driven from the prey. 4 FLORIDA SCIENTIST [VOL 57 Fic. 2. Terrapin hatchlings fed on by ghost crab: the two on the left were fed upon by captive crab, the three on the right were found in the wild. Regarding the possibility that all or some of the hatchlings were not killed by the crab but rather were the remains of other predators, or that they were scavenged by the crab after they were found dead or wounded, strong circumstantial, in addition to the direct evidence, indicates that they were crab-caught. 1) Of 15 hatchlings examined for presumed crab bite marks on the shell, 11 showed such marks on the anterior edges of the carapace and/or plastron, or small holes in the carapace. The former marks matched those noted on the shell of the 2 hatchlings fed upon by captive crab (Arndt 1991). 2) Other possible predators on hatchlings on the island are Norway rat, raccoon, red fox, otter, and several species each of gulls, herons, and egrets (pers. obs., and N. Sterling, pers. comm. 1993). However, the small and still- soft hatchlings would probably have been swallowed intact by these birds, or been totally macerated and then ingested by the mammals, with thus no or little likelihood of a shell remaining intact. 3) I noted several cases of hatchling and crab tracks to intersect, after which the former ended, but hatchling fates are not known. 4) No proof that any vertebrate or invertebrate predator or scavenger killed and/or cleaned a dead hatchling was ever obtained. ACKNOWLEDGMENTS—I thank A. M. Teti and several Stockton State College students for assistance in the field; N. Sterling for information on local mammals; C. K. Dodd, Jr., K. H. Peterson, G. M. Tilger, and K. C. Zippel for assistance in obtaining references; J. S. Garth for information on distribution of Ocypode; R. M. Vaughn for preparing the figures for publication; an anonymous reviewer for helpful comments; and D. L. Beall, Manager, Edwin B. Forsythe National Wildlife Refuge, for permission to work on Little Beach Island. No. 1, 2 1994] ARNDT—PREDATION ON DIAMONDBACK 5 LITERATURE CITED Arnpr, R. G. 1991. Predation on hatchling diamondback terrapin, Malaclemys terrapin (Schoepff), by the ghost crab, Ocypode quadrata (Fabricius). Florida Scient. 54(3/4):215-217. BEEBE, C. W. (1903) 1904. Five days among the birds of Cobb Island, Virginia. Eighth Ann. Rep. Zool. Soc. (New York) 8:161-181. Burcer, J., AND W. A. MonTEVECCHI. 1975. Nest site selection in the terrapin Malaclemys terrapin. Copeia 1975 No. 1:113-119. BustarD, H. R. 1966. Turtle biology at Heron Island. Austral. Mus. 15(8):262-264. . 1979. Population dynamics of sea turtles, Pp. 523-540. In: HarLess M. anD H. MorLOcCK (eds.), Turtles, Perspectives and Research. John Wiley & Sons, New York, NY. 695pp. Carr, A. 1967. So excellent a fishe. The Natural History Press, New York, NY. 248pp. Comins, D. M. 1961. A note on the sand crab (Ocypode sp.) as a predator of birds in South Africa. Bull. Brit. Ornith. Club 81:111-112. Cow es, R. P. 1908. Habits, reactions, and associations in Ocypoda arenaria. Paps. Mar. Biol. Lab. Tortugas 2(1):3-41. . Dopp, C. K., Jr. 1988. Synopsis of the biological data on the loggerhead sea turtle Caretta caretta (Linnaeus 1758). U.S.F.W.S. Biol. Rep. 88(14):1-110. Ernst, C. H., AND R. W. Barsour. 1972. Turtles of the United States. The Univ. Press of Kentucky, Lexington, KY. 347pp. Frey, R. W., anD T. V. Mayou. 1971. Decapod burrows in Holocene Barrier Island Beaches and Washover Fans, Georgia. Senckenbergiana maritima 3:53-77. HENDRICKSON, J. R. 1958. The green sea turtle, Chelonia mydas (Linn.), in Malaya and Sarawak. Proc. Zool. Soc. Lond. 130(4):455-535(+10 plates). MILNE, L. J., AND M. J. MILNE. 1946. Notes on the behavior of the ghost crab. Amer. Nat. 80(792):362- 380. Puituips, N. J. 1987. The breeding biology of white-tailed tropicbirds Phaethon lepturus at Cousin Island, Seychelles. Ibis 129:10-24. SpRUNT, A. JR. 1948. The tern colonies of the Dry Tortugas Keys. Auk 65(1):1-19. STERLING, N. 1993. Trapper, Port Republic, NJ. Pers. comm. Florida Scient. 57(1,2):1-5.1994 Accepted: April 10, 1993. 6 FLORIDA SCIENTIST [VOL 57 REPRODUCTIVE CYCLE OF THE INDO-PACIFIC GECKO, HEMIDACTYLUS GARNOTII, IN SOUTH FLORIDA WALTER E.. MESHAKA, JR. Archbold Biological Station P.O. Box 2057 Lake Placid, Florida 33852 Apstract: The reproductive cycle of the Indo-Pacific gecko, Hemidactylus garnotii, was studied in southern Florida from June 1991 to January 1993. Lizards were sexually mature at 49 mm SVL. Egglaying took place every month of the year and at least three clutches were produced annually. Year round reproduction may be added to the list of reproductive traits which make H. garnotii the most fecund and widespread species of the hemidactyline geckoes currently found in southern Florida. First recorded in Miami, Dade Co., Florida (King and Krakauer, 1966), the Indo-Pacific gecko H. garnotii has proven to be a successful colonizer of buildings in southern Florida (Voss, 1975; Wilson and Porras, 1983). Of the seven introduced species of geckoes known to Florida (Gekko gekko, Gonatodes albagularis, H. garnotii, H. mabouia, H.turcicus, Spherodactylus argus, and S. elegans), H. garnotii is currently the most widespread in its Florida distribution (Conant and Collins, 1991). Parthenogenesis in this species is thought to be a major cause for its rapid geographic expansion (Kluge and Eckardt, 1969; Wilson and Porras, 1983). In this paper, I present data on the gonadal cycle of H. garnotii from southern Florida which I relate to its superior colonization ability over Hie other introduced hemidactyline geckoes, H. turcicus and H. mabouia. METHODS—Visits were made biweekly from June 1991 to January 1993 to 18 buildings in Everglades National Park (Long Pine Key: and Florida B: ay) and in South Miami. Lizards were collected by he aud fixed in 10% formalin, ‘andl stored in 70% ETOH. Snout- vent length (SVL) was measured with vernier calipers, and diameters of ovarian follicles were measured with an ocular micrometer. All specimens will be deposited in the United States National Museum (USNM). Resu_ts—A total of 91 lizards was collected during the 21 months of this study (Fig. 1). Mean (+ 2 standard deviations) adult snout-vent length (SVL) was 55.0 + 2.95 mm (Range = 49.2-61.0 mm; N=52). Mean hatchling SVL was 23.7 + 1.40 mm (Range = 22-26; N= 6), and mean shelled egg diameter was 9.4 + 0.69 mm (Range= 7.9-10.2; N= 16). Sizes of hatchlings and shelled eggs from this study were similar to those eatiee (24-26 and 7-10 mm, respectively) reported by Voss (1975) from Miami. Based on the body sizes and reproductive condition of specimens in monthly samples (Fig. 1), it appeared that sexual maturity could be reached before the end of their first year at 49 mm SVL. Adults were active year round and contained enlarged follicles and/or shelled eggs each month (Fig. 2), even in months when adults were observed but not collected (Feb. 1992, Dec. 1992). The presence in some individuals of up to two additional sets of smaller follicles (1.6-2.0 mm) besides a set of enlarged follicles or shelled eggs indicated that at least three clutches were possible during the year. ~l No. 1, 2 1994] MESHAKA—INDO-PACIFIC GECKO 65 60 55 50 45 40 mm SVL 35 30 25 MONTH Fic. 1. Snout-vent lengths (SVL) of Hemidactylus garnotii collected in southern Florida from June 1991 to January 1993. Triangles represent immatures; squares, mature lizards. 11 10 a 9 & ia 8 a OOWMd DIAM. (mm) OF FOLLICLES 7 8 9 10 11 12 6 MONTH Fic. 2. Ova diameters of Hemidactylus garnotii collected in southern Florida from June 1991 to January 1993. Hollow squares represent enlarged follicles; solid squares, shelled eggs. 8 FLORIDA SCIENTIST [VOL 57 Discussion—Results of this study indicate that H. garnotii is a highly fecund species and possibly the most fecund of the three hemidactyline gecko species presently known to occur in Florida. Adults of H. garnotii in southern Florida could potentially lay more than three clutches annually if interclutch time was less than 4 months. This is likely, based on an interclutch time for H. turcicus of only 45 days found during the May to August egglaying season in Texas (Selcer, 1986). Thus, under a best case scenario, every adult H. garnotii could produce 16 eggs annually (2 eggs every 45 days). To compare, only a maximum of 6 eggs could be produced seasonally by one half of the population of H. turcicus in Louisiana (Rose and Barbour, 1968) and Texas (Selcer, 1986), and this estimate probably applies to southern Florida as well where H. turcicus is also a seasonal breeder (pers. obs.). Hemidactylus garnotii shares with the other hemidactylines introduced in south Florida the ability to thrive around buildings where food is abundant throughout the year (Rose and Barbour, 1968; Selcer, 1986; Wilson and Porras, 1983). For example, in Everglades National Park, areas surrounding lit walls of buildings may harbour up to 450 arthropods/ m2. In addition to this habitat adaptability, H. garnotii would be predicted to be a more successful colonizer than the other species based on its reproductive traits. These include paedogenesis (Kluge and Eckardt, 1969; Wilson and Porras, 1983) and production of eggs adapted for survival and dispersal (Bustard, 1968) together with early sexual maturity and the potential for year round egg production shown by this study. In this connection, it is interesting that of the 18 randomly-selected sites I sampled, H. garnotii was the only hemidactyline present at 15 (83 %) and occurred with H. mabouia at 3 (17 %), where it was numerically superior. Hemidactylus turcicus was absent from all sites. The ubiquity of H. garnotii in southern Florida is especially striking in light of its first recorded appearence in mainland Florida more than 50 years after that of H. turcicus in the Florida Keys (Fowler, 1915) and at least 30 years after the first recorded presence of H. turcicus in mainland Florida (Barbour, 1936). Hemidactylus mabouia, also a recent invader (Lawson et al., 1991), is also a year round breeder (Butterfield et al., 1993; Vitt, 1986) and more common within its range than H. turcicus (Butterfield et al., 1993). Opportunity to disperse notwithstanding, other life history requirements may play a greater role than fecundity in determining the colonization success of these species in drier (Florida Keys or western U.S.) and cooler (northern U.S.) locations where conditions may not be so amenable to H. garnotii as they are in southern Florida. ACKNOWLEDGMENTS—Collection of specimens in Everglades National Park was made possible by collection permit # 910039, United States Department of the Interior. I wish to thank Dan Foxen, William B. Robertson, Jr., and Mike Soukup of Everglades Research Center for their encouragement and support regarding this project in Everglades National Park. Samuel D. Marshall commented on an earlier draft of this manuscript. No. 1, 2 1994] MESHAKA—INDO-PACIFIC GECKO 9 LITERATURE CITIED Barzour, T. 1936. Two introduced lizards in Miami, Florida. Copeia 1936:113. BustarD, H. R. 1968. The egg shell of gekkonid lizards: a taxonomic adjunct. Copeia 1968:162-164. BuTTerFIELD, B. P., J. B. Hauce, anD W. E. MeEsnaka, Jr. 1993. The occurrence of Hemidactylus mabouia on the United States mainland. Herp. Review 24:111-112. Conant, R., AND J. T. CoLtins. 1991. A Field Guide to Reptiles and Amphibians of Eastern and Central North America. 3rd ed. Houghton Mifflin Company, Boston, MA. 450 pp. Fow er, H. W. 1915. Cold-blooded vertebrates from Florida, the West Indies, Costa Rica, and Brazil. Proc. Acad. Natur. Sci. Philadelphia 67:244-269. Kine, W. AND T. Krakauer. 1966. The exotic herpetofauna of southeast florida. Quart. J. Florida Acad. Sci. 29(2):144- 154. KucE, A. G., AND M. J. Eckarpr. 1969. Hemidactylus garnotii Dumeril and Bibron, a triploid all-female species of gekkonid lizard. Copeia 1969:651-664. Lawson, R, P. G. FRANK, AND D. L. Martin. 1991. A gecko new to the United States herpetofauna, with notes on geckoes of the Florida keys. Herp. Review. 22:11-12. Rose, F. L. anp C. D. Barsour. 1968. Ecology and reproductive cycles of the introduced gecko, Hemidactylus turcicus, in southern United States. Amer. Midland Natural. 79(1):159-168. SELCER, K. W. 1986. Life history of a successful colonizer: the mediterranean gecko, Hemidactylus turcicus, in southern Texas. Copeia 1986:956-962. Vir, L. J. 1986. Reproductive tactics of sympatric gekkonid lizards with a comment on the evolutionary and ecological consequences of invariant clutch size. Copiea 1986:773-786. Voss, R. 1975. Notes on the introduced gecko, Hemidactylus garnotii in south florida. Florida Scient. 38(3):174. Witson, L. D. aNnD L. Porras. 1983. The ecological impact of man on the south Florida herpetofauna. Univ. Kansas Mus. Natl. History Special Publ. No.9. 89 pp. Florida Scient. 57(1,2):6-9.1994 Accepted: October 26, 1993. 10 FLORIDA SCIENTIST [VOL 57 Environmental Chemistry EFFECT OF TREATED KRAFT BLACK LIQUOR ON HYDRILLA VERTICILLATA (ROYLE) ANDREW L. HassEL_L”, BARBARA B. MarTIN), DEAN F. Martin”, and Jess M. VAN Dyke”? ‘Institute for Environmental Studies, Department of Chemistry, University of South Florida, 4202 East Fowler Avenue, Tampa, FL 33620 Bureau of Aquatic Plant Management, Florida Department of Environmental Protection, 3917 Commonwealth Boulevard, Tallahassee, FL 32399 Asstract: Kraft black liquor (KBL), a by-product of the paper pulp industry, was neutralized, treated and diluted to produce a material (10% neutralized KBL) that has the potential to retard the growth of Hydrilla verticillata (Royle) by shading. Phase one of the research is concerned with the mechanism of action: light reduction versus chemical toxicity. A Warburg bioassay indicated that addition of the dark liquid (10% neutralized KBL) inhibited oxygen production within an hour. A non-contact bioassay indicated the dark liquid significantly affected the light that hydrilla could receive; both the percent fresh weight change of hydrilla and light intensity were directly related to the log of the black liquor concentration (dilution with 10% Hoaglands). Change in fresh weight of hydrilla in 10% Hoaglands and diluted black liquor mixture was measured every 3-4 days. Pseudo-first-order rate constants for fresh weight change were calculated and found to be inversely related to the black liquor concentration, another indication of shading as a major, but not exclusive, mechanism of action. Hypritia, Hydrilla verticillata (Royle), is a perennial submersed, exotic, rootable nuisance aquatic plant that was first identified in Florida waters in 1959-60 (Holm et al., 1969). Subsequently, the plant was introduced to other waterways in the southeastern United States, Texas, California, North Carolina, and the Potomac River (Cook, 1985; Steward et al., 1984; Steward, 1991). Hydrilla is able to dominate the macrophytic community, owing to a mat-forming capability (Haller and Sutton, 1975), as well as some unique physiological properties (Bowes et al., 1979; Van et al., 1976). Because of the mats, hydrilla has caused serious problems for navigation, recreation, and flood control. Control of hydrilla has been achieved using biocontrol (chiefly herbaceous fish), chemicals (Sonar®, among other herbicides), and by physical-chemical approaches, such as shading. One product, Aquashade®, a mixture of approved inert dyes, controls the growth of hydrilla by limiting the availability of photosynthetically-active region light (see Martin and Martin, 1992 for a summary of the pertinent literature and Osborne, 1979 for a practical example). A previous study (Manker and Martin, 1984) demonstrated that though some dyes can serve as sensitizers and produce singlet oxygen, this mechanism does not apply to Aquashade® and the function seemed to be solely limitation of light. The current study was initiated because of an on-going interest in whether natural products might serve a shading function and/or be growth inhibitors. We studied inhibition of hydrilla growth in certain colored lakes (Dooris and Martin, No. 1, 2 1994] HASSELL ET AL.—EFFECT OF KRAFT BLACK LIQUOR 1] 1980), tried to “fingerprint” the responsible agents (Martin et al., 1986), and to identify significant responsible structural features through the use of model com- pounds (Martin and Martin, 1988). In addition, previous research (Pompey and Martin, 1993) showed that cellulose-utilitizing organism(s) isolated from Lake Starvation produced organic compounds that inhibited growth of hydrilla. Because the effects of reduced light on hydrilla have been demonstrated, finding less expensive natural products that reduce light pentration may also be useful in controlling hydrilla. Thus, we have investigated the effects of natural shading agents on hydrilla growth. Another candidate for physical-chemical control is a material called kraft black liquor (KBL), a product of the paper pulp industry that results from the treatment of wood with a sodium hydroxide/sodium sulfide mixture (pulping liquor) to remove lignin from wood and release the cellulose (McKean, 1980). Wood fibers are washed free of used pulping liquor and dissolved wood substances (lignin and hemicellulo- ses). About 96-98% of the dissolved wood constituents are recovered in effluent from the first wash stage, and this material is called black liquor. Typically, it is burned for steam to generate power, and the inorganic chemicals are recovered for recycling. Typically, no black liquor enters the raw waste sewer (McKean, 1980). Thus KBL, suitably treated, would seem to be a potentially useful source of chemicals for inhibition of hydrilla growth, either by physical means (“shading”) or chemical means, as was found for our natural inhibitors (Pompey and Martin, 1993). In this study, we investigate the treatment of this material and conduct bioassays using hydrilla to determine the importance of shading versus chemical action. MaTERIALS AND METHODs—Source of kraft black liquor—A 15-L sample of kraft black liquor was provided by Stone Container Corporation, Panama City. The sample had been treated by soap and turpentine removal, but it had not been subjected to bleaching treatment. The sample, as received, was analyzed by ABC Research, Gainesville, FL for heavy metals, selected pesticides and herbicides. The sample was subjected to the following treatment (Fig. 1). The sample was diluted 1:10 with (well or tap) water to give a 150-L sample that was slowly aerated outside. During a two-week period, the solution was slowly neutralized with concentrated (18 M) sulfuric acid. Aerobic digestion was continued for a three-month period to reduce biological oxygen demand, (BOD) and reduce toxic organic compounds. The resulting product was called 10% KBL. The sample 10% KBLwas analyzed for total and inorganic carbon using a Beckman Model 915 total carbon analyzer. Potassium hydrogen phthalate was the total carbon standard and potassium bicarbon- ate was the inorganic carbon standard. Spectra—The ultraviolet and visible spectra of diluted samples of kraft black liquor were measured using a 1-cm cell and a Beckman DU-64 recording spectrophotometer. Hydrilla source—Plants were obtained from a site in the Hillsborough River about 400 m north of the Fowler Avenue overpass. In the laboratory, the plants were washed thoroughly with tap water, then with deionized water, then placed in an aquarium containing water from the Floridan aquifer, obtained using the Science Center pump. A week before bioassay began, sprigs of the hydrilla were placed in 10% Hoagland’s solution (Steward and Elliston, 1973), a standard defined medium. Bioassays of the treated liquid —The effect of the treated liquid on hydrilla was studied using a Warburg apparatus as before (Barltrop et al., 1984). Specifically, relative rates of oxygen production were measured by treating hydrilla leaves with a 1 to 50 dilution of the liquid. The hydrilla leaves were suspended in 4 ml of 10% Hoagland’s solution, and the rate of oxygen production was measured. The black liquor (1 ml of 1 to 10 dilution, effectively diluting the KBL liquid to 1 to 50) was added and the rate was determined again. Effect of shading—The method of Manker and Martin (1984) was used to evaluate the effect of shading of the treated liquor. Hydrilla sprigs (approximately 1 g.) were placed in 500-ml. flasks containing 10% Hoagland’s solution. The flasks were inverted so they were hanging in a rack and 12 FLORIDA SCIENTIST [VOL 57 KRAFT BLACK LIQUOR Treatment Soap and turpentine removed TREATED BLACK LIQUOR |—————»Heavy metal/pesticide/ herbicide analysis Dilute 1:10 Aerate Neutralize 18 M Sulfuric Acid ——————- Acceptable pH Aerat eae ———__§_§—-» Acceptable BOD Bioassay with TREATED, NEUTRALIZED , Hydrilla 10% BLACK LIQUOR Fic. 1. Treatment of kraft black liquor prior to dilution for hydrilla bioassay. shielded from light except through the flask bottom (Fig. 2) containing four replicates each of control (deionized water), 1:50 dilution and 1:200 dilution of the black liquor. These flasks were randomly placed above the hydrilla flasks. Two fixtures containing cool-white fluorescent lamps set to a 12-hour photoperiod were suspended above the flasks, which were filled to the rim and covered with watch glasses. Direct bioassay—The growth of hydrilla in the presence of dilutions of treated black liquor was measured as before (Dooris and Martin, 1980; Martin and Martin, 1988; Pompey and Martin, 1993) using 500 ml Erlenmeyer flasks stoppered and inverted and illuminated with 40-W cool-white fluorescent lamps (150uEs/m/sec’), 1-gram (fresh weight apical tips), and 10% Hoagland’s (Steward and Elliston, 1973). Each dilution was replicated four times. Black liquor was diluted with 10% Hoagland’s (1:50, 1:100, and 1:200) for test system samples. The control medium consisted of 10% Hoagland’s. No. 1, 2 1994] HASSELL ET AL.—EFFECT OF KRAFT BLACK LIQUOR i) Hydrilla samples were weighed every four days during the two-week bioassay and were returned to the flasks. ResuLts—Analyses—Several analyses indicate significant chemical properties of the material that has been neutralized and digested. The concentrated sample (original 15-L sample) was analyzed for heavy metals (Table 1). Organophosphorus and chlorinated pesticide residues [Aldrin, benzene hexachloride (BHC), DDD, DDE, DDT, Diazinon, Dieldrin, Endrin, Ethyl Parathion, Ethion, Guthion (Azinophos-methyl), Heptachlor, heptachlorepoxide, hexachlorobenzene (HCB), and Lindane, Malathion, Methoxychlor, Mirex, Trithion (carbophenothion)], were also analyzed but were present at less than the detection limit 0.01 ppm. Dioxins were less than 100 pg/L detection limit, and chlorophenoxyherbicides (2,4-D, Silvex, and 2,4,5-T), were present at less than 0.02 ppm. Lead analyses performed on KBF at Stone Container Corporation were non-detectable (indicated less than 2mg/L) for a typical sample and two other mill liquors (Berdon, 1993). Aerobic digestion of kraft effluents for only 5-10 days reduced BOD by 90% and acute toxicity to zero (Borton, 1991). Total carbon in the final sample (Fig. 1), used for bioassays was measured. This sample had 58 ppm total carbon (1:500 dilution) and 3.] mg/L inorganic carbon (1:50 dilution), and thus the 10% KBL sample contained 28,900 mg/L organic carbon and 150 mg/L inorganic carbon. Spectra—A 1:1000 dilution sample had an absorbance of 0.470 at 280 nm. The Fic. 2. Schematic representation of the non-contact bioassay to test the effect of shading by dilutions of kraft black liquor. 14 FLORIDA SCIENTIST [VOL 57 TABLE 1. Heavy metal content of kraft black liquor*> Component Concentration, mg/kg Aluminum < 1.81 Antimony fo) = [—) ° x< (e) = oe ° © = (—) Intensity, uEs/m?2/sec Fic. 5. Percent growth as a function of time and dilution for four different black liquor concentrations. Data were used to calculate pseudo-first order rate constants. These results indicate that treated black liquor affects the growth of hydrilla, chiefly through the shading effect, and dilutions of 0.005 (1/200) produce a notable effect on reduction in growth constant (from 0.149 to 0.106 day"). A dilution of 1/100 (0.01) reduced the constant to 0.051 day". The interpretation of the third bioassay is complicated by the fact that it is difficult to separate the effect of shading from a chemical effect because of the impossibility in determining what amount of light the 18 FLORIDA SCIENTIST [VOL 57 hydrilla is receiving under these conditions, more than with the second bioassay, of course, but by how much is uncertain. Discussion—Granted that the effect of the kraft black liquor is chiefly, if not exclusively, a shading effect, what are the implications of these results? Several come to mind: the possibility of some chemical effects, the extent of light limitation as a practical result, and the effect of organic matter on other organisms present. These implications can be considered in order. First, the possibility of some chemical effects cannot be absolutely excluded. Two come to mind. The first is chelation, possibly by weak aromatic phenolic and/ or carboxylic acids. Given the organic content, and the charge-transfer band that was observed in the ultraviolet spectrum, it would be surprising if some aromatic chelating agents were not observed. Previously, we have studied examples of humic- and lignin-like materials using such model compounds as dihydroxybenzoic acids and cinnamic acids. The results indicate that the percent inhibition of growth (fresh- weight change) for hydrilla was a linear function of the pK, for the weak acid (Martin and Martin, 1989). A second chemical effect is the action of a colored organic material serving as a sensitizer and converting triplet or ordinary oxygen to the reactive form called singlet oxygen (Bland, 1976), which results in the light-lethal effect called photodynamic action. One hydrilla-growth inhibitor was demonstrated to be a singlet oxygen generator (Barltrop and Martin, 1983). The material inhibited the germination of lettuce seeds, the effect was concentration dependent, was observed in the light, but not in the dark, and was inhibited by sodium azide (Barltrop and Martin, 1983). The absence of intense absorption in the visible region raises questions about the probability that kraft black liquor can function in this manner. The practical aspects of light limitation as a means of controlling hydrilla must be considered. This approach was effective using Aquashade® (Osborne, 1979; Martin and Martin, 1992). Previous research (Van et al., 1976) indicated that about 15 uEs/m?/sec was the amount of light required by hydrilla to attain the compensa- tion point. Our results indicate that dilutions of 1:50 of treated 10% KBL were sufficient to reduce light levels below compensation point light for hydrilla in these experiments. Organic color was one of the factors used to predict Secchi disc depths in Florida lakes (Canfield and Hodgson, 1983). Using 205 lakes, the best estimate of the lake Secchi disc depths could be obtained from a statistically derived relationship (Eqn. DF In (SD) = 2.01 -0.370 In (Chla) -0.278 In (C) (1) Here, SD is the Secchi disc depth (in meters), Chla is the chlorophyll a content (mg/m*), and (C) is the organic color concentration expressed as ppm (platinum units). This equation “indicates organic color concentrations do not affect lake Secchi disc depths as much as algal levels” (Canfield and Hodgson, 1983). The effect of KBL on the dissolved oxygen content of receiving waters should No. 1, 2 1994] HASSELL ET AL.—EFFECT OF KRAFT BLACK LIQUOR 19 not be a problem because the material has been treated (Fig. 1) to produce an acceptable D. O., and previous studies have been thorough, as the following examples should indicate. The effect on macroinvertebrates has been considered by others. For example, the effects of the discharge from an unbleached linerboard mill on characteristics of the Withlacoochee River were summarized (Holton et al., 1992). These discharges contained some KBL, but were not exclusively KBL. Community structure was evaluated using Hester-Dendy multiple plate samplers at four locations below and four above the discharge point. Data were analyzed for species richness and abundance, using the Shannon-Weaver diversity, and using the Florida Index. There was no significant impact on the downstream portion as a result of the discharge. Another study was concerned with effects of biologically treated bleached kraft mill effluent. Experimental streams (model systems constructed near New Bern, North Carolina) were studied from 1979 to 1981 using the periphyton community (aquatic organisms that grow on submerged substrates; bacteria, yeasts and molds, fungi, protozoans, algae, small invertebrates). The community includes primary producers, primary consumers, and decomposers. The report (Rupp and Borton, 1984) noted, “No clear pattern of effluent concentrations or color addition...significantly affected the ash-free dry weight of periphyton collected at any of the four depths studied. At the highest color additions tested no consistent differences in chlorophyll a biomass was observed until depths of 30 to 35 cm.” At greater depths, a reduction in overall autotrophic organisms (chlorophyll a biomass), but nota reduction in the total biomass of secondary producers, was observed (Rupp and Borton, 1984). Suspended solids in the effluent may have served as food sources to offset the effect of diminished algal growth. This indicated that the effects they saw on reduced algal growth at greater depths were due to light reduction, not toxicity effects. The 1993 meeting of the American Society of Limnology and Oceanography had a session on large rivers and pulp mills. Hall and Haley (1993), for example noted that secondary treated bleached kraft mill effluent has effects on periphyton that were “a function of reduced light availability and not toxicity.” On the other hand, Zimmerman and Livingstone (1976ab, 1979) have investi- gated the effects of kraft mill effluents on benthic organisms in shallow bays. Their work indicates the need for appropriate caution in treating kraft liquor and any proposed uses. The research described here and that done by others indicates that kraft black liquor has the potential to reduce Hydrilla verticillata growth under appropriate conditions by attenuation of light, and the research adds to the information gained through previous studies of naturally occurring hydrilla growth inhibitors (cf Dooris and Martin, 1993). What those appropriate conditions might be, however, requires further reflection as well as further testing and more extensive bioassays. ACKNOWLEDGMENTS— We thank Shannon Smith, who assisted with one of the bioassays, and we are grateful to Drew Leslie, who supplied analytical results from ABC research. We thank Bill Berdon of Stone Container Corporation for supplying the original study of kraft black liquor, and we are grateful to a reviewer for very helpful comments. We are grateful to Dr. P. M. Dooris, who served as Consulting Editor for this manuscript. 20 FLORIDA SCIENTIST [VOL 57 LITERATURE CITED Barxtrop, J. AND D. F. Martin. 1983. Evidence for photodynamic action by a naturally occurring hydrilla-growth inhibitor. J. Environ. Sci. Health A18:29-36. , B. B. Martin, AND D. F. Martin.1984. Activity of naturally occurring Hydrilla growth inhibitor: initial studies. J. Aquat. Plant Manage. 22: 84-87. Berpon, W. H. 1993. Technical Director, Panama City Mill, Stone Container Corporation, P. O. Box 2560, Panama City, FL 32402. BLAND, J. 1976. Biochemical effects of excited state molecular oxygen. J. Chem. Educ. 53:274-279. Borton, D. 1991. NCASI, P. O. Box 12868, New Bern, NC 28561, Pers. Comm. Bowes, G., A. S. HoLiDay, AND W. T. HALLER.1979. Seasonal variation in the biomass, tuber density, and photosynthetic metabolism of hydrilla in three Florida lakes. J. Aquat. Plant Manage. 17: 61-65. CanFIELD, D. E. anD L. M. Hopcson. 1983. Prediction of Secchi disc depths in Florida lakes: Impact of algal biomass and organic color. Hydrobiolog. 99:51-60. Cook, C. D. K. 1985. Range extensions of aquatic vascular plants. J. Aquat. Plant Manage. 23: 1-6. Doorts, P. M. anD D. F. Martin. 1980. Growth inhibition of Hydrilla Verticillata by selected lake sediment extracts. Water Res. Bull. 16: 112-117. AND D. F. Martin. 1993. Studies of the natural control of hydrilla in Florida. Aquatics [5C)\2 17-20: Finney, D. J. 1971. Probit Analysis. Cambridge University Press. Cambridge, U.K. Hau, T. M. AND R. K. Hatey. 1993. Effects of treated bleached kraft mill effluent on periphyton in outdoor experimental streams. Abstr. American Society of Limnology and Oceanography, Edmonton, Alberta, Canada, May 30-June 3. HALLER, W. T. anp D. S. Surron. 1975. Community structure and competition between Hydrilla and Vallisneria. Hyacinth Contr. J., 13: 48-50. Hou, L. G., L. W. WELDON, AND R. D. BLackBurn. 1969. Aquatic weeds. Science 166: 699-709. Ho tron, V. C., H. D. Putnam, AND D. L. Evans. 1992. Effects on the macroinvertebrate community in a north Florida river from the discharge of an unbleached linerboard mill. Pp. 183-204. In: 1992 Environmental Conference, April 12-15. Tappi Press, Atlanta, GA. MaNkKER, D.C. anp D. F. Martin. 1984. Investigations of two possible modes of action of the inert dye Aquashade® on hydrilla. J. Environ. Sci.. Health A19: 433-443. Martin, B.B. AND D. F. Martin. 1988. Influence of substituted phenols. J. Aquat. Plant Manage. 26: 74- TD. AND D. F. Martin. 1989. Bio-assays of humic-like model compounds: chelation versus acid-base effects. J. Environ. Sci. Health. A24: 167-174. Martin, D. F. anp B. B. Martin. 1992. Aquashade®: an annotated bibliography. Florida Scient. 55(4): 264-266. , P. M. Dooris, G. M. Dooris, anp R. J. Bova, JR. 1986. Analysis of hydrilla-inhibiting fractions in natural waters. The concept of “fingerprinting” through liquid chromatography. Water Res. Bull. 22: 283-287. McKean, W. T. 1980. Pulp and paper industry. Pp. 210-226. In: Guturik, F. E. anp J. J. Perry (eds.). Introduction to Environmental Toxicology. Elsevier. New York, NY. OspornE, J. A. 1979. Use of Aquashade® to control the reinfestation of hydrilla after herbicide treatment. Aquatics 1(4): 14-15. Pompey, B. AND D. F. Martin. 1993. Effects of metabolic products of cellulose-utilizing organisms on Hydrilla. J. Aquat. Plant. Manage. 31:109-113. Rupp, R. E. anp D. L. Borton. 1984. Effects of biologically treated bleached kraft mill effluent on the periphyton community in southern experimental streams for 1979 to 1981. Tech. Bull 423. NCASI, 260 Madison Ave., New York, NY. STEWARD, K. K. 1991. Light requirements for monoecious Hydrilla from the Potomac River. Florida Scient. 54; 204-214. AND R. A. ELLIsTon. 1973. Growth of Hydrilla in solution culture at various nutrient levels. Florida Scient. 36: 228-233. , T. K. VAN, V. CARTER, AND A. H. PIETERSE, 1984. Hydrilla invades Washington, D.C., and the Potomac. Amer. J. Bot. 71: 162-163. Van, T. K., W. T. HALLER, AND G. Bowes. 1976. Comparison of the photosynthetic characteristics of three submersed aquatic plants. Plant Physiol. 58: 761-768. No. 1, 2 1994] HASSELL ET AL.—EFFECT OF KRAFT BLACK LIQUOR 2] ZIMMERMAN, M. S. AND R. J. Livincstone. 1976a. The effects of kraft mill effluents on benthic macrophyte assemblages in a shallow bay system (Apalachee Bay, north Florida, USA). Mar. Bio. (Berlin) 34:297-312. AND R. J. Livincstone, 1976b. Seasonality and physico-chemical ranges of benthic macrophytes from a north Florida estuary (Apalachee Bay). Contrib. Mar. Sci. Univ. Tex. 20:34- AND R. J. Livincstong, 1979. Dominance and distribution of benthic macrophyte assemblages in a north Florida estuary (Apalachee Bay, Florida). Bull Mar. Sci. 29:27-40. Florida Scient. 57(1,2):10-21.1994 Accepted: December 2, 1993. 99 FLORIDA SCIENTIST [VOL 57 Biological Sciences MACROINVERTEBRATES OF THE NORTHERN EVERGLADES: SPECIES COMPOSITION AND TROPHIC STRUCTURE RUSSELL B. RADER USDA, Forest Service, Rocky Mountain Experiment Station, 222 South 22nd Street, Laramie, WY 82070-5299 AsstrAct: Macroinvertebrates of the northern Everglades are adapted to temporary waters with fluctuating oxygen concentrations, and periodic high (> 35. 0 C) temperatures. Despite harsh conditions, the macroinvertebrate assemblage was surprisingly diverse with species richness values similar to well- oxygenated, lotic ecosystems. One hundred and forty-eight taxa of macroinvertebrates were identified from sloughs in the northern Everglades. Although the diversity and density of macroinvertebrates in sloughs were dominated by Chironomidae, Gastropoda, and Coleoptera, the amphipod Hyallella azteca was the most abundant species collected. Macroinvertebrate diversity and density was reduced in dense sawgrass. Almost all taxa, however, were well represented within man-made canals where amphipods and the > freshwater shrimp Palaemonetes paludosus were extremely abundant. Trophic categories were dominated by grazers and collector-gatherers. Energy flow was equally distributed through grazer and detritivore pathways. Macroinvertebrates of the northern Everglades are characterized by common species from North America and the southeastern United States with a few colonists from Central and South America. However, two species, the most abundant snail Planorbella duryi and the mayfly Callibaetis floridanus, are endemic to the Florida peninsula. MACROINVERTEBRATES Constitute an important link between primary producers and higher trophic levels. In freshwater wetlands, including the Everglades, inver- tebrates form the energy base and major food resource of most birds, fish, and mammals (Mitch and Gosselink, 1986; Kushlan, 1991). In addition, macroinvertebrates are important indicators of ecological integrity (e.g. Rosenberg and Resh, 1993). Alterations in patterns of diversity, relative abundance, and spatial and temporal distribution provide valuable insight as early warning signals of environmental degradation. Despite its prominence as one of the largest and best known wetlands in the world, the Everglades invertebrate fauna remains relatively unknown. Rader and Richardson (1992) reviewed previous reports and available information concerning both macroinvertebrates and algae in the Everglades. Based on hydroperiod and dominant macrophytes, the Everglades can be divided into 4 habitats: sloughs, wet prairies, dense sawgrass, and tree islands (Gleason, 1974). Wet prairies are transition zones between sloughs and dense sawgrass stands. Because wet prairies are open water habitats that support abundant growths of algae, submersed vegetation, and macroinvertebrates, they were not distinguished from sloughs in this study. Samples were collected in both wet prairies and sloughs (open water habitat) with no effort to distinguish between them. Tree islands were not sampled because they represent a small fraction of the landscape in the northern Everglades. Early investigations indicate that sloughs were the center No. 1, 2 1994] RADER—MACROINVERTEBRATES 23 of biological diversity in the Everglades (Reark, 1961, Craighead, 1968; Gleason, 1974). Therefore, I have focused my efforts in sloughs while making additional comparisons to sawgrass and man-made canal habitats. The purpose of this study was to characterize and quantify the taxonomic composition and trophic structure of macroinvertebrates in the northern Ever- glades. Specifically, I sought to answer three questions: (1) Is macroinvertebrate diversity and density low compared to other freshwater environments?, (2) Are macroinvertebrates of the Everglades peninsula isolated and composed of rare and endemic species?, and (3) Is the macroinvertebrate assemblage composed of species that primarily consume detritus or are grazers also abundant? What is the relative abundance of grazers compared to detritivores? MeETHops—Macroinvertebrates were collected bimonthly (6 sampling dates) for one year (1990) at eight sites established along a transect running south through the northern half of Water Conservation Area 2a (WCA-2A; Fig. 1). Sampling was confined to open water, slough habitats. The environmental characteristics of each site and differences in macroinvertebrates between enriched and unenriched locations have been previously described (Rader and Richardson, in press). Six to eight samples were collected from submersed vegetation at each site on each date using a D-framed sweep net (2.0 - 2.5 mm mesh). Because sweeps samples filter large volumes of water, trap mobile invertebrates that easily avoid other methods, and facilitate access to all available sub-habitats, they are the best method of determining macroinvertebrate species composition and abundance in wetlands (Cheal et al., 1993). Macroinvertebrate density estimates (numbers/m?) were calculated from sweep samples by determining the volume of water filtered; the distance of the sample (usually 1 m) multiplied by the area of the net opening (0.04 m2). When possible, all macroinvertebrates collected from sloughs were identified to species. Both local and regional keys were used for species identifications (e.g. Young, 1954; Thompson, 1984; Merritt and Cummins, 1984; Epler, 1992). Mean annual density estimates of the species list from each site were used to calculate Shannon’s diversity index (Shannon and Weaver, 1963). Trophic structure was based on sweep samples from open water, slough habitats. Each species was classified into functional feeding groups based on their feeding mode and type of food consumed (Merritt and Cummins, 1984; Pennak, 1989). “Collector-Gatherers” collect and gather dead organic particles. “Grazers” remove algae from solid surfaces. “Herbivores” and “Predators” consume living macrophyte tissue and animal tissue, respectively. “Filterers” remove seston from the water column whereas “Shredders” reduce coarse particulate organic material into fine particles. Although some taxa were assigned to more than one functional feeding group, the first category listed represents the major feeding mode. Macroinvertebrates were also collected from dense sawgrass and canal habitats for comparison to the open water sloughs. Twelve sweep samples were collected at 7 sites in dense sawgrass (Cladium jamaicense Crantz) habitats of WCA-2A during August of 1990 (Fig. 1). During May of 1992, 12 sweep samples were also collected in each of 3 vegetation types (Grass, Floating Plants, and Reeds) located along the margins of Hillsboro Canal (Fig. 1). “Grass” and “Reeds” habitats were primarily composed of Panicum spp. and Phragmites spp., respectively. The dominant floating plants were water-lettuce (Pistia stratiotes L.) and water-hyacinth (Eichhornia crassipes Solms.). All invertebrates collected from sawgrass and canal habitats were enumerated, but were not identified below order or family. Statistical comparisons of macroinvertebrates among different habitats were not attempted because sawgrass and canal habitats were only sampled on a single date. ResuLts—Based on sweep samples, 148 macroinvertebrate taxa were identi- fied, including 83 species identifications, from slough habitats in the northern Everglades (Table 1). Shannon’s (1963) diversity index calculated for each site ranged from 2.79 to 3.08. Diptera (44 taxa), Coleoptera (39 taxa), Gastropoda (16 taxa), and Oligochaeta (11 taxa) were the most diverse groups comprising over 74% of the total number of taxa (Table 1). Several chironomid genera (e.g. Tanytarsus, 24 FLORIDA SCIENTIST [VOL 57 x Obhwh-= oF \ HILLSBORO CANAL f GENERAL ae DIRECTION OF Sg 2 sL° ° WATER Sg FLOW Fic. 1. Water Conservation Area 2A (WCA-2A) showing the location of slough (“sl”), dense sawgrass (“sg”), and canal (“c”) sites. Polypedilum, Kieferulus, and Parakiefferiella) are comprised of numerous undescribed species. This study used designations by Epler (1992) for undescibed larval chironomids (e.g., Tanytarsus sp. A). Because of a relatively large mesh size (sweep samples), smaller species of some groups (Chironomidae, Oligochaeta, Ostracoda, and Nematoda) were likely under-represented. Although the amphipod Hyallella azteca Saussure was the most abundant species collected from sweep samples in open water sloughs, Diptera, Gastropoda, No. 1, 2 1994] RADER—MACROINVERTEBRATES TaBLE 1. A list of macroinvertebrate taxa from slough sites in the northern Everglades. Class/Order ACARINA AMPHIPODA COLEOPTERA COLEOPTERA COLLEMBOLA COPEPODA DECAPODA Family Crangonyctidae Hyalellidae Chrysomelidae Dryopidae Dytiscidae Gyrinidae Haliplidae Hydrophilidae Noteridae Scirtidae Entomybryidae Argulidae Cambaridae Palaemonidae Genus/Species Several Unidentified Species Crangonyx sp. Hyalella azteca Donacia sp. Pelonomus obscurus Agabetus sp. (larva) Bidessonotus pulicarius Celina slossoni Celina imitatrix Celina spp. (larva) Cybister fimbriolatus (larva) Desmopachria grana Hydroporus sp. Hydrovatus pustulatus compressus Ilybius sp. (larva) Laccophilis gentilis gentilis Gyrinus aneolus Gyrinus elevatus Haliplus havenensis Haliplus mutcherli Haliplus spp. (larva) Peltodytes dietrichi Berosus infuscatus Berosus pugnax Berosus spp. (larva) Chaetarythria sp. (larva) Crenitulus sp. Derallus altus Enochrus consortus Enochrus hamiltoni Enochrus ochraceus Enochrus pygmaeus pygmaeus Enochrus sayi Enochrus spp. (larva) Helobata sp. Helophorus sp. (larva) Hydrobiinae (unidentified adult) Hydrobiomorpha sp. (larva) Hydrochus sp. (larva) Paracymus sp. Tropisternus blatchleyi blatchleyi Tropisternus lateralis nimbatus Tropisternus spp. (larva) Hydrocanthus oblongus Suphis inflatus Suphisellus gibbulus Prionocyphon sp. Entomobrya sp. Several unidentified species Argulus sp. Procambarus alleni Palaemonetes paludosus Pre Gol @) 24 i? D a gg) tag) lag) lac) Ine) tas} lao) Iga) Ine) tao}! Ine) tae) tag Hele H, C/G H, C/G HaAc/E lsh, COLE P H, C/G H, C/G H, C/G H, C/G He ¢/¢€ lal COKE P (Ge CKE ? ? Sh, G 5 A/C/R2 P= (2) 2d (@) ee) leo (@) 22 P= (Qed 22 es eee 2) 22) (@) (@) (@) 22 29 (©) es (Ged 23 eS es) es be) Ps) (@) es) (@) ee eh esl te) es) tec) esl (@) 2s) (=o) tee) [=o] (@) [eel > Ee) Pe! 26 FLORIDA SCIENTIST [VOL 57 TABLE 1 (con't). A list of macroinvertebrate taxa from slough sites in the northern Everglades. Class/Order DIPTERA DIPTERA EPHEMEROPTERA GASTROPODA Family Ceratopogonidae Chironomidae Ephydridae Psychodidae Stratiomyidae Tabanidae Tipulidae Tipulidae Baetidae Caenidae Ancylidae Hydrobiidae Lymnaeidae Genus/Species Bezzia/Palpomyia complex Dasyhelea spp. Forcipomyia sp. Ablabesmyia karelia sp. Ablabesmyia peleensis Ablabesmyia rhamphe group Ablabesmyia sp. Asheum beckae Chironomus sp. Chironomus stigmaterus Cladopelma sp. Cladotanytarsus sp. Dicrotendipes modestus Endochironomus nigricans Goeldichironomus holoprasinus Goeldichironomus natans ? Keifferulus sp. Kiefferulus sp. A Labrundinia neopilosella Larsia decolorata Nimbocera sp. Parachironomus directus Parakiefferiella sp. C Paramerina sp. ‘Polypedilum halterale group Polypedilum sp. Polypedilum sp. A Polypedilum sp. G Polypedilum trigonus Polypedilum tritum Procladius sp. Pseudochironomus sp. Tanypus carinatus Tanytarsus sp. Tanytarsus sp. G Tanytarsus sp. J Tanytarsus sp. R Hydropyrus sp. Pericoma sp. Odontomyia spp. Tabanus sp. Limonia spp. Polymera sp. Tipula sp. Callibaetis floridanus Caenis diminuta Ferrissia sp. Littoridinops monroensis Fossaria cubensis Lymnaea stagnalis Micromenetus dilatatus avus Pseudosuccinae columella F.F.G.1 Ine) tao). lao} Iqel Iqg) Ine) !ne) !ag ere aD anges 2 ’ ’ QO ’ v aspitac| fab) fart ee) fa OOO) Qa e A/C/R2 FODF er OP FADO DAPDOOSFePDDP DOP DAO DAODADADAPPADOOF DAO DDOONAADDAADAONYS SYS No. 1, 2 1994] RADER—MACROINVERTEBRATES a7 TaBLE | (con't). A list of macroinvertebrate taxa from slough sites in the northern Everglades. Class/Order Family Genus/Species F.F.G.1 A/C/R2 Physidae Physella spp. Pilidae Pomacea paludosa Planorbidae Biomphalaria havanensis Drepanotrema sp. Gyraulus parvus Helisoma sp. Planorbella duryi Planorbula armigera wheatleyi Planorbula sp. Snail (?) HEMIPTERA Belostomatidae Belostoma flumineum Belostoma testaceum Belostomatidae spp. (early instars) Lethocerus americanus Corixidae Palmacorixa gillettei Trichocorixa minima Macroveliidae Oravelia sp. ing} tao} laa) Iy9) tas}! Iacfi Ine) tne) Ine no) ns) ©) CO) @) QD) QD) @ QD OQ © A R G R R R A R R C R R G R R C R Mesoveliidae Mesovelia sp. R Naucoridae Pelocoris femoratus balius A HIRUDINEA Erpobdellidae | Mooreobdella sp. C Glossiphoniidae Helobdella sp. R ISOPODA Asellidae Caecidotea spp. C/ExE R LEPIDOPTERA Noctuidae Simyra sp. Sh, H R Pyralidae Acentria sp. H, Sh R Parapoynx sp. H, Sh R ODONATA Aeshnidae Coryphaeschia ingens P R Coenagrionidae Enallagma sp. P C Ischnura sp. P R Telebasis byersi P R Libellulidae Erythemis simplicicollis 12 A Pachydiplax longipennis P R OLIGOCHAETA Lumbriculidae — Eclipidrilus sp. C/G R Naididae Allonais pectinata C/G R Bratislavia unidentata C/G C Dero digitata C/G R Dero furcata C/G R Dero obtusa C/G R Dero pectinata C/G R Dero sp. C/E C Dero trifida C/G R OLIGOCHAETA Naididae Pristina aequiseta C/G R Stylaria lacustris C/G R OSTRACODA Cyprididae Several Unidentified Species (CHE, GE A Physocypria sp. CKELE R Scottia sp. C/G, G R PHYLACTOLAEMA Plumatellidae —_ Plumatella repens F R POLYCHAETA Nereidae Namanereis hawaiiensis C/G R PORIFERA Spongillidae Spongilla lacustris F R TRICHOPTERA Hydroptilidae | Qxyethira sp. H, C/G R Leptoceridae Nectopsyche sp. Sh, H R 1- “F.F.G.” designates the functional feeding group. 2- Abundant, common and rare taxa (A/C/R) are represented by more than 100, between 100 and 20, and less than 20 individuals, respectively. 28 FLORIDA SCIENTIST [VOL 57 and Coleoptera were the most abundant orders (Fig. 2). Eighty-five percent of the Diptera were in the family Chironomidae. However, two genera, Odontomyia spp. (Tipulidae) and Dasyhelea spp. (Ceratopogonidae), also were abundant. Seventy percent of the Chironomidae collected were comprised of four species: Polypedilum trigonus Townes, Goeldichironomus holoprasinus Goeldi, Larsia decolorata Malloch, and Pseudochironomus sp. Adult Hydrocanthus oblongus Sharp and larval stages of Berosus spp., Pelonomus spp., and Tropisternus spp. were the most abundant Coleoptera. Seventy eight percent of the Gastropoda were represented by three taxa (Planorbella duryi Weatherby, Physella spp., and Littoridinops monroensis Frauenfeld). Grazers comprised the largest proportion of the macroinvertebrate assemblage (Fig. 3). However, when all macroinvertebrates that primarily collect and gather detritus were combined (37%) and compared to all individuals that primarily graze algae (33%), collector-gatherer density slightly exceeded grazer density. Caenis diminuta Walker (Ephemeroptera) and three snail taxa (Planorbella duryi, Physella spp., and Littoridinops monroensis) were the most abundant invertebrate grazers. Although numerous species collect and gather detritus, Hyallella azteca, ostracods, and three chironomids (Polypedilum trigonus, Goeldichironomus holoprasinus, and Pseudochironomus sp.) were most abundant. Odonate nymphs (primarily Erythemis simplicicollis Say) and the hemipteran, Pelocoris femoratus balius LaRivers were the most abundant large invertebrate predators. The most abundant small predaceous invertebrates were Dasyhelea sp. (Ceratopogonidae), Larsia decolorata (Chironomidae), and Hydrocanthus oblongus (Coleoptera). Filtering invertebrates Other Amphipoda Gastropoda Ephemeroptera Diptera Fic. 2. Relative abundance (%) of selected taxonomic groups based on density estimates of sweep samples collected from slough habitats. No. 1,2 1994] RADER—MACROINVERTEBRATES 29 Other C/G G, C/G CG, G C/G, H P Fic. 3. Relative abundance (%) of selected functional feeding groups based on density estimates of sweep samples collected from slough habitats. The abbreviations, “C/G”, “G”, “H”, and “P” represent Collector/Gatherer, Grazer, Herbivore, and Predator, respectively. See the text for a description of each functional group. (3%), shredders (1.5%), and herbivores (6%) were represented by a relatively small number of species with low densities. Mean densities were reduced and several taxa were absent from dense sawgrass habitats (Table 2). Almost all taxa were well represented in habitats along the margin of the canal. Amphipods and decapods (almost exclusively Palaemonetes paludosus Gibbes) were extremely abundant. Amphipods were most dense on floating plants, whereas freshwater shrimp were abundant in grass habitats along the shallow margins of the canal (Table 2). Oligochaeta and Ostracoda were not collected from the canal because samples were not taken from the bottom substrate. Discussion—Macroinvertebrate diversity in the Everglades was surprisingly high. Both species richness and Shannon’s diversity index were comparable to well- oxygenated lotic ecosystems (e.g. Rader and Ward, 1988). Although dispersal capabilities and colonization barriers determine the pool of species capable of inhabiting the Everglades, adaptations to temporary, fluctuating conditions deter- mine which species eventually become established members of the Everglades community. 30 FLORIDA SCIENTIST [VOL 57 TABLE 2. Macroinvertebrate mean annual densities (numbers/m®*) are shown for slough habitats in WCA-2A (Marsh). “Sawgrass” and “Canal” habitats are mean densities based on sweep samples collected on a single sampling date. Numbers in parentheses are standard deviations. A dash indicates taxa not collected from the specified habitat. Taxa Marsh Canal Slough Sawgrass Grass Floating Plants Reeds Amphipoda 104.3 68.3 787.5 10768.8 787.5 (241.0) (159.8) (934.2) (7982.8) (1034.3) Chironomidae 304.3 5.0 = Ses 56.3 (725e1) (10:5) (63.4) (106.4) Coleoptera eas 8.8 6.3 193.8 18.8 (209.3) (16.7) (13.5) (208.7) (33.4) Decapoda Oo.d 25.0 2075.0 550.0 725.0 (476.4) (45.3) (1134.6) (432.1) (988.0) Ephemeroptera 84.4 8.8 43.8 = S13 (596.1) (16.7) (53.2) (71.8) Gastropoda 199.7 = 6.3 S183 68.8 (317.2) (11.6) (56.2) (37.4) Hemiptera 27.9 : 62.5 131.3 6.3 (1732) (12.4) (89.9) (10.3) Hirudinea 6.1 E “ D5 6.3 (91.4) (421.6) (76.3) Odonata 214 5:3 50.0 68.8 oles (127.4) (10.6) (ir) (218.4) (324.0) Oligochaeta 3.0 1225 2 = - (196.8) (S17) Ostracoda 18.2 feo: = - E (108.4) (16.9) Because of low flow temporary conditions, the Everglades is a harsh aquatic environment. Although some habitats (e.g. sloughs) typically remain inundated, much of the Everglades landscape dries on a seasonal cycle. The occurrence of a terrestrial phase is an enormous obstacle for many aquatic macroinvertebrates (Williams, 1987). Adaptations by which species survive desiccation during the dry phase vary, but many species have evolved similar mechanisms. For example, mayflies, ostracods, many midges and mosquitoes survive as dormant or diapausing eggs (Lehmkuhl, 1973; Williams and Hynes, 1976); amphipods find permanent pools and crayfish burrows (Williams and Hynes, 1976); while dragonflies, most Nox 1.2 1994] RADER—MACROINVERTEBRATES al beetles and hemipterans survive as terrestrial adults. In addition to desiccation, fluctuating oxygen concentrations (hyperoxia and hypoxia), and high temperatures (metabolic regulation) establish physiological barriers that filter the pool of available colonists. Many macroinvertebrates that evolved in lotic environments and are poorly adapted to low flow, temporary conditions (mayflies, stoneflies, and caddisflies) are either under-represented or absent from the Everglades. Summer temperatures in the Everglades (Gleason, 1974; Rader and Richardson, 1992) can often exceed the upper lethal limit (30 to 40° C) for most macroinvertebrates (Pennak, 1989). Many taxa common in the Everglades tolerate high temperatures by reducing oxygen consumption producing a dormant or quiescent condition and by finding thermal refugia in the mud or in shaded habitats (Ward, 1992; Young and Zimmerman, 1956). Diurnal oxygen concentrations in the Everglades fluctuate between super- saturation and anaerobiosis as a function of photosynthesis and respiration (e.g., Gleason, 1974; Rader and Richardson, 1992). Seasonal oxygen concentrations in macrophyte beds measured near the plant/water interface where most macroinvertebrates reside fluctuated between supersaturation (30 mg/l!) during the day to less than 3 mg/l at night (Rader and Richardson, 1992). Oxygen measurements near the soil/water interface within various marsh habitats and mid-way in the water column within sawgrass habitats were often anoxic and consistently < 3.0 mg/l (Rader and Richardson, 1992). Most macroinvertebrates in the Everglades can utilize atmospheric oxygen. For example, 84% of the snails, all but 2 species (the apple snail and Littoridinops monroensis), are secondarily aquatic, pulmonate snails that have a modified “lung”. Only a few taxa (mayflies, some chironomid larvae, and oligocha- etes) use cutaneous respiration and they rely on respiratory pigments (e.g., hemoglo- bin) and body undulations to supplement cutaneous diffusion. Despite these adaptations, oxygen concentration appears to play an important role in limiting the habitat utilization of macroinvertebrates. Most macroinvertebrates inhabit sloughs which contain abundant growths of algae. Both the diversity and density of macroinvertebrates were lower in dense sawgrass where shading limits algal production resulting in consistently low (< 3.0 mg/L) seasonal and diurnal oxygen concentrations (Rader and Richardson, 1992). Additionally, macroinvertebrates may be reduced in sawgrass because of reduced food resources caused by limited algal abundance. Man-made canals are a dominant feature of the Everglades landscape. Macroinvertebrate densities were surprisingly high in floating vegetation and on grasses and reeds growing along the canal margins. Shrimp and amphipods were especially dense in aquatic grasses and on floating vegetation, respectively. Most of the general taxonomic groups (e.g., Hemiptera, Odonata, and Coleoptera) found in sloughs were present in the canal vegetation. As a prominent water source for newly inundated marshes throughout the Everglades, canals likely represent a source of colonists for many invertebrates and fish. The same adaptations that permit macroinvertebrates to survive and flourish in harsh, temporary environments also produce excellent dispersal capabilities (Will- iams, 1987; Thorp and Covich, 1991). Some species in the Everglades are the most 39 FLORIDA SCIENTIST [VOL 57 common members of their family and are ubiquitous in both the neotropical and nearctic regions of the world |e.g. Hyallella azteca (Amphipoda), Spongilla lacustris (Porifera), Plumatella repens (Bryozoa), and Pseudosuccinae columella (Gastropoda)]. Many species common or abundant in the northern Everglades also occur through- out North America and/or the southeastern United States [Berosus infuscatus (Coleoptera), Enochrus hamiltoni (Coleoptera), Littoridinops monroensis (Gastropoda), Erythemis simplicicollis (Odonata), Palaemonetes paludosus (Decapoda)]. Only three species, Derallus altus (Coleoptera), Biomphalaria havanensis (Gastropoda), and Pomacea paludosa (Gastropoda) also occur in Central and South America (Young, 1954; Thompson, 1984). Isolation of the Florida peninsula likely accounts for the under-representation of Central and South Ameri- can species that must cross oceanic barriers to reach the Everglades. Therefore, the Everglades macroinvertebrate assemblage is characterized by common species from North America, especially the southeastern U.S., with a few colonists from Central and South America. Speciation and the occurrence of endemic species is rare in temporary environments (e.g. MacArthur, 1972). However, two species are endemic to the Florida peninsula, Callibaetis floridanus (Ephemeroptera) and Planorbella duryi (Gastropoda). Planorbella duryi comprised over 70% of the total density of all snails, whereas, Callibaetis floridanus is one of only two mayflies found in the Everglades (Berner, 1950). Energy flow through food webs from primary producers to higher trophic levels can follow both grazer and detritivore pathways. Based on the relative abundance of macroinvertebrates that primarily graze living plants versus those that primarily consume detritus, the grazer and detritivore pathways achieve approximately equal representation in the Everglades. These data suggest that higher trophic levels may be equally dependent upon algal primary production and detritus accumulation. ACKNOWLEDGMENTS—I thank John G. Zahina and Robert R. Johnson for technical assistance. This research was funded bya grant from the Everglades Protection District of Florida to the Duke University Wetland Center. LITERATURE CITED BERNER, L. 1950. The Mayflies of Florida. Univ. Florida Stud. Biol. Sci. Series 4:1-267. CralGHEaD, F. C., Sr. 1968. The role of the alligator in shaping plant communities and maintaining wildlife in the southern Everglades. Florida Natural. 41:2-7. Cut, F., J. A. Davis, J. E. GRowns, J.S. BRADLEY, AND F. H. WuITTLEs. 1993. The influences of sampling method on the classification of wetland macroinvertebrate communities. Hydrobiologia 257:47- 56. EpLer, J. H. 1992. Identification manual for the larval Chironomidae (Diptera) of Florida. Biology Section, Florida Department of Environmental Regulation, Tallahassee, FL 299 pp. GLEASON, P. J. (ed.). 1974. Environments of South Florida: Present and Past. Miami Geological Society, Miami, FL 551 pp. KusHLan, J. A. 1991. The Everglades. Pp. 141-142. In: Meyers, R. L.anp J. J. EWEL (eds.), Rivers of Florida. Springer-Verlag Inc., New York, NY. LEHMKUHL, D. M. 1973. Anewspecies of Baetis from ponds in the Canadian Arctic, with biological notes. Canad. Entomol. 10:343-346. MacarTHuR, R. H. 1972. Geographical ecology. Harper and Row, New York, NY 367 pp. Merritt, R. W. aND K. W. Cummins, (eds). 1984. An introduction to the aquatic insects of North America (2nd ed.). Kendall/Hunt, Dubuque, IA, 722 pp. No. 1, 2 1994] RADER—MACROINVERTEBRATES Bu) Mitcu, W. J. AND J. G. GossELINK. 1986. Wetlands. Van Nostrand Reinhold, New York, NY, 539 pp. PENNAK, R. W. 1989. Freshwater Invertebrates of The United States (3rd ed.). John Wiley & Sons, New York, NY, 628 pp. Raper, R. B. anp C. J. Ricnarpson. 1992. The effects of nutrient enrichment on algae and macroinvertebrates in the Everglades: A review. Wetlands 12:121-135. AND C. J. Ricuarpson. 1994. Response of macroinvertebrates and small fish to nutrient enrichment in the northern Everglades. Wetlands (in press). AND J. V. Warp. 1988. Influence of regulation on environmental conditions and the macroinvertebrate community in the upper Colorado River. Regul. Rivers 2:597-618. Reark, J. B. 1961. Ecological investigations in the Everglades. Pp. 1-19. In: Second Annual Report, Everglades National Park, Homestead, FL. RosENBERG, D. M. AND V. H. Resu. 1993. Freshwater Biomonitoring and Benthic Macroinvertebrates. Chapman and Hall, New York, NY, 488 pp. SHANNON, C. E. AND W. Weaver. 1963. The Mathematical Theory of Communication. Univ. Illinois Press, Urbana, IL. THompson, F. G. 1984. Freshwater snails of Florida. Univ. Florida Press, Gainesville, FL, 94 pp. Tuorp, J. H. anp A. P. Covicn. 1991. Ecology and classification of North American freshwater invertebrates. Academic Press, San Diego, CA, 911 pp. Wakb, J. V. 1992. Aquatic insect ecology. 1. Biology and habitat. John Wiley & Sons, New York, NY, 438 ee D. D. 1987. The Ecology of Temporary Waters. Timber Press, Portland, OR 205 pp. and H. B. N. Hynes. 1976. The ecology of temporary streams I. The faunas of two Canadian streams. Internationales Revue ges. Hydrobiologia 61:761-787. Younc, F. N. 1954. The Water Beetles of Florida. Univ. Florida Press, Gainesville, FL 238 pp. AND J. R. ZIMMERMAN. 1956. Variations in temperature in small aquatic situations. Ecology 37:609-611. Florida Scient. 57(1,2):22-33.1994 Accepted: January 25, 1994. 34 FLORIDA SCIENTIST [VOL 57 Biological Sciences SILVER ACCUMULATION IN THREE SPECIES OF FISH (FAMILY: CENTRARCHIDAE) IN STORMWATER TREATMENT PONDS Kym RousE CAMPBELL! St. Johns River Water Management District, 618 E. South St., Orlando FL 32801 and Department of Biology, University of Central Florida, Orlando FL 32816 Asstract: Redear sunfish (Lepomis microlophus), largemouth bass (Micropterus salmoides), and bluegill sunfish (Lepomis macrochirus) were collected from stormwater ponds and natural lakes and ponds in the Greater Orlando area and analyzed for silver in order to determine: (1) if fish that live in stormwater treatment ponds bioaccumulate significant concentrations of silver and (2) if differences in silver concentrations between species with different foraging strategies occur. Redear sunfish from stormwater ponds (0.458 mg/kg) contained significantly higher (p<0.005) concentrations of silver than fish from control sites (0.001 mg/kg). Largemouth bass and bluegill collected from stormwater ponds (0.419 mg/kg; 0.0278 mg/kg, respectively) contained higher concentrations of silver than those from control sites (0.0418 mg/kg; 0.0014 mg/kg, respectively), but the differences were not statistically significant. No significant (p<0.05) correlations between silver concentration and length and weight of any of the three fish species were found. REGULATIONS to address stormwater were created in Florida in the late 1970s. When land is developed in Central Florida, stormwater management systems must be constructed to treat the stormwater associated with the project. The St. Johns River Water Management District requires that many stormwater ponds be veg- etated for the biological treatment of stormwater. Also, the District commonly approves plans that create habitat for fish and wildlife by planting desirable wetland and aquatic vegetation in the littoral zones of stormwater ponds to mitigate for the loss of wetland habitat as a result of land development. Studies have shown that urban runoff entering stormwater ponds contains a significant amount of heavy metals (Wilber and Hunter, 1979; Owe et al., 1982). Heavy metal sources are largely associated with the operation of motor vehicles, atmospheric fallout, and road surface materials (Harper, 1985). Several studies have suggested that heavy metals accumulate in bottom sediments of stormwater ponds (Wigington et al., 1983; Harper, 1985; Hampson, 1986; Nightingale, 1987). The objective of this research was to determine if the fish that live in stormwater treatment ponds in the Orlando area bioaccumulate significant concentrations of silver because the various species of fish that inhabit stormwater ponds serve as a food source to wildlife, especially wading birds (Campbell, 1993). Most fish are capable of accumulating metals from their diet and directly from water via various membrane surfaces, particularly the gills. Seelye and co-workers ' Current address: Oak Ridge National Laboratory, Environmental Sciences Division, P.O. Box 2008, Oak Ridge TN 37831 No. 1, 2 1994] CAMPBELL—SILVER ACCUMULATION 35 (1982) have shown that sediments contain toxins that may accumulate in fish indirectly through the food web or directly from exposure due to resuspended sediments. Fish exposed to high levels of trace metals in the water can take up substantial quantities of these metals (Atchison et al., 1977). Patrick and Loutit (1978) found that a metal-contaminated diet is a significant source of increased metal levels in fish. The fish tissue concentrations of most metals appear to be more greatly influenced by association with bottom sediments than by the position in the food chain (Wren et al., 1983). Silver is one of the most toxic but least studied of the heavy metals in aquatic ecosystems (Coleman and Cearley, 1974). Although silver has received little envi- ronmental interest, it is a very toxic metal, does occur in industrial discharges, and it must be considered in any classification of highly toxic potential pollutants (Bowen, 1966). Silver is used in jewelry, silverware, ink, electroplating, and photographic processes (Lima et al., 1982). Silver, in minute amounts in water, is very toxic to fish, probably through interference with gas exchange by the gills (Gough et al., 1979). Coleman and Cearley (1974) determined that death from silver toxicity in largemouth bass and bluegill possibly occurred from gill damage and/or central nervous system damage. In addition to the direct toxicity of silver to fish, another and possibly more serious threat exists through the ability of these organisms to concentrate this metal (either directly from the water or indirectly via the fish food chain) (Coleman and Cearley, 1974). In reviewing the literature, very little information was found regarding the accumulation of silver in whole fish tissue. Coleman and Cearley (1974) found that the rate of weight gains of largemouth bass and bluegill sunfish exposed to silver decreased as the concentration increased; bass were more sensitive to silver than bluegill. Bass and bluegill accumulated silver in concentrations greater than those of the exposure water with a subsequent equilibrium developing between the water and tissue concentrations; the quantity of metal accumulated increased as a function of time and concentration. In order to determine if there were differences in silver concentrations in fish with different foraging strategies, three species of the sunfish family, Centrarchidae, with substantially different foraging strategies were selected for this study: redear sunfish, Lepomis microlophus; largemouth bass, Micropterus salmoides; and bluegill sunfish, Lepomis macrochirus. The redear sunfish is a bottom feeder; it depends largely on mollusks for food and does not compete severely with insect-eating fish (McClane, 1978). Redear sunfish in Florida eat midge larvae, snails, scuds, prawns, and mayfly and dragonfly naiads (Wilbur, 1969). Upon finding a snail on the bottom, the redear assumes a vertical position and literally dives into the sediment, head first (Wilbur, 1969). Largemouth bass are considered predators; they eat mainly fish (McClane, 1978). Bass in ponds without forage fish species rely on crayfish, frogs, large insects, and young bass (Carlander, 1969). The food of the bluegill sunfish consists of insects and some vegetation (McClane, 1978); they are considered omnivores. The greater consumption of free-swimming organisms indicates that bluegills are not as oriented to bottom feeding as are the redears (Wilbur, 1969). 36 FLORIDA SCIENTIST [VOL 57 METHODS AND MATERIALS—Stormwater ponds selected for the study met the following criteria: (1) the stormwater pond was of a wet design; (2) it was located in the Orlando area; (3) the project associated with the pond was a shopping center, an apartment complex, or a road; (4) the project was built between 1983 and 1988; (5) seining was possible; and (6) wading birds had been observed feeding. Control sites selected met the following criteria: (1) the pond or lake did not receive any urban or road runoff; (2) it was located within the Greater Orlando area; (3) it was accessible by boat, and (4) permission had been obtained for access. The locations of the stormwater ponds and control sites selected are shown (Fig. 1) and described (Table 1). Fish were collected in December 1991 and January 1992 from stormwater ponds using seines and a gill net. Because all three species of fish did not occur in each stormwater pond, seven ponds were required for the study (Table 1). Five redear were collected from each of three stormwater ponds containing them; five largemouth bass were collected from each of three ponds; and five bluegill were collected ‘icon each of three of the stormwater ponds (Table 1). A total of fifteen individuals of each species was collected from the stormwater ponds. Fish were collected from control sites with an electrofishing boat or a seine in December 1991 and January and March 1992. Fifteen individuals of each species were collected from the control sites: five redear, bass, and bluegill were collected from each of the control sites containing them (Table 1). Upon collection, the fish were measured (cm). Each fish was tagged with a numbered tag and placed in a sterile plastic bag. The fish were placed on ice and later frozen until they were taken to the laboratory for analysis. A composite sediment sample was collected from each of the seven stormwater ponds and four control sites on, or near, the date that the fish were collected. A sediment sample was collected from three different locations in each pond with an Ekman dredge. The top three inches of each of the three samples was combined in a bucket in order to obtain a composite sample. Each composite sediment sample was placed in a sterile plastic bag and kept on ice until it was taken to the laboratory for analysis. The fish (90 individuals total) and the composite sediment samples (11 samples) were taken to Flowers Chemical Laboratories, Inc. (451 Newburyport, Altamonte Springs, Florida) for heavy metal analysis. At the laboratory, each fish was weighed (g). Each whole fish was pureed in a Waring blender. A subsample (0.2:to'0:5 g) of each puree was cused ae the microwave digestion procedure al digested according to EPA Method 3051 (USEPA, 1986). Each fish was analy welt for silver (Ag) using the Neomne Absorption Direct Aspiration Method, EPA Method 7760 (USEPA, 1986). The minimum level of detection was 0.001 mg/kg wet wt. A representative sediment sample (0.5 g) from each site was digested Fic. 1. Location map of stormwater ponds and control sites where fish were collected for silver analysis, December 1991-March 1992, Orlando, Florida. No. 1, 2 1994] CAMPBELL—SILVER ACCUMULATION 37 TaBLE 1. Description of sites numbered on location map (Fig. 1.), December 1991-March 1992, Orlando, Florida. Site Number i | 10 Type of Site Stormwater Pond Stormwater Pond Stormwater Pond Stormwater Pond Stormwater Pond Control Stormwater Pond Stormwater Pond Control Control Control Site Description Apartment Complex, Built in 1987, Approx. 0.5 acre Road, Built in 1986, Approx. 1.5 acres Shopping Center, Built in 1986, Approx. | acre Shopping Center, Built in 1988, Approx. 2 acres Shopping Center, Built in 1986, Approx. | acre Sand Lake, located in Wekiva Springs State Park, Approx. 3 acres Road, Built in 1986, Approx. | acre Apartment Complex, Built in 1988, Approx. 0.75 acre Buck Lake, located near Geneva, Approx. 30 acres 0.5 acre pond on E.H. Kilbee Ranch, near St. Johns River Lake located near Chuluota, Approx. 10 acres Species Collected Redear Bass and bluegill Bass and bluegill Redear Redear Redear, bass, and bluegill Bass Bluegill Redear, bass, and bluegill Redear Bass and bluegill and analyzed as described above. All concentrations were determined on a wet weight basis; units of measurement used were mg/kg wet wt. Analysis of Variance (ANOVA) was used to determine differences in silver concentrations (SAS Institute, 1985). The Bonferroni Multiple Comparisons Procedure was used to compare mean concen- trations. The level of significance was set at p<0.005. The Wilkes-Shapiro Test was used to test for normality. The Dixon and Massey Degrees of Freedom Adjustment for Unequal Variances was used to determine if the equal variance assumption was met (Zar, 1984). A correlation analysis was used to determine the relationship between length and weight of the fish and the silver concentration. ResuLts—The mean length and weight of fish collected are listed in Table 2. The mean length and weight of the redear and bass from the stormwater ponds and 38 FLORIDA SCIENTIST [VOL 57 TABLE 2. The mean length (cm) and weight (g) of fish collected, December 1991-March 1992, Orlando, Florida. Species Mean Length Range Mean Weight Range Stormwater Ponds Redear sunfish 15 Ae ala) Sor 5.9 110.0 Largemouth bass 14.6 7.8 2598 55.6 3.8 193.2 Bluegill sunfish 9.3 5.6 126 14.9 3:6. 8915 Control Sites Redear sunfish kOe 9:1. 15.0 27.8 8.9 49.1 Largemouth bass 20.4 PES =30'3 116.7 12.5 323.8 Bluegill sunfish 14.6 92 20.6 57.8 14.7 154.7 control sites were not significantly different (p>0.005). However, the differences were significant (p<0.005) between the mean length and weight of bluegill collected from stormwater ponds and controls. Redear sunfish from the stormwater ponds contained the highest mean silver concentration of any fish collected (Fig. 2, Table 3). Silver concentrations in redear sunfish from stormwater ponds were significantly higher (p<0.005) than from control sites. No significant differences between silver concentrations of the bass and bluegill from stormwater ponds and control sites were observed (Fig. 2, Table 3). Largemouth bass from stormwater ponds did not contain statistically elevated silver concentrations; however, the actual p value (0.0057) indicated that the difference was very close to being significant and, therefore, considered biologically meaningful (Table 3). Bluegill from stormwater ponds also contained a higher mean concentra- tion than control site fish which also may be biologically significant (p=0.1905) (Table 3). Composite sediment samples collected from stormwater ponds contained a similar silver concentration to the samples from control sites (Table 3). No significant correlation between the silver concentrations and the length and the weight of the three species of fish was observed (Table 4). Results of the Wilkes-Shapiro Test showed that the data did not deviate significantly from normality. The Dixon and Massey Degrees of Freedom Adjust- ment for Unequal Variances test determined that the variances were equal enough to meet the assumption of equal variances. Discuss1oN—Silver accumulation in fish from the stormwater ponds was consis- tent with many studies on other metals (Sorensen, 1991); the fish species most associated with the bottom sediments (redear, in this case) contained the highest metal concentration of the fish sampled. The results of this study indicated that a No. 1, 2 1994] 0.5 MEAN AG CONCENTRATION (MG/KG) CAMPBELL—SILVER ACCUMULATION 39 — BLUEGILL fii] STORMWATER PONDS [_] CONTROLS Fic. 2. The mean silver concentrations of redear sunfish, largemouth bass, and bluegill sunfish collected from the stormwater ponds and the control sites, December 1991-March 1992, Orlando, Florida. TABLE 3. Silver concentrations (mg/kg wet wt.) of fish and composite sediment collected, December 1991-March 1992, Orlando, Florida. Species N Mean Ag SoD), Range Concentration Stormwater Ponds Redear sunfish 15) 0.458 0.413 <0.005, 1.31 Largemouth bass 15 0.419 0.462 <0.005, 1.15 Bluegill sunfish 15 0.0278 0.0763 <0.001, 0.297 Composite sediment 7 Omi 0.116 <0.001, 0.326 Control Sites Redear sunfish 15 0.001° N/A N/A Largemouth bass 15 0.0418 0.158 <0.001, 0.613 Bluegill sunfish 15 0.0014 0.0016 <().001, 0.0070 Composite sediment 4 0.337 0.233 0.0904, 0.604 * all values below limits of detection 40 FLORIDA SCIENTIST [VOL 57 TABLE 4. Correlation coefficients of silver concentration vs. length and weight of fish. All correlations are non-significant (p>0.05). Species Length Weight Redear sunfish 0.1339 0.0919 Largemouth bass 0.1225 0.3657 Bluegill sunfish 0.0094 -0.0356 metal-contaminated diet or association with the bottom sediment appeared to be a more significant source of silver accumulation in fish than direct exposure from the water. Only one value of silver concentration in whole fish tissue was available in the literature with which to compare the results. In a summary of 1975-1979 data on whole fish tissue from various species from EPA’s STORET database, the mean concentration of silver in 211 samples was 0.225 mg/kg wet wt., with a range of 0.004- 1.900 mg/kg (Scow et al., 1981). These values are lower than those obtained for redear and bass from stormwater ponds. However, because of the paucity of data available in the literature, it is not known if the average silver concentrations in fish from stormwater ponds represented high levels. The lack of a significant correlation between the silver concentration and the length and the weight of redear, bass, and bluegill was consistent with the literature regarding other heavy metals (Sorensen, 1991). Liebscher and Smith (1968) have suggested that the lack of a correlation between non-essential elements (such as silver) and fish weight might be due to the fact that non-essential elements are merely contaminations of tissue and have no significant function. Giesy and Wiener (1977) suggest using caution in the analysis of data on metal concentrations in fish because metal concentrations in fish often are highly variable. Consequently, the data often are non-normally distributed, and the assumptions for parametric statistics are not met. For this reason, the data in this study were analyzed to determine if the assumptions of normal distribution and equal variances were met. There was variability in the data; however, the data did not deviate significantly from the assumptions, and using parametric methods to analyze the data was acceptable. ConcLusions—The effect on the wading birds and other wildlife that are feeding on the fish living in stormwater ponds is unknown and was beyond the scope of this study. Research is needed to determine how often predators forage in stormwater ponds and the effects of predators eating fish containing silver. Studies also should be done on the silver concentrations present in the other faunal components of stormwater ponds. Stormwater ponds help keep the lakes and rivers of Florida from becoming polluted from urban runoff; however, the fish that live in these ponds represent a potentially contaminated food source for both wildlife and humans. Now 1,2 1994 | CAMPBELL—SILVER ACCUMULATION 4] ACKNOWLEDGMENTS—I would like to thank the St. Johns River Water Management District for funding this project. Thanks to the many people who work for the District for their help; the Florida Game and Freshwater Fish Commission for lending me their seines and taking me out on their electrofishing boats; and my graduate committee for all their help with this project: Lorrie Hoffman, Buck Snelson, Jack Stout, and David Vickers. LITERATURE CITED Atcuison, G. J., B. R. Murpuy, W. E. Bisnop, A. W. McINTosH, AND R. A. Mayes. 1977. Trace metal contamination of bluegill (Lepomis macrochirus) from two Indiana lakes. Trans. Am. Fish. Soc. 106(6):637-640. Bowen, H. J. M. 1966. Trace Elements in Biochemistry. Academic Press, New York. 241 pp. CampBELL, K. R. 1993. Bioaccumulation of heavy metals in fish living in stormwater treatment ponds. Masters Thesis, University of Central Florida, Orlando, Florida. 123 pp. CarLANDER, K. D. 1969. Handbook of Freshwater Fishery Biology, Volume 2. The Iowa State University Press, Ames, Iowa. 431 pp. CoLeMaN, R. L. AND J. E. Cear ey. 1974. Silver toxicity and accumulation in largemouth bass and bluegill. Bull. Environm. Contam. Toxicol. 12(1):53-61. Giesy, JR., J. P. AND J. G. WIENER. 1977. Frequency distributions of trace metal concentrations in five freshwater fishes. Trans. Am. Fish. Soc. 106(4):393-403. Goucu, L. P., H. T. SHACKLETTE, AND A. A. Case. 1979. Element concentrations toxic to plants, animals, and man. Geological Survey Bulletin 1466, U.S. Govt. Printing Office, Washington, D.C. 80 pp. Hampson, P. S. 1986. Effects of detention on water quality of two stormwater detention ponds receiving highway surface runoff in Jacksonville, Florida. U.S. Geol. Sur., Water-Resources Investigations Report 86-4151. pp 44-50. Harper, H. H. 1985. Fate of heavy metals from runoff in stormwater management systems. Ph.D. Dissert., University of Central Florida, Orlando, Florida. 389 pp. LIEBSCHER, K. AND H. Situ. 1968. Essential and non-essential trace elements. Arch. Environ. Health. 17:881-890. Lia, A. R., C. Curtis, D. E. HAMMERMEISTER, D. J. CALL, AND T. A. FELHABER. 1982. Acute toxicity of silver to selected fish and invertebrates. Bull. Environm. Contam. Toxicol. 29:184-189. McC1ang, A. J. 1978. McClane’s Field Guide to Freshwater Fishes of North America. Henry Holt and Company, New York. 212 p. NIGHTINGALE, H. I. 1987. Accumulation of As, Ni, Cu, and Pb in retention and recharge basins soil from urban runoff. Wat. Res. Bull. 23(4):663-672. Owe, M., P. J. CRAULAND H. G. HaLverson. 1982. Contaminant levels in precipitation and urban surface runoff. Wat. Res. Bull. 18(5):863-868. PaTRICK, F. M. aND M. W. Loutir. 1978. Passage of metals to freshwater fish from their food. Wat. Res. 12:395-398. SAS InstituTE INc. 1985. SAS User’s Guide: Statistics, Version 5 Edition. Cary, NC: SAS Institute Inc. 956 pp. Scow, K., M. Goyer, AND L. NELKEN. 1981. Exposure and risk assessment for silver. Report to the USEPA, Office of Regulations and Standards, Washington, D.C. PB85-211993. SEELYE, J. G., R. J. HESSELBERG, AND M. J. Mac. 1982. Accumulation by fish of contaminants released from dredged sediments. Environ. Sci. Technol. 16:459-464. SORENSEN, E. M. B. 1991. Metal Poisoning in Fish. CRC Press, Boca Raton, FL, 374 pp. USEPA (United States Environmental Protection Agency). 1986. Test methods for evaluating solid waste, Volumes 1A and 1B. SWA-846, Office of Solid Waste and Emergency Response, Washington, D.C. WHALEN, P. J. AND M. G. CuLium. 1988. An assessment of urban land use/stormwater runoff quality relationships and treatment efficiencies of selected stormwater management systems. SFWMD Tech. Pub. 88-9. 55 pp. WIGcINGTON, JR., P. J., C. W. RANDALL, AND T. J. Grizzarp. 1983. Accumulation of selected trace metals in soils of urban runoff detention basins. Wat. Res. Bull. 19(5):709-718. Wixsur, R.L. 1969. The Redear Sunfish in Florida. FGFWFC. Fishery Bull. No. 5. Dingell-Johnson F- 22. 64 p. Wizsur, W.G. anp J. V. Hunter. 1979. Distribution of metals in street sweepings, stormwater solids, and urban aquatic sediments. Res. J. Wat. Poll. Contr. Fed. 51:2810-2822. Wren, C. D., H. R. MaccrimMMon, AND B. R. LoescuEr. 1983. Examination of bioaccumulation and 49 FLORIDA SCIENTIST [VOL 57 biomagnification of metals in a precambrian shield lake. Wat. Air Soil Pollut. 19:277-291. Zak, J.H. 1984. Biostatistical Analysis, 2nd ed. Prentice-Hall, Inc., Englewood Cliffs, NJ. 718 p. Florida Scient. 57(1,2):34-42.1994 Accepted: January 7, 1994. REVIEW Alan Holden, The Nature of Solids, Dover Publications, Mineola, NY, 1992, Pp. v + 241. Price: $6.95. Tus book is a republication of the text by Alan Holden on the nature of solids, which was first published in 1965 by the Columbia Univ. Press. This represents the continuation of the program by Dover Publications to make available certain classic texts on science, particularly ones such as this, in which the author utilizes a non- mathematical approach to explain the contemporary theories and understand the models which solid state physicists have developed to account for the observed physical, electrical and magnetic properties of solids. As indicated in the Foreword by Dr. Holden, the study of solids “is commonly divided between the specialities of physicists, chemists and crystallographers”. The fifteen chapters of the text divide off to provide the reader in the early chapters with a review of the fundamental concepts and language characteristic of each of these three specialities and then goes on to apply the combined knowledge to the understanding of some of the larger issues in solids such as atomic motion (Chapt. IX dealing with lattice vibrations, ionic conduction and heat conduction) and electronic motion (Chapt. X - XIV dealing with electrical conduction in metals and semiconductors). This reviewer, as a chemist with strong interest in solid state chemistry, found the book extremely useful for the insights it provided into many of the concepts which solid-state physicists find so useful in describing solids, but which are not commonly covered in the chemistry curriculum. For example Dr. Holden takes the mystery out of Bloch functions for electrons in metals when he illustrates these as simply the combination of orbitals (e.g. 1s, 2s, 2p etc.) for isolated atoms with the orbitals (wave functions) deduced for a particle in a box(see pp. 172 - 177). The important concept of the p-n junction and its application in the rectifier is clearly explained using actual illustrations of the chemical substitution into the silicon crystal structure supplemented with schematic illustrations of the charge distribu- tion in the junction region(see pp. 208 - 215). Necessarily, in view of the original publication date, some of the illustrations in the text and suggested home experiments are dated. One example is the use of the regular grooves of a phonograph record(see Chapt. IV, Fig.1) to diffract light as an analogy to x-ray diffraction from solids. Nevertheless, imaginative readers can easily See REVIEW, page 47 No. 1, 2 1994] YOUNT—COMPUTER MODELING 43 Science Teaching THE USE OF COMPUTER-ASSISTED MOLECULAR MODELING IN COLLEGE GENERAL CHEMISTRY JAMEs R. YOUNT Brevard Community College, 1311 North U.S. Highway 1, Titusville, Florida 32796 Apstract: Students enrolled in General Chemistry II at Brevard Community College were given assignments requiring the use of a computer and molecular modeling software in an effort to improve their understanding of molecular geometry. These assignments were given as part of a unit on organic chemistry and involved (a) learning the computer's operation, (b) becoming familiar with the basics of molecular modeling, and (c) constructing organic molecules such as amino acids, isomeric forms of hydrocarbons, and steroids. Verbal feedback, evaluations, and classroom observations showed that college freshman students were not only able to master the software and perform the assignments, but showed increased understanding of molecular geometry and heightened enthusiasm for both chemistry and computers. TuE visualization of molecules in three dimensions has always been one of the more awkward tasks facing those beginning the study of chemistry. When presented with molecular geometry, discussions of bond angles and lengths and other aspects of structure, students are expected to develop and understand concepts that are highly abstract, that is, to “visualize the unseeable” (Cohen and Denisovich, 1991). Teaching aids in these areas have traditionally been “ball and stick” models, straws, cardboard cut-outs, gumdrops, or any miscellaneous bric-a-brac that could be employed to depict these elusive particles. One of the answers to this educational dilemma may lie in the computer molecular modeling packages that are currently available for use on the microcomputer, capable of allowing “visualization of structures in three-dimensional space [as well as] techniques that represent nonstructural features and properties” (Mohamadi et al., 1990). Computer-assisted molecular modeling tools are widely utilized in professional laboratories and in university research. Programs in a variety of forms are being used for research on coal structures (Haggin, 1989), pharmaceuticals (Erikson, 1990), drug design (Vinter et al., 1987), and chemical databases (Borman, 1989) as well as for puzzling out structures of enzymes and other macromolecules (Cohen and Denisovich, 1991). Computers often can (a) provide information that is not available from experiment, (b) allow a deeper understanding of chemical problems, and (c) be more economical than complementary experimental studies (Wilson, 1986). The future of computer-aided visualization is becoming quite clear: each issue of Protein Science is accompanied by a diskette containing 3D molecular models (Borman, 1992). Unfortunately, most undergraduate programs lack instruction in computer molecular modeling and computation chemistry methods (Lipkowitz, 1989). This 4A FLORIDA SCIENTIST [VOL 57 omission is especially lamentable since exercises in this field are excellent educa- tional opportunities to encourage practical learning of difficult concepts and tend to strongly stimulate student interest. The successful use of computerized molecular modeling in undergraduate education has been reported by several authors. These include exercises on the binding of metals to ligands for the inorganic laboratory (Canales et al., 1992), stereochemistry of organic molecules (Jarret and Sin, 1990), and correlation of odors with molecular structure (Lipkowitz, 1989). The ongoing project described here was initially intended to assess the ability of second semester college general chemistry students to learn a molecular modeling program and utilize it to produce correctly drawn organic molecules. With the success of the pilot project, the exercises were expanded to include more complex molecules and some reaction chemistry. Important factors governing success in the project were (a) time required to learn to operate the computer and the modeling software, (b) accuracy of the molecules drawn, and (c) degree of student interest and understanding resulting from the experience. MATERIALS AND METHODS—The molecular modeling software used was Chem 3D Plus™ from Cambridge Scientific Computing, Inc. This application allows students not only to draw and rotate molecules, but to analyze the structural geometry (i.e., bond length, bond angle, and others). A set of substructures is provided for quicker model building. Models can be minimized for closures and structural error, as well as for strain energy using MM2 field calculations. All atom types are available for use (unlike some modeling programs where only a selection of atoms are available), and nontypical valences may be used. All features are operated by a mouse using pull-down menus. One of the handiest features of the application is that it can be asked to fill all open valences on drawn atoms; for example, drawing a carbon atom results in a model of methane, correctly drawn to proper bond angles. The production of larger molecules then simply involves the replacing of a hydrogen atom with another carbon atom, after which ethane is displayed, and so on. (For a more complete review of Chem 3D Plus™, see Bays, 1992.) A Macintosh LC computer with a math coprocessor was installed in the chemistry lab so that student access would be unrestricted during normal lab operating hours. (The coprocessor is necessary for some of the energy minimization calculations.) Ease of use and minimal training time were the primary reasons for selecting the Macintosh, and, while the software will run on black and white units, the color screen of the LC provided the best visualization at a nominal price. A grant from the BCC Staff and Program Development office was used to purchase both the software and the computer supplies. In order for the students to have some guide both to the assignments and to the workings of the computer and software, a manual was prepared and distributed. In the pilot project during the Spring of 1992, only three sections were included: (1) Using the Macintosh, (2) Chem 3D Plus™ Tutorial, and (3) Amino Acids. The first two sections directed the students through interactive tutorials of the Macintosh (provided with the computer) and the Chem 3D Plus™ software, which came from the manufacturer. The software tutorial included such topics as drawing and selecting atoms, bonding, use of substructures, rotation, saving, and energy minimization. Students generally required about one to one and a half hours to complete these assignments. The third section of the manual contained the modeling assignment. Students constructed one amino acid and one hydrocarbon (e.g., normal or branched pentanes, cyclic aliphatics, and benzene compounds, etc.) and performed an energy minimization. The instructor then verified the correctness and completeness of the structures and assigned grades. In addition, students filled out evaluations of the program. During the Spring of 1993, additional sections were added to the assignment. Each assignment required a weekly written report as well as molecular drawings saved on disc. The revised manual contains the following sections: (a) Introduction to Modeling; (b) Computer Usage Rules; (c) Using the Macintosh; (d) Organic Structure and Nomenclature; (e) Using Chem 3D Plus™ Modeling Software; (f) Basic Organic Molecules; (g) Functional Groups I, including alcohols, halogens, and acids; (h) Functional Groups I, including aldehydes, ketones, ethers, and amines; (i) Molecules Associated with Odors (partially derived from Lipkowitz, 1989, and Sbrollini, 1987)); (j) Amino Acids; and (k) Steroids, No. 1, 2 1994] YOUNT—COMPUTER MODELING A5 including cholesterol and hormones. Alookat the section Basic Organic Molecules will serve to illustrate the nature of the exercises. The student is first required to draw methane and get a report of the bond angles and lengths. Comparisons are then made with measurements of ethane and propane; answers are written on a prepared form. Next, as a study in isomers and branching, students draw models of all isomers of butane, pentane, and butene and discuss any differences in bond angle and bond length, especially concerning double bonds versus single bonds. The assignment concludes with a study of bond angles of cyclohexane versus those of benzene. For grading, the student must hand in the written assignment and (electronically) models of neopentane, butene, methylbenzene, neopentylbenzene, and n-pentylbenzene. Other sections of the manual follow similar formats based on increasingly complex molecules, with emphasis on structure, reactions/bonding, and isomerization. RESULTS AND Discussion—Nine students enrolled in General Chemistry II during the Spring of 1992 and 12 in the Spring of 1993 participated in the project. By observation and grading of their completed molecules, it was determined that each student had achieved the basic objectives of the project: (a) learn the operation of the Macintosh computer and (b) properly operate molecular modeling software to create accurate representations of small organic molecules. These objectives were accomplished in a surprisingly small amount of time: the total mean time spent by a single student on the computer, for all assignments, was about 5 hours. It was expected that this amount of time would be needed just to learn the computer and the modeling program. In some cases, it was apparent that even LESS time may have been needed, as the students seemed in no hurry to get off the machine. Students observed working at the computer appeared to be enjoying themselves. (One student was irritated: he had asked for help from another student, who had obligingly done his entire assignment for him. The first student went back to the computer and REDID it since he felt he had missed the fun!) Several students returned after the assignment was completed wanting to use the program, some just for fun and others as an aid in their organic chemistry course work. | Student evaluations appear to support the positive impression made by the project (Table 1). The overall attitude that the computer and software were easy to learn and the manual was informative provides ample evidence that the project was successful in capturing student interest , especially in light of a few of the students’ written comments: e “The project was a great learning experience. It gave me a better understand- ing of molecular geometries.” e “The computer made it easier to visualize the actual structure of the molecules. It was also enjoyable and more interesting than learning from the book.” ¢ “Before participating in this project, I was intimidated by organic chemistry. After using [the software] I feel more comfortable about it.” e “All my time spent on the computer was informative. [The project was] very useful for completion of future chemistry courses.” A potential problem with electronic homework was the possibility of students’ getting into each other's files. The encryption program FileGuard™ (ASD Software, Inc., see Scheffer, 1992) was employed and has so far been effective in preventing such tampering. Once the student finishes the assignments, the files are “deposited” AG FLORIDA SCIENTIST [VOL 57 TABLE 1. Summative evaluation results from students compiled for two years of the project (n = 14). Aspects of Project Easy to learn/good Adequate Hard to learn/poor Macintosh LC computer Ease of learning use 12 2 0 Ease of actual use fe) 5 0 Suitability for project 13 1 0 Chem 3D+ modeling software Ease of learning use 8 6 0 Ease of actual use 10 4 0 Suitability for project 13 1 0 Instruction manual Content glk 3 0 Ease of use fe) 5 0 Suitability for project 10 4 0 in a special “drop box” folder that only the instructor can enter. FileGuard™ allows either folders or files themselves to be protected from prying eyes, while the instructor (“administrator” in the program’s language) has total access to them. In sum, college freshman chemistry students were able to learn and utilize molecular modeling software in a short period of time and with minimum instruc- tion. As a result of this, they were able to better comprehend the subtleties of molecular geometry and the use of computers in the chemical laboratory. The interest and enthusiasm that the project sparked was in itself reason to further explore the use of computers in the laboratory as a learning tool and as a retention strategy for the physical sciences. ACKNOWLEDGMENTS—This work was supported by a grant from the Brevard Community College Staff and Program Development Office. Thanks are extended to Dr. Rosemary Layne and Dr. Joe Lee Smith for their support. Special thanks are offered to Dr. Mary Crume for editing the manuscript. No. 1, 2 1994] YOUNT—COMPUTER MODELING AT LITERATURE CITED Bays, J. P. 1992. So you want to do molecular modeling? J. Chem. Educ. 69(3): 209-215. Borman, S. 1989. Software adds new dimension to structure searching. Chem. Eng. News 67(28): 28- oo. . 1992. “Kinemages” bring journal illustration into the computer age. Chem. Eng. News 70(7): 26-27. CAMBRIDGE SCIENTIFIC ComPuTING. 1990. Chem 3D: The Molecular Modeling System [Computer program manual]. Cambridge, MA. Cana .es, C., L. EGAN, AND M. ZIMMER. 1992. Molecular modeling as an inorganic chemistry exercise. J. Chem. Educ. 69(1): 21-22. Couen, P. anpD M. Denisovicu. 1991. Molecular modeling. SunWorld 4(9): 58-68. Erikson, D. 1990. Rational drugs. Scient. Amer. 262(1): 102-104. Haccin, J. 1989. Research on coal structures forges ahead. Chem. Eng. News 67(40): 27-28. Jarret, R. M. ANDN. Sin. 1990. Molecular mechanics as an organic chemistry exercise. J. Chem. Educ. 672) 53-150: ; Lipxowitz, K. B. 1989. Molecular modeling in organic chemistry: Correlating odors with molecular structure. J. Chem. Educ. 66(4): 275-277. MouamaDI, F., N. G. J. Ricuarps, W. C. Guia, R. Liskamp, M. Lipron, C. CauFIELD, G. CHANG, T. HENDRICKSON, AND W. C. Stitt. 1990. MacroModel — An integrated software system for modeling organic and bioorganic molecules using molecular mechanics. J. Comput. Chem. 11(4): 440-467. SCHEFFER, M. 1992. FileGuard [Computer program manual]. ASD Software, Inc., Montclair, CA. SBROLLINI, M. C. 1987. Olfactory delights. J. Chem. Educ. 64(9): 799-801. VINTER, J. G., A. Davis, AND M. R. SauNDERS. 1987. Strategic approaches to drug design: An integrated software framework for molecular modeling. J. Computer-Aided Molec. Design 1: 31-51. Witson, S. 1986. Chemistry by Computer: An Overview of the Applications of Computers in Chemistry. Plenum, New York, NY. Florida Scient. 57(1,2):42-47. 1994 Accepted: January 28, 1994. REVIEW (CON’T) replace these examples with ones from currently available technology and with which current students are more familiar, such as the bright, holographically imprinted side of a compact disk (CD) and that gives even better diffraction of light than the phonograph record. This text should be available in all libraries for use by both undergraduate and graduate students in particular to supplement their texts in the area of the study of solids. Instructors in Chemistry, Physics and Crystallography who must present material dealing with solids should consider owning the text as a resource for preparing lectures based on the insights it provides into the fundamental concepts and to better appreciate how the separate ways of speech of these disciples are unified in this author’s presentation. —Joseph A. Stanko, Dept. of Chemistry, Univ. of South Florida, Tampa, FL. 48 FLORIDA SCIENTIST [VOL 57 Biological Sciences A MASS STRANDING OF LEACH’S STORM-PETREL IN GEORGIA AND FLORIDA CarOoL A. RUCKDESCHEL, C. ROBERT SHOOP AND GEORGE W. SCIPLE Cumberland Island Museum, P.O. Box 796, St. Marys, GA 31558 Asstract: This is the first report of a mass stranding of Leach’s Storm-Petrel, Oceanodroma leucorhoa, along the southeast United States coast and the first onshore record for the species in Georgia. Seventy-two birds were found on the beaches of Cumberland Island and Little Cumberland Island, Georgia during a 12-day period (22 May-2 June 1991). Strandings were documented between St. Catherines Island, Georgia and St. Augustine, Florida. Leacu’s Storm-Petrel, Oceanodroma leucorhoa, is a pelagic migrant between its breeding grounds along the coast of eastern North America from Greenland south to Maine and its wintering habitat to the equator (Bent, 1922). The species has been reported from all southeastern coastal states except Georgia and Mississippi, and most observations have been of live animals with <6 birds per sighting (Clapp et al., 1982). This is the first report of the species onshore in Georgia and the first report of mass mortality along the southeastern United States coast. During a 12-day period (22 May-2 June 1991), a total of 72 Leach’s Storm- Petrels was found dead or dying on the beaches of Cumberland Island (65) and Little Cumberland Island (7), Camden County, Georgia. Several birds were observed alive, performing characteristic foot pattering on the water just outside the breaking surf, and live birds were seen standing on the low ocean beach. Cumberland Island was apparently the center of the stranding event which covered approximately 216 km of the Georgia-Florida coast between St. Catherines Island, Liberty County, Georgia (30° 40' N) and St. Augustine, St. Johns County, Florida (29° 50" N). Five dead petrels were recorded by B. Winn (1991) on St. Catherines Island, Georgia; one live bird was seen on Jekyll Island, Glynn County, Georgia, but no strandings were recorded on Jekyll Island (Stewart, 1991). Ten specimens from Nassau and Duval counties, Florida were deposited in the Florida State Museum of Natural History (Robertson and Woolfenden, 1992). Belcher (1991) reported 20+ observations on Amelia Island, Nassau County, Florida and, along with Lowenstein-Whaley (1992), verified the southern limit of the sightings. Negative reports of the species north of St. Catherines Island were provided by Coolidge (1991) and Ross (1992). Leach’s Storm-Petrel is listed by Haney and co-workers (1986) as a rare to uncommon offshore transient during the spring (3 May-20 June) and fall (1 August). Mass strandings appear to coincide with migration periods but have been reported only from the Canadian maritime provinces and Western Europe following gales and storms (Palmer, 1962). During the week before and after the first Storm-Petrel No. 1, 2 1994] RUCKDESCHEL ET AL.—_LEACH’S STORM PETREL 49 stranding in Georgia, peak wind gusts were all from the east or southeast while “resultant wind direction” varied between northeast and southwest at the municipal airport in Savannah, Georgia (National Climatic Data Center). Average wind speed ranged between 11.1 and 4.4 mph with a weekly average only 7.7 and 8.7 mph, respectively. We determined the sex of 29 individuals (17 2 2, 12 64) and examined the crop/gizzard of 48 birds. The crop/gizzard was empty in three animals and 45 contained some material. The most common item represented was eye lenses found in 19 specimens, but no distinction was made between the lenses of fish or squid. A small teleost fish was found in one specimen and squid beaks occurred in five. Other items included parts of crustacea (8), seeds or algal floats (7), polychaete worms (3), mollusca (cf. cassidid larva) (2), algae (2), gall wasp (Hymenoptera; Cynipidae) (1), and dipteran larvae (1). One cestode and three nematodes were found. Pieces of feather were in four birds. No oil was noted in or on the specimens. Cause of the strandings is unknown. Crop/gizzard contents revealed some of the animals had been feeding on a variety of items prior to death, and generally the specimens did not appear emaciated, although a moribund individual collected at Little Cumberland island appeared emaciated when prepared as a museum speci- men. Much sargassum washed ashore concomitantly with the strandings. Specimens were deposited in the U. S. National Museum (USNM 608564), the Field Museum of Natural History, and the Univ. of Georgia Museum of Zoology. All other specimens (skins and formalin-fixed) are in the Cumberland Island Museum. ACKNOWLEDGMENTS—Identification of crop/gizzard contents was accomplished with the help of several individuals. Dr. Robert Bullock (Univ. of Rhode Island) generously confirmed identification of the invertebrate material. Dr. Cecil Smith (Univ. of Georgia) supplied information on the insects, and Dr. R. D. Suttkus (Tulane Univ.) confirmed fish identification. Rebecca Bell counted and collected stranded birds on Little Cumberland Island. Specimens were collected under Georgia permit 29-No.000027. LITERATURE CITED BE.Lcuer, H. 1991. 109 S.18th St., Fernandina, FL 32034, Pers. Commun. Bent, A. C. 1922. Life histories of North American Petrels and Pelicans and their allies. Bull. U.S. Natl. Mus. No. 121. xii and 343 pp. Capp, R. B., R. C. Banks, D. Morcan-Jacoss, AND W. A. HorrMan. 1982. Marine birds of the southeastern United States and Gulf of Mexico. Part I. Gaviiformes through Pelecaniformes. U.S. Fish and Wildlife Service, Office of Biological Services, Washington, D. C. FWS/OBS-82/01, 637 pp. Coo.wwcE, H. 1991. 14 Liberty Cr. Dr., Savannah, GA 31406. Pers. Commun. Haney, J. C., P. Brisse, D. R. Jacobson, M. W. Oberle, and J. M. Paget. 1986. Annotated checklist of Georgia birds. Occasional Pub. No. 10, GA Ornithol. Soc. 51 pp. LOWENSTEIN-WHALEY, J. 1992. Marineland, 9507 Ocean Shore Blvd., Marineland, FL32086. Pers. Commun. PauM_r, R. S. 1962. Handbook of North American birds. Vol. 1, Yale Univ. Press, New Haven CT. ROBERTSON, W. B. JR., AND G. E. WOOLFENDEN. 1992. Florida Bird Species, An Annotated List. Special Publ. No. 6, FL Omithol. Soc., Gainesville, FL. Ross, D. 1992. Wassaw Island, c/o Isle of Hope Marina, 50 Bluff Dr.,Savannah, GA 31406. Pers. Commun. Stewart, D. 1991. 4H Center, 201 South Beachview Dr., Jekyll Island, GA 31527. Pers. Commun. Winn, B. 1991. Wildlife Survival Center, N.Y. Zool. Soc., St. Catherines Island, Midway, GA 31320. Pers. Commun. Florida Scient. 57(1,2):48-49.1994 Accepted: February 4, 1994. 50 FLORIDA SCIENTIST [VOL 57 Biological Sciences YEAST INTERACTIONS INFERRED FROM NATURAL DISTRIBUTION PATTERNS Puitip F. GANTER’, EDUARDO BuSTILLO®’, AND JENNIFER PENDOLA®’ “Tennessee State University, Biology Department, 3500 John Merritt Blvd., Nashville, TN 37209- 1561, USA; Barry University, School of Natural and Health Sciences, Miami Shores, FL 33161, USA Apstract: Cactophilic yeast were sampled from decaying stems and fruit of two Opuntia species at four locations in Florida. Species composition of the community differed between stems and fruits with very few species occurring in both habitats, although fruits and stems were interspersed. Stem communities were more homogeneous than fruit communities. The yeast communitites from stems of different species of cacti were dominated by different species of yeast. Analysis of the pH of decaying tissue supported the hypothesis that the number of yeast species per stem increased through time. There were several significant pair-wise associations (both positive and negative) in the distributions of the most common cactophilic species. Some negative associations could be explained as the outcome of the ability of some of yeast to kill yeasts with which they were negatively associated. WHEN some cacti are injured, the damaged tissue (necrosis; soft rot) may be colonized by a succession of bacteria, fungi, and insects (Fogleman and Foster, 1989; Ganter et al., 1986; Heed, 1977). Many of the fungi are ascomycetous yeast found only in the cactophilic habitat (Fogleman and Starmer, 1985; Ganter et al., 1986; Phaff et al., 1972; Starmer et al., 1987b; Starmer et al., 1982: Starmer et al., 1984). These yeast and the insects that eat and vector them have become a model system for the study of the ecology and evolution of microbial communities (Barker, 1992; Barker and Starmer, 1982; Heed and Mangan, 1986; Starmer et al., 1990). The mechanisms that structure cactophilic communities are only partially understood. Many cactophilic yeast species are restricted to a subset of cacti (Starmer et al., 1990) although virtually all of the yeast will grow in the laboratory on tissue from any of the cacti. Studies have correlated yeast distributions with variation in physical parameters (Barker et al., 1987) or variation in the distributions of vectoring insects (Barker et al., 1984; Ganter, 1988; Heed et al., 1976; Starmer et al., 1987). Experimental evidence implicates interactions among the yeast species as the most important factor influencing patterns of species occurrence (Ganter and Starmer, 1992; Starmer and Fogleman, 1986). Theoretical considerations make demonstration of some pairwise interactions between species difficult to discern in collection data (Hastings, 1987), and this has been true of many cactophilic yeast collections restricted to a single site or to a single time (Ganter et al., 1986). An exception to this generality occurs when interspecific interactions are strong, as is the case for the distributions of killer yeast species and species sensitive to their toxins in the Sonoran Desert (Ganter and Starmer, 1992). In this paper, we examine the distribution of cactophilic yeast at several sites sampled at more than one time and find evidence that cactophilic yeast in Florida Opuntia No. 1, 2 1994] GANTER ET AL—YEAST INTERACTIONS 5] form distinct communities, that strong pair-wise interactions occur within these communities, and that these interactions result from the ability of some yeasts to kill other members of their community. MATERIALS AND METHODs—Cacti were collected from four sites in Florida (Figure 1): Big Pine Key; Rowdy Bend Trail, near Flamingo in Everglades National Park; Archbold Biological Station, near Lake Placid; and Canaveral National Seashore, near Titusville. Opuntia stricta was sampled at all sites except Archbold Biological Station and Opuntia humifisa was sampled at Archbold Biological Station and Canaveral National Seashore. There are several species of columnar cactus and at least eight species of Opuntia found in Florida (Benson, 1982). Most species were excluded from this study. The populations Canaveral National Seashore Archbold Biological Station Rowdy Bend Everglades National Park Cactus Hammok *, Big Pine Key s ? Fic.1. Map of central and southern Florida with collection sites. 52 FLORIDA SCIENTIST [VOL 57 of columnar cacti and three of the Opuntia species found in the Keys are small and obtaining an adequate sample from these populations is difficult. One Opuntia species is naturalized from central Mexico and is known froma single population in central Florida. One is restricted to the northern quarter of the state and another toa small area in south central Florida. The wide distributions of O. stricta (from sand dunes along both east and west coasts) and O. humifisa (found throughout the state except for the very southernmost portion) make them suitable for regional comparisons. Cactophilic yeast communities have been described previously from only O. stricta (Starmer et al., 1988). Samples were taken by opening a necrosis with a sterilized knife and placing from 1 to 5 g of the semi-liquified cactus tissue into a sterile Whirl-pak®. In the laboratory, approximately 1 g of tissue was then diluted with 9 ml of sterile water, vortexed, and 0.1 ml of the suspension plated onto YM Agar® (Difco) that had been acidified with 0.1 N HCl toa pH of 3.8. The acidity of the plate inhibits the growth of many bacteria. At the same time that the tissue was sampled for yeast, its pH was recorded using pHydrion pH papers, allowing estimation of pH in half-pH units. The use of half-pH units is justified by the large range in pH measurements (3.0 to 10.0). Finer measurements would not affect variances. After three and six days of incubation at ambient temperature, the plates were inspected for morphologically distinct yeast colonies. Distinct colony types from each sample were purified by subculturing and identified by replica plating of the strain onto a standardized battery of defined media (van der Walt and Yarrow, 1984) modified for cactophilic yeast (Phaff et al., 1985). Sixty five tests were used. All strains were tested for the ability to kill Candida glabrata (strain Y-55, obtained from Dr. T. W. Young), a yeast sensitive to most toxins found in cactophilic yeast (Starmer et al., 1987a). Identifications were made by comparing the resulting physiological profile and the results of micro- scopic examination to published profiles and descriptions for all yeast (Barnett et al., 1983; Kreger-van Rij, 1984) and for cactophilic yeast (Lachance et al., 1988). Correlations among the proportional species compositions were used to compare the collections from different hosts taken at different sites and times. The data were angular transformed to obviate the relationship between mean and variance present in proportional data before the correlations were done (Sokal and Rolf, 1969). Clustering of the collections based on these correlations was done with the unweighted pair-group method using arithmetic averages (UPGMA) (Sneath and Sokal, 1973). The choice of UPGMA as a clustering algorithm allows comparison between our results and those from earlier publications about Opuntia community structure. ResuLts—A total of 337 yeast strains from 33 species were isolated from 131 Opuntia stem and fruit necroses (2.57 strains/rot — Table 1). Multiple isolates of a single species from the same rot are considered here as the same strain. By considering each location and host type as a separate location, the total can be divided into eleven separate collections. TaBLE 1. Number of yeast strains isolated from Opuntia cladode and fruit necroses at four locations in Florida. Each table entry represents the number of strains isolated of a particular species at a particular time and place. The total number of rots sampled (and therefore, the maximum number of strains possible) is entered at the top of the table Location! fi CNS CNS ENP ENP BPK BPK ABS_ ABS CNS CNS ABS Host? fi Os Os Os Os Os Os Oh Oh Oh OsF OhF Dates fi 12/89 1/91 12/90 2/91 6/90 12/90 12/89 1/91 12/89 12/89 12/89 1/91 1/91 Number of rots 22; 14 10 10 ll 1l 8 10 5 vu) 7 Yeast Pichia amethionina vf 12 eae Dire CCl 4 Candida boidinii 6 2 it Cryptococcus laurentii 3 1 i Kloeckera apiculata 5 9 2 No. 1, 2 1994] GANTER ET AL.—YEAST INTERACTIONS 53 TABLE 1. (Con’t) Location! fi CNS CNS ENP ENP BPK BPK ABS ABS CNS CNS ABS Host? fi Os Os Os Os Os Os Oh Oh Oh OsF OhF Dates fi 12/89 1/91 12/90 2/91 6/90 12/90 12/89 1/91 12/89 12/89 12/89 1/91 1/91 Number of rots 22 14 10 10 11 11 8 10 5 De 7 Yeast Pichia barkeri Cryptococcus ater Cryptococcus flavus Cryptococcus luteolus Rhodotorula minuta Candida guilliermondi Candida sonorensis Pichia cactophila Cryptococcus cereanus com.* Prototheca sp. Cryptococcus albidus Candida mucilagena Kloeckera apis Rhodotorula graminis Pichia kluyveri Rhodotorula javanensis 1 Clavispora opuntiae 8 4 8 a 3 3 i 2 Debaryomyces hansenii il Hansenula polymorpha 2 3 Hansenula nonfementans 2 Clavispora lusitaniae y) Metschnikowia bicuspidatus 1 Issatchenkia terricola Candida deserticola 1 Rhodotorula aurantiaca il 2 Pichia carsonii 1 Pichia membranefasciens 1 Cryptococcus heveanensis 1 Cryptococcus hungaricus il Total Isolates 64 42 26 aD, 31 26 18 24 19 47 18 ‘CNS = Canavera National Seashore, ENP = Everglades National Park, BPK = Big Pine Key, ABS = Archbold Biological Station : EHRN PRP RRB OCR DR Re pre re +l w w Os = Opuntia stricta, Oh = O. humifisa, OsF = Opuntia stricta fruit, OhF = O. humifisa fruit *y. f. = variety fermentans ‘com. = complex, used when impossible to distinguish related speces. 54 FLORIDA SCIENTIST [VOL 57 The clustering of the collections suggests that the species composition of Opuntia rots are influenced by host plant type, geography and time (Figure 2). To widen the data base, the clustering includes data from Starmer and co-workers (Starmer et al., 1988). They sampled O. stricta pads and fruit from two sites at Canaveral National Seashore in 1986 using techniques comparable with ours. We have grouped the data from both of their sampling sites for this analysis. They isolated 74 strains of yeast from 31 stem necroses (2.39 strains/rot) and 66 strains from 31 fruit necroses (2.19 strains/rot). Their overall average strain/rot value (2.26 Host Locale Date O. stricta BPK 12/90 O. stricta ENP 12/90 O. stricta ENP 2/91 O. stricta BPK 6/90 O. humifisa ABS 12/90 O. stricta CNS 1986 O. stricta CNS 12/90 O. stricta CNS 12/89 O. humifisa CNS 12/89 O. humifisa ABS 12/89 12/89 O. stricta fruit CNS & 12/90 12/89 O. humifisa fruit ABS & 12/90 Q. stricta fruit CNS 1986 1.0 0.8 0.6 0.4 0.2 0 -0.1 Correlation Coefficient Fic. 2.Tree diagram based on UPGMA clustering of yeast collections. CNS = Canaveral National Seashore, BPK = Pig Pine Key, ENP = Everglades National Park, ABS = Archbold Biological Station. The collections in boldface type are from Starmer et al. 1988 and from Starmer and Lachance [previously unpublished data] No. 1, 2 1994] GANTER ET AL—YEAST INTERACTIONS 55 strains/rot) is close to ours, reinforcing the assumption that our data and theirs are comparable. In addition, Starmer and Lachance collected in Everglades National Park in 1986, although they did not publish the data. They have kindly made the unpublished data available to us and it is included (Figure 2). All Opuntia stem communities cluster before joining any from Opuntia fruit necroses. In general, the correlations among fruit communities are lower than those between stem communities. This result does not seem to be due to sample size, as the fruit and cladode data from Starmer and co-workers (1988) represent the largest collections from fruit or stems and their fruit community is the last community to join the clustering. This result confirms studies from other cactophilic systems which have concluded that the fruit community is more “open” than the stem community (Starmer et al., 1990) (perhaps deserving the designation “aggregation” rather than “community” ). The clustering of Opuntia stricta stem samples suggests that the yeast commu- nity in the southern sites (Everglades National Park and Big Pine Key) differs from that in the northern site (O. stricta is present only at Canaveral National Seashore). With the exception of the 1986 Everglades collection, samples from the two areas belong to different clusters. The north-south difference is primarily due to the absence of “Pichia amethionina var. fermentans” , Candida boidinii, and Cryptococ- cus laurentii from the southern sites and the absence of Hansenula species from the northern site. “P. amethionina var. fermentans” is an undescribed variety of P. amethionina (Lachance et al., 1988) from Florida and Caribbean islands. The exceptional 1986 Everglades sample clusters with the northern samples because of the presence of “P. amethionina var. fermentans” (10% of the total strains isolated in that collection). The subsequent disappearance of this yeast from our samples is probably not due to seasonal variation in community composition, as the 1986 and 1990 samples were both collected in December. There was no clear trend for differences in community composition due to year of collection. The yeast communities from the two species of Opuntia sampled in this study differ. The six most abundant yeast species from Opuntia stricta represent 68% of all isolates from that cactus. They are (in order of abundance) Candida sonorensis > Clavispora opuntiae > Cryptococcus cereanus > P. amethionina var. fermentans = Prototheca sp. > P.cactophila > Candida boidinii (Prototheca is a green algae lacking chloroplasts). The same ranking for O. humifisa is C. sonorensis > C. mucilagina = C. guilliermondii > Clavispora opuntiae > P. amethionina var. fermentans = Hansenula polymorpha. These 6 species comprise 66% of all isolates from O. humifisa. The two rankings share only half of the six species and only C. sonorensis occupies the same rank in both. These rankings are quite distinct and deserve further attention. It is well documented that there are distinct communities of yeast associated with different cacti (Fogleman and Starmer, 1985; Ganter et al., 1986; Starmer, 1982; Starmer and Phaff, 1983). Congeneric species of Opuntia have been sampled by Barker and co- workers (Barker et al., 1984) from Opuntia naturalized in Australia. They found that differences in the yeast communities among Opuntia species were greater than differences among collection sites. In North America, comparisons of differences in 56 FLORIDA SCIENTIST [VOL 57 Opuntia yeast communities has usually been confounded by a lack of overlap in cactus distributions. The allopatry made it impossible to separate the effect of different Opuntia species on yeast community composition from geographic effects (edaphic differences, different vectors, etc.). Our results confirm that sympatric Opuntia have distinct yeast communities in their native habitats too. Discussion—The data collected confirm that there are patterns in cactophilic yeast species distributions attributable to distance, host cacti, and time. Further examination of the data provides evidence of patterns based on species-species interactions also. The simplest method of demonstrating convergent or divergent species distributions is contingency analysis. However, Hastings (Hastings, 1987) points out the difficulty of demonstrating associations with this technique related to the predominance of samples without either species being compared. This has been true of cactophilic collections (personal observations of the first author), and so it is unusual that these Florida collections provide several instances of significant associations (Fig. 3). The pH of the rots can be used as an indicator of the age of the rot. The pH of Opuntia stricta and O. humifisa tissue, measured between 11 AM and 3 PM, is near 4.5, as expected for a CAM plant (Gibson and Nobel, 1986). As a necrosis ages, the pH can rise as high as 10. In the field, it is not known what is responsible for this Cre Pro Pichia amethionina var. fermentans Clavispora opuntiae aan Candida 'mucilagena' eae Prototheca sp. Cryptococcus cereanus Fic. 3.Results of contingency analysis of associations between the seven species most commonly collected from Florida Opuntia stem rots. A plus (+) indicates a significant positive association, a minus (-) is a significant negative association (a = 0.05, G-test based on a 2 x 2 contingency table Sokal and Rolf, 1969). No. 1, 2 1994] GANTER ET AL.—YEAST INTERACTIONS 57 change in pH, but cactophilic yeast can effect the change in a week on homogenized Opuntia stricta and O. humifisa tissue in the laboratory (P. Ganter, unpublished data). Thus, yeast associated with high pH rots are associated with older rots. Of the seven most common species, two (C ryptococcus cereanus and Prototheca sp.) are over-represented in rots with high pH (Table 2). This may be due to a physiology best suited for the conditions characteristic of older rots or to the vectoring of these species by an insect attracted to older rots. There were no species associated with younger (low pH) rots, an outcome that may be an artifact of the technique. If pH increases as the rot ages, then association with low pH rots would be detectable only if those species that arrived early later disappeared from the rot. The data indicate that this is not so. Regression of number of species onto pH (r = 0.32, p < 0.015) demonstrates that the number of species increases as the rot ages (the pH rises). “Pichia amethionina var. fermentans”, Candida mucilagina, and C. sonorensis are candidates for early-arriving species, because they are positively associated and some have negative associations with Cr. cereanus and Prototheca sp., species associated with older rots (Fig. 3). It is not surprising that there are more positive associations than negative. Little experimental data exists about natural yeast interactions and the many possible interactions suggested by laboratory experiments makes predicting the direction of interaction in nature difficult. Yeast might compete for carbon or nitrogen resources or may poison a habitat for other species with metabolic waste. They may crossfeed on the products of other species’ extracellular enzymatic activity or from vitamins or nitrogenous compounds released from other yeast. Some yeast secrete pectinolytic enzymes and may modify the host tissue in a manner that inhibits or promotes other yeast’s growth. The interaction may be indirect. A modification of the necrosis that affects the attractiveness of the rot as a place for vectors to oviposit or feed may TABLE 2. Comparison of pH of necroses with and without one of the seven most common species collected from Florida Opuntia stem necroses. Ordered from most common species at the top to least common at the bottom. Significance level for comparison of mean pH of necroses with and without a particular species of yeast = 0.05, t-test. With Without Significant Candida sonorensis 6.8 al Clavispora opuntiae 6.9 CS - Pichia amethionina var. fermentans 7.0 Tell - Cryptococcus cereanus 6.8 7.8 yes Candida mucilagina 6.9 7.4 - Prototheca sp. 6.3 Ta yes Pichia cactophila 7.0 Coll - 58 FLORIDA SCIENTIST [VOL 57 change the probability that subsequent yeast are vectored there. There is no reason to assume that species interact in only one way or that the interaction does not change as the rot ages. All of which makes biologically meaningful interpretation of associations difficult without data that can promote or eliminate some possibilites. Killer toxins (proteins secreted by a yeast that kill sensitive yeast) are known to influence species distributions (Ganter and Starmer, 1992) and may help in interpeting the associations found here. A subset of strains from Florida have been cross-tested for their ability to kill one another (P. G. Ganter, E. Jacques, and N. Lima, unpublished data). Strains from Pichia kluyveri and “P.amethionina var. fermentans” were able to kill other yeast from their communities. “P. amethionina var. fermentans” is also negatively associated with Cryptococcus cereanus and Clavispora opuntiae. Here, association and mechanism coincide. “P. amethionina var. fermentans” killed strains of both species. “P. amethionina var. fermentans” killer toxin is unusual in that it reaches maximal killing activity at a pH greater than six (P. Ganter and E. Jacques, unpublished data). If“P. amethionina var. fermentans” is excluding Cr. cereanus and Cl. opuntiae from rots via production of killer toxin, then one would expect that the rots with both killer and sensitive yeast to be rare, as is true (Table 3) and to have a low pH (due to the pH-dependent toxin activity), as is also true (Table 3). This is an exception to the general finding that species do not disappear from rots (as discussed above) and is expected only if the interactions between “P. amethionina var. fermentans” and the sensitive species are pH dependent in such a way as to discourage co-habitation at higher pH. In fact, if there were no interaction between “P. amethionina var. fermentans” and the sensitive species, then rots at which both arrived should be found in at least equal proportion in young and old rots. Of all the significant associations found, both positive and negative, these are the only ones in which the rots containing both species were the youngest rots (Table 4). It is interesting to note the close agreement between the Opuntia stricta communities collected in 1986 (Starmer et al., 1988) and our collections. With the addition of a second cactus species, more can be added to the conclusions drawn by Starmer and co-workers. Neither collection contained Pichia norvegensis, a yeast TABLE 3. Mean pH of rots with and without P. amethionina var. fermentans and those yeasts negatively associated with it. One-way ANOVA indicates that the overall model is significant at the 0.005 level. Yeast Present: Mean pH N eee Cr. c. and/or Cl. o0.° Yes Yes 6.0 + No No 6.2 On Yes No 7.4 18 No Yes en 39 °P. a. f. = “Pichia amethionina fermentans”, Cr. c. = Cryptococcus cereanus, Cl. 0. = Clavispora opuntiae No. 1, 2 1994] GANTER ET AL.—YEAST INTERACTIONS 59 TaBLe 4.Mean pH of necroses with one, both or neither species of yeast present. The pairs of yeast species are those with significant associations. Those comparisons with stars have signficantly different pH distributions (ANOVA, p< 0.05) only only Sp. A! Sp.B! Neither Sp. A Sp. B Both Negative Associations Prototheca C. sonorensis 6.3 Tell tall Toll g Cl. opuntiae C. sonorensis 6.5 Doll 7.0 Ue Cl. opuntiae C. mucilagina 6.6 We es 7.0 Positive Associations Cl.opuntiae Prototheca 6.8 7.4 7.0 el Cr. cereanus Cl.opuntiae 6.6 8.3 (3 7.4 : Cr. cereanus C. sonorensis 6.8 6.8 6.9 8.1 si C. mucilagina C. sonorensis 6.8 6.0 6.9 Tol P. a. fermentans C. mucilagina 6.9 Cit 7.6 Gil Prototheca P. cactophila 6.8 7.8 6.7 C® P. = Pichia, a. = amethionina, Cr. = Cryptococcus, Cl. = Clavispora, C. = Candida common in O. stricta rots from Caribbean islands and in Opuntia rots from Baja California and Baja California Sur, Mexico, but not from Opuntia rots in the northern Sonoran desert (Starmer et al., 1990). Starmer and co-workers (Starmer et al., 1988) suggest that P. norvegensis is associated with semi-arid conditions, livestock, and human habitation. However, the collections reported here are as close to human habitations as were the Mexican collections and ranching is a major industry in Florida, especially near the Archbold site. Florida is not as dry as the Mexican sites nor are summer temperatures as high, but the northern Sonoran sites, which also lack P. norvegensis, are both as dry and as hot as the Mexican sites. The causes of P. norvegensis’ disjunct distribution may be a combination of chance and regional variation in the communities inhabiting Opuntia. It has been suggested that the distribution of Clavispora opuntiae may be associated with that of Cactoblastis cactorum, a phyticid moth used in the biological control of Opuntia (Dodd, 1940; Murray, 1982; Starmer et al., 1988; Starmer et al., 1990). Our data is consistent with that hypothesis. We found Cl. opuntiae in rots from Big Pine Key, Everglades National Park, and Archbold Biological Station, but not at the most northern site, Canaveral National Seashore (Table 1 and F ig. 1). We also found larvae and eggs sticks of the moth at Big Pine Key and at Archbold Biological 60 FLORIDA SCIENTIST [VOL 57 Station (from O. humifisa). A sample of frass collected from larvae in O. humifisa yielded no Cl. opuntiae (but did contain Candida sonorensis, P. kluyveri, and Kloeckera apis). The moth is not native to Florida, but may have invaded from Caribbean islands where it has been introduced as a control agent. This is somewhat alarming, as many of the species of Opuntia in Florida are known only from small populations and may disappear if the moth becomes established. ACKNOWLEDGMENTS—This work was supported by NIH MBRS Grant 1-S06-GM4545501, sub- project 3. Thanks to Archbold Biological Station, Canaveral National Seashore, and the Everglades National Park for permission to collect, to Tom Starmer and Andre Lachance for access to their data, and to Tom Starmer, Alan Sanborn, and Jeremy Montague for reading an earlier version. LITERATURE CITED BarkKER, J. S. F. 1992. Genetic variation in cactophilic Drosophila for oviposition on natural yeast substrates. Evolution 46:1070-1083. , P. D. East, H. J. PHAFF, AND M. Miranpa. 1984. The ecology of the yeast flora in necrotic Opuntia cacti and of associated Drosophila in Australia. Microb. Ecol. 10:379-399. AND W. T. STarMER. 1982. Ecological genetics and evolution: the cactus-yeast-Drosophila model system. Academic Press, New York, NY. , W. T. STARMER, AND D.C. VaceEk. 1987. Analysis of spacial and temporal variation in the community structure of yeasts associated with decaying Opuntia cactus. Microb. Ecol. 14:267- 276. BarNETT, J. A., R. W. PayNE, AND D. Yarrow. 1983. Yeasts: Characteristics and Identification. Cambridge University Press, Cambridge. Benson, L. 1982. The Cacti of the United States and Canada. Stanford University Press, Stanford, CA. Dopp, A. P. 1940. The Biological Campaign Against the Prickly Pear. Government Printer, Brisbane, Australia. FOGLEMAN, J. C. AND J. L. Foster. 1989. Microbial colonization of injured cactus tissue (Stenocereus gummosus) and its relationship to the ecology of cactophilic Drosophila mojavensis. Appl. Environ. Microbiol. 55:100-105. FOGLEMAN, J.C. AND W. T. STARMER. 1985. Analysis of the community structure of yeasts associated with the decaying stems of cactus. III Stenocereus thurberi. Microb. Ecol. 11:165-173. GanTeR, P. F. 1988. The vectoring of cactophilic yeasts by Drosophila. Oecologia 75:400-404. AND W. T. StarMeEr. 1992. Killer toxin as a mechanism of interference competition in cactophilic yeast communities. Ecology 73:54-67. , W. T. STARMER, M. A. LACHANCE, AND H. J. Parr. 1986. Yeast communities from host plants and associated Drosophila in southern Arizona: new isolations and analysis of the relative importance of hosts and vectors on community composition. Oecologia 70:386-392. Gipson, A. C. AND P. S. NoBEL. 1986. The Cactus Primer. Harvard University Press, Cambridge, MA. Hastincs, A. 1987.Can competition be detected using species co-occurrence data? Ecology 68:117-123. HEED, W. B. 1977. Ecology and genetics of Sonoran Desert Drosophila. Pp. 109-126. In: Brussarb, P. F. (ed.) Ecological Genetics: the Interface. Springer-Verlag, New York, NY . AND R. L. Mancan. 1986. Community ecology of the Sonoran Desert Drosophila. Pp. 311- 345. In: ASHBURNER, M., H. L. Carson, Anp J. N. THompson (eds.) The Genetics and Biology of Drosophila. Academic Press, New York, N.Y . , W. T. STARMER, M. Miranpa, M. W. MILLER, AND H. J. Puarr. 1976. An analysis of the yeast flora associated with cactiphilic Drosophila and their host plants in the Sonoran desert and its relation to temperate and tropical associations. Ecology 57:151-160. KREGER-VAN Ry, N. J. W. 1984. The Yeasts, a Taxonomic Study. Elsevier Science, New York, NY. LacuanceE, M. A., W. T. STARMER, AND H. J. Puarr. 1988. Identification of yeasts found in decaying cactus tissue. Can. J. Microbiol. 34:1025-1036. Murray, N. D. 1982. Ecology and evolution of the Opuntia-Cactoblastis ecosystem in Australia. Pp. 17- 30. In: BARKER, J. S. F. AND W. T. STARMER (eds.) Ecological Genetics and Evolution: the Cactus- Yeast-Drosophila Model System. Academic Press, Sydney, Australia . PuarF, H. J., M. W. MILter, M. YONEYAMA, AND M. Sonepa. 1972. A comparative study of the yeast flora associated with trees on the Japanese islands and on the west coast of North America. P. 759-774. No. 1, 2 1994] GANTER ET AL.—YEAST INTERACTIONS 61 In: Terul, G. (ed.), 4" International Fermentation Symposium in Osaka, Japan. , W. T. STARMER, J. TREDICK, AND M. Miranpa. 1985. Pichia deserticola and Candida deserticola, two new species of yeasts associated with necrotic stems of cacti. Int. J. Syst. Bact. 35:211-216. SneaTH, P. H. AND R. R. Soka. 1973. Numerical Taxonomy. W. H. Freeman, San Francisco, CA. SoKAL, R. R. AND F. J. RoLF 1969. Biometry. W. H. Freeman, San Francisco, CA. STARMER, W. T. 1982. Analysis of the community structure of yeasts associated with the decaying stems of cactus. I. Stenocereus gummosus. Microb. Ecol. 8:71-81. , V. ABERDEEN, AND M. A. LacHANCE. 1988. The yeast community associated with Opuntia stricta (Haworth) in Florida, with regard to the moth Cactoblastis cactorum (Berg). Florida Scient. 51:7-11. AND J.C. FoGLEMAN. 1986. Coadaptation of Drosophila and yeasts in their natural habitat. J. Chem. Ecol. 12:1035-1053. , P. F. GANTER, V. ABERDEEN, M. A. LACHANCE, AND H. J. PHaFF. 1987a. The ecological role of killer yeasts in natural communities of yeasts. Can. J. Microbiol. 33:783-796. , M. A. Lacuance, AND H. J. PHarF. 1987b. A comparison of yeast communities found in necrotic tissue of cladodes and fruits of Opuntia stricta on islands in the Caribbean Sea and where introduced into Australia. Microb. Ecol. 14:179-192. , M. A. Lacuance, H. J. PHarr, AND W. B. HEED. 1990. The biogeography of yeasts associated with decaying cactus tissue in North America, the Caribbean, and Northern Venezu- ela. Pp. 253-296. In: HEcut, M. K., B. WALLACE, AND R. J. MACINTYRE (eds.) Evolutionary Biology. Plenum Press, New York, NY. AND H. J. Puarr. 1983. Analysis of the community structure of yeasts associated with the decaying stems of cactus. II. Opuntia species. Microb. Ecol. 9:247-259. , H.J. Puarr, J. M. BowLEs, AND M. A. LacHance. 1987. Yeasts vectored by insects feeding on decaying saguaro cactus. Southwestern Naturalist 33:362-363. , H.J. Parr, M. Miranpa, M. W. MILLER, aND W. B. HEED. 1982. The yeast flora associated with the decaying stems of columnar cacti and Drosophila in North America. Pp. 269- 295. In: HECHT, M. K., B. WALLACE, AND C. T. PRANCE (eds.) Evolutionary Biology. Plenum Press, New York, NY. , H. J. Puarr, J. TREDICK, M. MIRANDA, AND V. ABERDEEN. 1984. Pichia antillensis, a new species of yeast associated with necrotic stems of cactus in the Lesser Antilles. Int. J. Syst. Bact. 34:350-354. VaNn Der WA_r, J. P. AND D. Yarrow. 1984. Methods for the isolation, maintenance, classification and identification of yeasts. Pp. 45-104. In: Ry, N. J. W. K.-v. (ed.) The Yeasts, a Taxonomic Study. Elsevier Science, New York, NY. Florida Scient. 57(1,2):50-61.1994 Accepted: February 15, 1994. 62 FLORIDA SCIENTIST [VOL 57 BOOKS RECEIVED AUDUBON, JOHN JAMES. 1993. Treasury of Audubon Birds in Full Color: 224 Plates for the Birds of America. Dover Publications, Inc., Mineola, NY, 112 pp. plates, 16 pp., text, $15.95 (paper). BILDSTEIN, KEITH L. 1993. White Ibis. Smithsonian Institution Press. Washing- ton. D. C., 272 pp., $22.50 (cloth). FORD, LEONARD, and E. WINSTON GRUNDMEIER, (ED.). 1993. Chemical Magic. 2 ed., Dover Publications, Inc., Mineola, NY, 128 pp., $5.95 (paper). FORRESTER, DONALD J. 1992. Parasites and Diseases of Wild Mammals in Florida. University Press of Florida, Gainesville, FL, 460 pp. $59.95 (cloth). GILBERT, CARTER (ed. ). 1992. Rare and Endangered Biota of Florida. II, Fishes. University Press of Florida, Gainesville, FL 246 pp., $59.95 (cloth), $26. 95 (paper). HOLDEN, ALAN. 1993. The Nature of Solids. Dover Publications Inc., Mineola, NY, 256pp., $6.95 (paper). HUMPHREY, S. R. (ed.). 1992. Rare and Endangered Biota of Florida. I, Mam- mals. University Press of Florida, Gainesville, FL 420 pp., $59.95 (cloth), $26. 95 (paper). JONES, DAVID L. 1993. Cycads of the World. Smithsonian Institution Press. Washington. D. C., 312 pp., $45.00 (cloth). KILHAM, LAWRENCE. 1983. Woodpeckers of Eastern North America. Dover Publications, Inc., Mineola, NY, 256 pp. $7.95 (paper). LAVENDA, BERNARD H. 1978. Thermodynamics of Irreversible Processes. Dover Publications, Inc., Mineola, NY, 182 pp., $7.95 (paper). MENNINGER, KARL. 1992. Number Words and Number Symbols: A Cultural History of Numbers. Dover Publications, Inc., Mineola, NY, 480 pp., $14.95 (paper). (Reprint of 1969 edition). PLATER, ZYGMUNT J. B., et al. 1992. Environmental Law and Policy: Nature, Law, and Society. West Publishing Co. St. Paul, Minnesota 1039 pp. (cloth). PLATT, RUTHERFORD. 1993. 1,001 Questions Answered About Trees. Dover Publications, Inc., Mineola, NY,352 pp., $7.95 (paper). SABROSKY, CURTIS W., GORDON F. BENNETT, and TERRY L. WHITWORTH. 1989. Bird Blow Flies (Protocalliphora) in North America. Smithsonian Institution Press. Washington. D. C., 312 pp., $16.95. Selected Environmental Law Statues 1992-1993. West Publishing Co. St. Paul, Minnesota 1295 pp. (paper). THE NATURAL HISTORY MUSEUM, LONDON. 1993. Great Bird Illustrations Postcards in Full Color: 24 Cards. Dover Publications, Inc., Mineola, NY, 24 cards, $3.95 (paper). THE NATURAL HISTORY MUSEUM, LONDON. 1993. Great Flower Prints : Postcards in Full Color. Dover Publications, Inc., Mineola, NY, 24 cards, $3.95 (paper). No. 1, 2-1994] 63 WEISFELD, VICTORIA D., (ed.). 1991. AIDS Health Services at the Crossroads: Lessons for Community Care. The Robert Wood Johnson Foundation, Princeton, NJ 136 pp. (paper). REVIEW Leonard A Ford, Chemical Magic, Second Edition, Revised by E. Winston Grundmeier, Dover Publications, Inc., 31 East 2nd Street, Mineola, N.Y., 1993. Pp ix+ 109. Price $5.95 (paper) This book is a compilation of 104 demonstrations that have proven value for either teaching or for entertainment purposes. The book is oriented toward the entertainment value of the chemical trick and includes introductory tips on presen- tation that creates part of the atmosphere for successful entertainment. There is enough detail on each magic trick to provide guidance for a “magician” to perform the demonstration and expect reasonable success without the instructions becoming too detailed. There are no illusions in these magic tricks. The demonstrations are visual and sound effects caused by the chemical reaction or phase changes to create the entertainment. Since each demonstration is based upon sound physical and chemical phenom- ena, these demonstrations can be adapted to use in the classroom to illustrate scientific principles. The demonstrations are collected together in categories based upon similar subject matter. The categories include: foams, gas liberation, air pressure, vaporization, explosions, crystallization, freezing and gel formation, smoke and vapors, specific gravity, polymerization, delayed or consecutive reactions, and miscellaneous. However, the lecturer in the classroom will need to supply the background information to describe the scientific principles for the demonstration. Some of the demonstrations are regularly illustrated in the common textbooks. For example, the thermite reaction is presented as a demonstration and many textbooks feature pictures of this reaction. There is information warning of the possible hazards and recommended safety precautions for each demonstration. —Robert F. Benson, Institute for Environmental Studies, Department of Chemistry, University of South Florida, Tampa. 64 FLORIDA SCIENTIST [VOL 57 LETTER TO THE EDITORS We have not had a Letters column because the letters we receive are personal, professional ones dealing with individual manuscripts. But we recently received a letter from Dr. Car] Luer, Mote Marine Laboratory, Sarasota, that we should like to share. “Dear Dean and Barbara: As Chair of the Medalist Committee of the Florida Academy of Sciences, it is my pleasure to inform you that you have been selected as co-recipients of the 1994 Medalist Award. To my knowledge, this is the first time that the award has been shared, but in this case it seems highly appropriate since your significant contribu- tions to science have always been a shared venture. You are both to be congratulated for achieving this highest recognition the Academy can bestow. I look forward to seeing you both at the Annual Banquet, where the Medalist Award will be presented.” We are truly grateful. DFM, BBM INSTRUCTIONS TO AUTHORS Individuals who publish in the Florida Scientist must be active members in the Florida Academy of Sciences. Submit a typewritten original and two copies of the text, illustrations, and tables. All typewritten material—including the abstract, literature citations, footnotes, tables, and figure legends—shall be double-spaced. Use one side of 8'/2 x 11 inch (211/2 cm x 28 cm) good quality bond paper for the original; the copy may be xeroxed. Margins should be at least 3 cm all around. Number the pages through the Literature Cited section. Avoid footnotes and do not use mimeo, slick, erasable, or ruled paper. Use metric units for all measurements. Assistance with production costs will be negotiated directly with authors of pages which exceed 10 printed pages. Excess over 10 pages by an author in a single volume will result in charges for additional pages. AppreEss follows the author’s address. ApsTRACT—AIll manuscripts shall have a short, concise, single-paragraphed abstract. The abstract follows immediately the author’s address. ACKNOWLEDGMENTS are given in the body of the text preceding immediately the Literature Cited section. LITERATURE CITED section follows the text. Double-space every line and follow the format in the current issue. Manuscrirts of 5 or less, double-spaced typewritten pages should conform to the short article protocol. See recent issue for proper format of both long and short articles. TABLES shall be typed on separate sheets of paper, double-spaced throughout. Each table shall contain a short heading. Do not use vertical rulings. Maximum character width, including spaces, of tables is 98.0. This includes a minimum space of 5 characters between columns. Tables are charged to authors at $13 per quarter page, $29 per half page, and $39 per full page. ILLusTRATIONS—All drawings shall be done in good quality India ink, on good board or drafting paper. Letter by using a lettering guide or equivalent. Typewritten letters on illustrations are unaccept- able. Drawings and photographs should be large enough to allow 1/3 to 1/2 reduction in size. Photographs shall be glossy prints of good contrast. Whenever possible, mount photographs in lots size. The author’s name and figure number should be penciled lightly on the back of each figure. Figure legends must be listed on a separate page and not on the drawing or photograph. The legend must be double spaced. Illustrations are charged to authors at $6.00 per page or fraction. We ask authors to prepare accepted manuscripts in personal computer formats. Please do not submit a manuscript on disk. Once reviewers have recommended a manuscript be accepted, you will be given directions for submitting the final manuscript on disk. Proor must be returned promptly. Notification of address change and proofreading are the author’s responsibility. Alterations after the type has been set will be charged to the author. REPRINTS may be ordered through the FAS office from the printer on forms provided at the time that proofs are sent to authors for checking. FLORIDA ACADEMY OF SCIENCES INSTITUTIONAL MEMBERS FOR 1994 Archbold Biological Station St. John’s River Community College, Palatka Brevard Community College, Florida Advanced St. Petersburg (Florida) Public Library Technology Center Library, Palm Bay St. Thomas University, Miami, FL The British Library, Boston Spa, Weatherby, U.S. Department of Commerce, National UK Marine Fisheries (Panama City) Charles Courtney / Hillsborough County EPC U.S. Environmental Protection Agency Library, Florida Institute of Technology Gulf Breeze, FL Harbor Branch Oceanographic Institution, Inc. University of Central Florida, Division of Lake-Sumter Community College Sponsored Research Newfound Harbor Marine Institute University of South Florida CORPORATE MEMBERS FOR 1994 Mote Marine Laboratory, Sarasota Membership applications, subscriptions, renewals, changes of address, and orders for back numbers should be addressed to the Executive Secretary, Florida Academy of Sciences, P.O. Box 033012, Indialantic, FL 32903-0012 PUBLICATIONS FOR SALE by the Florida Academy of Sciences Complete sets. Broken sets. Individual numbers. Immediate delivery. A few num- bers reprinted by photo-offset. All prices strictly net. Special bargain: $500 for nearly complete sets. Prices quoted include domestic postage. Some issues may not be available. All are $20 per volume or $5 per issue, except for symposium issues. PROCEEDINGS OF THE FLORIDA ACADEMY OF SCIENCES (1936-1944) Volumes 1-7 QUARTERLY JOURNAL OF THE FLORIDA ACADEMY OF SCIENCE (1945-1972) Volume 8-35 FLORIDA SCIENTIST (1973) Florida’s Estuaries—Management or M ismanagement P—Academy Symposium FLORIDA SCIENTIST 37(4)—$5.00 Land Spreading of Secondary Effluent—Academy Symposium FLORIDA SCIENTIST 38(4)—$5.00 Solar Energy—Academy Symposium FLORIDA SCIENTIST 39(3)—$5.00 (includes do-it-yourself instructions) Anthropology—Academy Symposium FLORIDA SCIENTIST 43(3)—$7.50 Shark Biology—Academy Symposium FLORIDA SCIENTIST 45(1)—$8.00 Future of the Indian River System—Academy Symposium FLORIDA SCIENTIST 46(3/4)—$15.00 Second Indian River Research Symposium—Academy Symposium FLORIDA SCIENTIST 53(3)—$ 15.00 Please send payment with order. If required, an invoice will be sent on purchases over $20 from a recognized institution. Please include Florida sales tax if shipped to a Florida address, unless a copy of a valid Florida Consumer’s Sales Tax Certificate is enclosed. The Florida Academy of Sciences, Inc. NON-PROFIT P.O. Box 033012 ORGANIZATION Indialantic, Florida 32903-0012 U.S. Postage PAID Permit #26 Melbourne, FL F Gi ISSN: 0098-4590 H Florida cientist Volume 57 Summer, 1994 7 Number 3 CONTENTS Fog Temporarily Increases Water Potential in Florida Scrub Oaks. ............. Eric Menges _ 65 puprmmeiated Bibliography of Lyngbya ...............,s:0s-.dineessosnssonsseesseons debaowavee Dean F. Martin, Barbara B. Martin, Elsie D. Gross and Karen Brown 75 Morphometric Quality of Wild Turkeys Harvested from Public vs. Private Se tin Tl oarira ey Ba eae a rete ON tha Pee RE OLE UTR aR David T. Cobb 88 Observations on Reproduction of an Endangered Cactus, Cereus Robinii SWenmramce) Wis BEMS ONT wie, bccckelanceasscsc0.05). 68 FLORIDA SCIENTIST TABLE 1. Weather statistics for the 1988-1989 and 1989-1990 winter dry seasons. Month October November December January February March April May June Total Oaks at weather station, minimum predawn water potential from 2 individuals, MPa % fog = % of observed days with fog Normal Rainfall 6.30 6.98 5.92 10.87 Papas 76.35 PAE 60.42 1988-1989 Yo % Heavy Min.y* Fogt Fog} NA NA NA NA NA NA NA NA NA -0.54 = 53 38 -0.48 43 50 -0.32 50 59 -0.26 45 30 -0.32 31 50 -0.28 NA NA _ 45 +4 t t % heavy fog = % of observed foggy days with heavy fog NA data not available Rain (cm) 23.37 69.56 1989-1990 % Min. y* Fogt NA NA -0.28 59 -0.34 35 -0.52 64 -0.92 35 -0.90 40 -0.52 38 -0.50 48 -0.02 NA — 46 [VOL 57 % Heavy Fog} NA 46 33 75 57 M2: 38 30 NA 46 TABLE 2. Daily changes in soils and plant water relations parameters, for two winter dry seasons, in relation to whether rain occurred between the two days. Positive numbers indicate increases from one day to the next; negative numbers indicate decreases. Change in: Soil Moisture (%); 20-40 cm Quercus chapmanii Water potential (MPa) Quercus myrtifolia Water pontential (MPa) Year 1988-1989 1989-1990 1988-1989 1989-1990 1988-1989 1989-1990 * p value in t test contrasting days with rain vs. days without rain Days with Rain +0.09 +0.46 +0.03 +0.09 +0.03 +0.09 Days without Rain 0.07 0.18 0.01 0.03 0.00 0.02 No. 3 1994] MENGES—FOG INCREASES WATER POTENTIAL 69 In contrast, during the wetter 1989-1990 dry season, significant shifts in daily water relations parameters occurred on rainy days (Table 2). For example, on rainy days, mean plant water potential increased (on average) 0.09 MPa; while the 24-hour decrease when no rain occurred averaged 0.02—0.03 MPa (depending on species). The mean net effect was thus about 0.11—0.12 MPa. Light rains (less than 0.5 cm) did significantly affect water potentials of both plants (Duncan’s multiple range test, 3 classes, p<0.05). The amount of change in water relations parameters was positively (but not significantly) related to rainfall amounts. For all these analyses, there were no significant relationships between plant or soil water relations parameters and rainfall levels more than 24 hours previous (1— A days). In fact, soils tended to become drier and plants more stressed during the second day after rain. Given the potential of light rains to affect the next day’s water relations, days with rain were not considered in analyses of fog effects. Fog Effects on Water Relations—Day to day changes in plant water relations were affected by morning fog, but fog did not affect soil moisture. These patterns were consistent in both dry seasons (Table 3). Predawn water potentials in Quercus myrtifolia were more strongly affected by fog in 1989-1990 than in the previous year. Fog effects on water relations of Q. chapmanii were similar in both years (Table 3). The major effects of fog on plant water potential occurred with either light or heavy fog (Fig. 1). For Q. chapmanii in both dry seasons (Fig. la, 1b) and Q. myrtifolia in 1989-1990 (Fig. 1d), I found significant differences between no fog and both light and heavy fog (Scheffe’s test). For Q. myrtifolia in 1988-1989, the only TABLE 3. Fog effects on daily changes in soil and plant water parameters during two winter seasons, as detected by one-way ANOVAst. F value and level of significance Shallow Deep Quercus Quercus Soil Soil chapmanii _—myrtifolia Year Fog Index} Moisture Moisture w(predawn) yy (predawn) 1988-1989 Fog Rating 0.07 0.53 6.59** o29>= Visibility 0.48 0.75 3.03° 3.90° 1989-1990 Fog Rating 0.23 0.06 8:90*=* 16.06*** Visibility 0.32 0.10 SuWloye= o.00° °° * p<0.05; °* p<0.01; °°* p<0.001 t Analysis considers effects same morning as fog on rainless days. Effects of fog from previous days (up to 3 days) not significant. df (2, 29) for 1988-1989; df (2, 86) for 1989-1990. {Fog Rating: no fog, light or patchy fog, or heavy fog. Visibility: number of telephone poles visible. See methods for more detail. 70 FLORIDA SCIENTIST [VOL 57 0.12 QUERCUS CHAPMANII QUERCUS MYRTIFOLIA a 1988-1989 1988-1989 a = 0.06 =!) eosin Y = erythrosin >> fluorescein. Relative rate change (rate of test relative to control samples) was a linear function of the quantum yield for singlet oxygen, and the relationship was also applicable to methylene blue. Jorrner, J. 1987. Volatile organic substances. Chapter 18 In: Fay, P. AND C. VAN BaaLeN (eds.), The 89 FLORIDA SCIENTIST [VOL 57 Cyanobacteria. Elsevier, New York, NY. Includes occurrences of geosmin and 2-methylisoborneol in different axenic and non-axenic strains of cyanobacteria, including Lyngbya aestuarii and L. cryptovaginata. Kuwapa, Y. aND Y. Onta.1989. Hydrogen production and carbohydrate consumption by Lyngbya sp.(No.108). Agric. Biol. Chem. 53: 2847-2851. Reserve glycogen or other carbohydrate were used as sources of electron donors for hydrogen production, and the NaCl content of the medium affected the hydrogen production. Maximum production occurred at 3% NaCl. Monofluoroacetic acid (20 mmol/L) inhibited hydrogen production. , Y. INouE, K. KoIkE, AND Y. Outa. 1991. Functional and structural changes of PSII of Lyngbya sp. under hydrogen-producing conditions. Agric. Biol. Chem.55: 299-305. Structural and functional changes in Photosystem (PS)II in Lyngbya cells were studied. Changes occurred when cells were transferred from growing to hydrogen-producing conditions. Lyngbya sp. was isolated from estuary of Ashida River on the coast of the Seto Inland Sea. Martin, B. B. anD D. F. Martin. 1987. Effects of four dyes on rates of oxygen production by three filamentous algae. Microbios Letters, 35: 151-154. Lyngbya majuscula and two other filamentous algae were treated with four dyes, and the rates of oxygen production were measured with a Warburg apparatus. The dyes function as sensitizers for production of singlet oxygen. Rose bengal was the most effective, and uM concentration produced about a 60% reduction in the relative rate of oxygen production. , D. F. Martin, AND M. J. PEREZ-Cruet. 1987. Effect of selected dyes on the growth of filamentous blue-green alga, Lyngbya majuscula. J. Aquat. Plant Manage. Soc. 25:40-43. The effects of four selected dyes on the growth of Lyngbya majuscula were studied, both under ambient conditions, and by measuring the rate of oxygen production with a Warburg apparatus. Dyes act as photosensitizers, based upon a diagnostic test with sodium azide. Also the order of effectiveness of the dyes was related to their (known) quantum yields. Meta, R. AND K. J. Hawxsy. 1977. Use of UV to achieve bacteria-free algal culture. Proc. Okla. Acad. Sci. 57: 54-60. Lyngbya birgei and two other cultures were exposed to UV radiation for varying times and intensities. “No bacterial growth was observed for first 64 hours when treated algae were inoculated on nutrient agar plates. Algal pigments were not affected by radiation.” Mobs ey, A. 1992. Kings Bay/Crystal River. Hydroscope 23 (1): 10-11. Description of Southwest Florida Water Managment District SWIM (Surface Water Improve- ment and Management) program for Kings Bay/ Crystal River. Hydrilla verticillata and Lyngbya sp. “pose another set of water quality problems.” Though both are nuisance plants that impair navigation, at least hydrilla “does have one saving grace: it is a food source for the manatees.” Morkena, S. N.,G. W. Cuu, anp L. R. Bercer. 1971. Dermatitis- producing alga Lyngbya majuscula Gomont in Hawaii. I. Isolation and chemical characterization of the toxic factor. J. Phycol. 7:4-8. Purification by extraction of dried plant with chloroform-methanol (2:1), followed by TLC. General characterization. Bioassay with human amnion cells grown in tissue culture. No. 3 1994] MARTIN ET AL.—BIBLIOGRAPHY OF LYNGBYA 83 AND G.W. Cuu. 1971. Dermatitis-producing Lyngbya majuscula Gomont in Hawaii. II. Biological properties of the toxic factor. J.Phycol. 7:8-13. Intracutaneous injection of toxic factor produces severe acute inflammatory reaction in guinea pigs. Antibacterial activity studied. Marine isolate. MonecuE, R. L. anD E. J. Puuips. 1991. The effect of cyanophages on the growth and survival of Lyngbya wollei, Anabaena flos- aquae, and Anabaena circinalis. J. Aquat. Plant Manage. 29:88- 93. Three newly isolated cyanobacteria viruses tested on title organisms in laboratory culture experiments. “Cyanophage LW 1 significantly reduced the growth and survival of L. wollei....” Inhibition of growth occurred within 7 days and chlorophyll a concentrations were reduced 95% (relative to controls) within 14 days of inoculation. Moore, R. E. 1977. Toxins from blue-green algae. BioScience 27:797-802. “Only a few reports of toxicity among the marine blue-green algae have appeared in the literature. Unlike the freshwater species, the marine cyanophytes have not presented serious health and economic problems. Possibly the closest exception is L ynght ya a which i is responsible for sporadic outbreaks of contact dermatitis known as ‘swimmers’itch’ . AND M. ENTzEROTH. 1988. Majusculamide d and deoxymajusculamide D, two cytotoxins from Lyngbya majuscula. Phytochemistry 27:3101-3103. Title compounds are cytotoxic lipopentapeptides, and are minor constituents of deep-water variety of L. majuscula from Enewetak. MynbersE, J.S., A. H. Hunt, ann R. E. Moore. 1988. 57- Normajusculamide c, a minor cyclic depsipeptide isolated from Lyngbya majuscula. J. Nat. Prod. 57: 1299-1301. Title compound was isolated from “the deep-water variety of Lyngbya majuscula growing in the lagoon of Enewetak Atoll in the Marshall Islands.” Compound possesses significant activity against fungal plant pathogens — Phytophora infestans and P. viticola — that cause, respectively, tomato late blight and grape downy mildew. Neccui O., Jr. 1992. Microagal dynamics in a spring in Sao Paulo State, Southeastern Brazil. Arch. Hydrobiol. 124:489-499. Dynamics of a rheocrene-type spring studied. Four species dominate: Lyngbya putealis and Scytonema arcangeli (Cyanophyta) Klebsormidium subtile (Chlorophyta), and Batrachospermum delicatulum (Rhodophyta). “The dynamics of the microalgal community was marked by wide variability and impredictability, and showed no evident relation to any physical or chemical parameter of the water.” Paer_, H. W., L. E. PRuFERT, AND W. W. AmprosE. 1991. Contemporaneous N, fixation and oxygenic photosynthesis in the nonheterocystous mat-forming cyanobacterium Lyngbya aestuarii. Appl. Environ. Microbiol. 57: 3086-3092. The cyanobacterium was isolated “from laminated mats in an intertidal lagoon on Shackelford Banks, one of North Carolina’s coastal barrier islands.” Purified populations were isolated; bacteria-free cultures were prepared and studied. A photosystem I (O, evolution) inhibitor was added (<4 hr), and light-mediated N, fixation was stimulated. Results of electron microscopy and “CO, fixation studies suggest ‘lateral partitioning of photosynthesis and N, fixation during Ae aeiion with [the latter] being confined to terminal regions. 84 FLORIDA SCIENTIST [VOL 57 Pups, E. J., J. luNAT, AND M. Conroy. 1992. Nitrogen fixation by the benthic freshwater cyanobacterium Lyngbya wollei. Hydrobiologia 234:59-64. The mat-forming ability of L. wollei, collected from Lake Okeechobee, where it forms extensive mats, was studied using field samples and axenic cultures. Maximum rates of acetylene reduction were 39.1 and 200 nmole/mg dry wt/h for field and axenic cultures, respectively. Reduced oxygen levels and low light were required for nitrogenase development, and the level of irradiance had a significant impact on the rate of photosynthesis up to about 500 umol/m7/sec. , R. L. MONEGUE, AND F. J. ALDRIDGE. 1990. Cyanophages which impact bloom-forming cyanobacteria. J. Aquat. Plant Manage. 28:92-97. Describes isolation of cyanophages that affect L. wollei and two other cyanobacteria. Futher studies: see Monegue and Phlips (1991). Ramey, V. 1987a. Lyngbya biocontrol research. Aquaphyte 7(1):14,16. Summarizes proposed research by E. J. Phlips [See Phlips et al., 1990]. .1987b. Lyngbya physiology research. Aquaphyte 7(1):14. Summarizes research on Lyngbya sp. by Beer and Spencer (see above) REYMonp, O. L. AND B. HickEL. 1986 Large intracytoplasmic crystals in Lyngbya (Cyanophyceae). Phycologia 25: 397-411. Investigation of large crystalline inclusions on samples of Lyngbya subtilis W. West (Syn. L. limnetica Lemm) from eutrophic Plusse (a lake near Plén, Germany) REYNOLDS, C. S., J. G. TUNpIsI, AND K. Hino. 1983. Observations on a metalimnetic Lyngbya population in a stably stratified tropical lake (Lagoa Carioca, eastern Brasil). Arch. Hydrobiol. 97: 7-17. This is the first reported example of the thermal stratification of a cyanobacterium, Lyngbya limnetica Lemmerman, in a tropical lake. The stratification is discussed in terms of known buoyancy regulation in gas- vacuolate cyanobacteria. RoBINsoNn, C.P., D.R. FRANZ, AND M.E. Bonpura. 1991. Effects of lyngbyatoxin A from the blue-green alga Lyngbya majuscula on rabbit aorta contractions. Toxicon 29: 1009-1017. Title compound is a potent tumor promoter anda kinase C activator. Lyngbyatoxin A (1 UM) also slowly causes contractions in rabbit aorta rings becoming maximal at 3 hours. Pharmacological probes were used to evaluate the mechanism of action. ScuHMIDT, J.C. 1990. How to Identify and Control Water Weeds and Algae. 4th ed. Applied Biochemists, Inc. Mequon, MI 108 pp. Lyngbya is covered. ScHRADER, K. K. AND W. T. BLEvins. 1993. Geosmin-producing species of Streptomyces and Lyngbya from aquaculture ponds. Can. J. Microbiol. 39: 834-840. Authors analyzed water and sediment samples (monthly) from six aquaculture ponds in east- central Alabama for presence of actimomycetes and cyanobacteria capable of producing geosmin and/or 2-methylisoborneol. Lyngbya ctf. subtilis produced the first compound, but not the No. 3 1994] MARTIN ET AL.—BIBLIOGRAPHY OF LYNGBYA 85 second, and was prevalent during the sampling period in a “pond with a history of off-flavor problems.” SHANNON, K., E. D. Gross, AND D. F. Martin. 1992. Variation of growth of Lyngbya majuscula as a function of salinity. Biomed. Lett.47:29-33. Growth of the blue-green alga Lyngbya majuscula was studied in mixtures of freshwater growth medium and salt water to aid in identification and to determine the adaptability from fresh to saltwater. SonzocnlI, W. C., W. M. Prepavicn, J. H. STANRIDGE, R. E. WEDEPOHL, AND J.G. VENNIE . 1988. A note on algal toxins in Wisconsin waters experiencing blue-green algal blooms. Lake Reservoir Manage. 4:281-285. Samples from 86 lakes and ponds in Wisconsin were collected and tested for toxins produced by cyanophytes in an effort to account for sporadic incidences of animal deaths. Genera most commonly observed were: Microcystis, Anabaena, Aphanizomenon (in order of observation), Oscillatoria, Lyngbya, and Gloeotrichia. Samples from 25 sites were toxic (mouse bioassay), nearly all were hepatotoxic. “All tests of potable water supply systems, both raw and finished, were negative for algal toxins.” SPEZIALE, B. J. AND L.A. Dyck. 1992. Lyngbya infestations: Comparative taxonomy of Lyngbya wollei comb. nov.(cyanobacteria). J.Phycol. 28:693-706. Thorough consideration of conflicting identifications of the title species, which is “Exceptional in size and growth form and thus easily recognizable.” Tendency to confuse Plectonema wollei with L. wollei in the early literature is reviewed. False branching, a characteristic of the former species, is environmentally variable. Authors prefer to assign species name wollei to distinguish marine species (L. majuscula) and emphasize “ the essentially freshwater habitat and distinct growth form of massive floating mats.” , G. TURNER, AND L. Dyck. 1988.” Giant” Lyngbya. Aquatics 10:4-11. Considers classification and description, field identification, distribution and habitat, seasonal cycles and control techniques. “Giant” refers to dimensions of typical Lyngbya cell ( diameter 30- 70 um, length 4-7 ttm) that “could be packed with more than ten thousand bacterial cells.” Also coarse hair-like filaments of Lyngbya can reach lengths in excess of tens of centimeters, and these intertwine in a healthy mat. , G. Turner, AND L. Dyck. 1991. Physiological characteristics of vertically stratified Lyngbya wollei. Lake Res. Manage. 7:107-114. “In situ standing crops (up to 6.66 kg fresh weight m*) , seasonal vertical distribution (surface, benthic,, and mid-depth suspended) and metabolic characteristics of L. wollei mats in Marten’s Pond, S.C. are described and compared with midsummer infestations in three other southeast- ern impoundments....The bulk of L. wollei biomass (> 64%) in Marten’s Pond remained subsurface throughout the year.” THAJUDDIN, N. AND G. SUBRAMANIAN. 1992. Survey of cyanobacterial flora of the southern coast of India. Bot. Marina 35:305-314. Survey of 500 km of southern east coast of India from Nagore (Bay of Bengal) to Cape Comorin. Survey includes cyanobacteria in four areas: open sea, stagnant sea water, ponds/puddles, backwaters, and salt pans. Some 21 Lyngbya spp. were identified, some in all four areas. “Fifty species from 19 genera ... were found in salt pans with a salinity of over 50°/o0.” 86 FLORIDA SCIENTIST [VOL 57 TuomastTon, W. W. 1984. The status of undesirable aquatic weeds in Georgia during 1983. Aquatics 6(1):9-10. A review of troublesome aquatic plants. Lyngbya sp., present in various lakes, was well established in Lake Blackshear (near Cordele, GA), where it was a problem. Control was achieved using Aquazine (1 ppm ) early in the spring, and 100% control was observed, but the plant reappeared in 3-4 weeks. Control required repeated applications of herbicide. Diquat (1 gal/ acre) was successful in controlling a publicly-owned fishing lake (8 acres) near Milledgeville. TRIMBEE, A. M. AND G. P. Harris. 1984. Phytoplankton population dynamics of a small reservoir. Effect TUBEA, ZOLCZY of intermittent mixing on phytoplankton succession and the growth of blue-green algae. J. Plankton Res. 6:699-714. Seasonal successions in Guelph Lake Ontario compared for 1981 and 1982. Main difference: Aphanizomenon aeruginosa present in much greater abundance in 1982. Lyngbya birgei appeared earlier in 1982, but didn’t immediately increase in abundance as it had in 1981. AND G. P. Harris. 1984. Phytoplankton population dynamics of a small reservoir. Use of sedimentation traps to quantify the loss of diatoms and the recruitment of summer bloom- forming blue-green algae. J. Phytoplankton Res. 6:897-918. Specific migration rates of Lyngbya birgei fragments (and two other cyanobacteria) were high at 10 m in August (Guelph Lake, Ontario). About 2-4% of the maximum standing crop was due to recruitment from the sediments. Appearance of L. birgei corresponded to high surface temperature and anoxic conditions at 10 m. B., K. Hawxsy, AND R. Menta. 1981. The effects of nutrient, pH and herbicide levels on algal growth. Hydrobiologia 79: 221-228. Growth inhibition of Lyngbya birgei increased with increasing concentrations of herbicides (prometryn and fluometuron and dinoseb). Picloram had no effect. Growth rates of L. birgei were unaffected by different levels of K or P, growth rates were higher with high levels of N or pH, except when combined with prometryn or fluometuron. NSKI, S. J. AND B. W. Situ. 1980. Evaluation of white amur (Ctenopharyngodon idella) for contol of Lyngbya in a 32 hectare public fishing lake. Proc. South. Weed Soc. 33:196- 203. White amur (Ctenopharyngodon idella), 74 per hectare, were stocked (fall and winter of 1975- 76) in Coffee County Public Fishing Lake (in southeast Alabama, a 32.4 ha lake) in an effort to control Lyngbya sp. “Surveys and samples collected over a four-year period indicated an 84% control by weight and an 87.7% control by area.” Total number of fish was kept constant by monitoring removals and restocking. ACKNOWLEDGMENTS— Weare grateful for the helpful comments of the reviewers and those of Dr. Patricia M. Dooris, who served as consulting editor. OTHER LITERATURE CITED Cow LL, B. C. 1991. Department of Biology, University of South Florida, Tampa, FL, Pers. Commun. LovELL, T. 1974. Environment-related off-flavors in intensively cultured fish. Pp. 259-262. In: KREUZER, R. (ed.) Fishery Products, Fishery New (Books) Ltd., Surrey., England. Persson, P. E. 1984. Uptake and release of environmentally occurring odorous compounds by fish. Water Res. 18:1263-1271. ScHARDT, J.D. AND D.C. Scuitz. 1991. 1990 Florida Aquatic Plant Survey, Tech. Rpt. 91-CGA, Bureau No. 3 1994] MARTIN ET AL.—BIBLIOGRAPHY OF LYNGBYA 87 of Aquatic Plants, Florida Dept. Environ. Protection, Tallahassee, FL. SPEZIALE, B. 1990. Department of Biological Sciences, Clemson University, Clemson, SC, Pers. Commun. Tarver, D. P., J. A. Ropcers, M. J. MAHKER, AND R. L. Lazor. 1978. Aquatic and Wetland Plants of Florida. Florida Department of Natural Resources, Tallahassee, FL. Florida Scient. 57 (3):75-87.1994. Accepted: March 22, 1994. 88 FLORIDA SCIENTIST [VOL 57 Biological Sciences MORPHOMETRIC QUALITY OF WILD TURKEYS HARVESTED FROM PUBLIC VS. PRIVATE LANDS IN FLORIDA Davip T. Coss Florida Game and Fresh Water Fish Commission, Rt. 7, Box 3055, Quincy, FL 32351 Apstract: Morphometric variables (body weight, beard length, spur lengths) were compared from male turkeys harvested from private tracts and Wildlife Management Areas during the 1992 and 1993 spring turkey seasons. Differences across years and between land type, age class, and regional zone were not significant. In data pooled across year and zone, differences in variables were not significant for hatching-year individuals. Differences in all variables for after-hatching-year (AHY) individuals were significant (P<0.0003), as was age at harvest. Compared to AHY turkeys harvested from public lands, those harvested from private lands were older (X=8.5 months), weighed significantly more, and had significantly longer beards and spurs. THE Florida Game and Fresh Water Fish Commission (FGFWFC) presently serves as the lead managing agency for public hunting on 1,703,586 ha in 71 Wildlife Management Areas (WMAs). Spring or fall turkey hunting is conducted on 53 of these WMAs. Since 1988, the FGFWFC has quantified the influence of various hunting, wildlife management, and land use conditions in defining quality turkey hunting in Florida (e.g., see Eichholz, 1990). Based upon input from turkey hunters to FGFWFC Wild Turkey Management Section staff, it appeared that a segment of the turkey hunting public differentiated between a high and low quality turkey based on differences in weight, and beard and spur lengths, and preferred to hunt in areas with a high probability of harvesting a mature bird. Hunters appeared to believe that their probability of harvesting a mature, physically large bird was highest on areas where turkey population densities were high, hunter densities were low, and there were restrictions on harvest of hatching-year (HY, i.e., <1 year old) birds. This study was conducted as part of continuing efforts by the FGFWFC to evaluate turkey hunting quality on public and private lands in Florida. Because morphometric quality is a component of hunter satisfaction, the objective in this study was to collect and analyze data to test the statistical hypothesis that there was no difference in major morphometric characteristics (i.e., weight, beard length, and spur lengths) between turkeys harvested on FGFWFC-managed WMAs and those harvested from private lands. METHODS—Data were collected from selected private hunting lessees (n=12, Fig. 1) and WMaAs (n=11) during the 1992 and 1993 spring turkey hunting seasons. A random sample of all hunting leases and all WMAs in Florida was not possible, nor would this sampling approach have resulted in adequate sample sizes because of variations in turkey harvest. Potential private leesees were those involved with management programs administered by FGFWFC Regional Resource Biologists (e.g., No. 3 1994] COBB—MORPHOMETRIC QUALITY OF WILD TURKEYS 89 private lands deer management program). All of these private leesees that allowed turkey hunting and who were willing to participate in data collection were included in the sample. Proximate WMAs that had a spring turkey season were then matched with private leases in a treatment-control design (Fig. 1). Because latitudinal morphometric variation may occur in Florida turkeys (FGFWFC, unpubl. data), the statewide area from which data were to be collected was divided into four zones (Fig. 1) to prevent geographic factors from masking morphometric differences resulting from differential management between public and private lands. Within each sampling zone the number and location of control areas was fixed. Private cooperators and WMA check-station operators were requested to collect data from harvested turkeys to include: date of harvest, sex, age class, weight, beard length, and right- and left-spur lengths. Data were analyzed immediately following each hunting season using PC-SAS Version 6.04 (SAS, 1988) licensed to the FGF WFC. The significance criterion was P=0.01. Analyses were conducted at three levels: 1) each variable for each combination of year/land type/age between zones (n=32), 2) each variable pooled across zones for each combination of land type/age between years (n=16), and 3) each variable pooled across zones and years for both ages between land type (n=8). To test for significance in after-hatching-year (AHY) age at harvest between males harvested on public vs. private lands, individuals in this gross age class were subdivided using criteria of Kelly (1975) and Williams (1981) into three annual age classes based on spur lengths. Individuals were classified as 2, 3, >3 years old when mean right- and left-spur lengths were in the ranges 1.9-2.5, 2.5-3.2, and >3.2 cm, respectively. Data for all variables failed to meet normality assumptions and could not be transformed so analyses were conducted using a nonparametric Wilcoxon Rank Sums Test through PROC NPARIWAY (SAS, 1988). PRIVATE TRACTS * @ WILDLIFE MANAGEMENT AREAS Fic. 1.-Sample sites for collection of morphometric data from turkeys harvested during 1992 and 1993 spring seasons. Stars represent private lands; ovals represent Wildlife Management Areas; numbers represent analysis zones. 90 FLORIDA SCIENTIST [VOL 57 Table 1.-Comparison of morphometric data from male turkeys harvested on public vs. private lands in Florida, 1992 and 1993. Age/Variable* Land Type? Pe Public Private Hatching-year weight 4.9(92) 5.6(15) 0.0233 beard length 9.4(75) 9.8(14) 0.7566 right-spur length 0.7(76) 0.9(15) 0.1678 left-spur length 0.8(76) 0.9(15) 0.1910 After-hatching-year weight 7.1(198) 8.1(129) 0.0001 * beard length 22.2(189) 23.8(130) 0.0001* right-spur length 2.6(189) 2.8(130) 0.0003° left-spur length 2.5(189) 2,.8(130) 0.0002* * Hatching-year and after-hatching-year males were < and > 1 year old when harvested, respectively. > Data are means expressed as kilograms or centimeters (and associated sample sizes). “ * indicates statistical significance in variables between land types. RESULTS AND DISCUSSION—Usable data were collected from 435 tur- keys. The HY and AHY harvest from public and private lands were 92 and 198, and 15 and 130 turkeys, respectively. Most analyses at level 1 (i.e., by year/land type/age between zones) were clearly non-significant (0.05 1 m/s) and lower sediment accumulation were more speciose and exhibited higher coral cover. Stony corals exhibited clumped distributions on nearshore sites, while octocorals were uniformly to randomly distributed in belt quadrats. Temperature fluctuations and sediment transport are inferred to be the most important factors in controlling coral distribution on nearshore communities. Life histories of stony corals on nearshore communities are characterized by small colony size, brooding of larvae, and high recruitment rates. No significant negative associations among coral species were observed, suggesting that competition for substratum availability does not significantly influence observed distribution patterns. Vectorial (physical-chemical) and reproductive factors were inferred to exert the most influence on observed coral abundance patterns at spatial scales investigated. SPATIAL patterning of organisms is an important feature of marine benthic communities. Hutchinson (1953) outlined several factors affecting spatial pattern: (1) reproductive, (2) social, (3) co-active (predator-prey, competition), (4) vectorial (physical-chemical gradients), and (5) stochastic events. The spatial distribution of organisms may therefore be a result of functional relationships among multi-species populations and the relationship of organisms to the environment (Buss, 1986). The arrangement of organisms in ecological communities reflects a history of one or a combination of these factors influencing species. Despite the relative importance of patterning in marine benthos, only recently have investigators focused on identifying the causal factors for shallow-water marine benthic communities (Dana, 1976; Buss and Jackson, 1979; Bradbury and Young, 1983; Reichelt and Bradbury, 1984; Yoshioka and Yoshioka, 1989a; Chadwick, 1991). Shallow-water (< 10 m) marine hard-bottom communities occur throughout the western Atlantic and are an important component of coastal marine ecosystems (Voss and Voss, 1955; Rodriguez, 1959; Macintyre and Pilkey, 1969; Opresko, 1973; Sullivan and Chiappone, 1992). Marine hard-bottom areas encompass different types of communities, including reefal hard-bottoms that differ in the relative abundance of calcifying organisms (hermatypic corals, coralline algae, serpulid worms) (reviewed in Fagerstrom, 1987) and non-reefal hard-bottoms. Hard-bottom No. 3 1994] CHIAPPONE ET AL.—PATTERNS OF CORAL ABUNDANCE 109 communities vary in: (1) distance from shore, (2) depth, (3) geological setting, (4) topographic relief, and (5) dominant biotic components. Although non-reefal hard- bottom communities often encompass large areas of coastal marine ecosystems (Marszalek, 1981; Jaap, 1984), many studies have focused on ecological patterns and processes of high-diversity reef communities. The marine hard-bottom complex along the south Florida shelf represents a mosaic of communities that exhibits extreme variability in all parameters used to evaluate biological communities (Jaap, 1984). Many of the hard-bottom communi- ties in south Florida can be characterized as shallow-water, wave-resistant, three- dimensional carbonate accretions constructed by limestone-secreting organisms on a pre-existing hard substrate (Ginsburg and Shinn, 1964; Hoffmeister and Multer, 1968). Offshore bank-barrier reefs in the Florida Keys are sites of high diversity and biomass of marine flora and fauna, though low-relief hard-bottom communities on the south Florida shelf cover a larger area (Marszalek, 1981). Low-relief hard- bottom communities are characterized by a close proximity to shore ( < 1 km), variable depth ( 1 to 10 m), and are often found adjacent to large tidal passes in the middle Florida Keys. These communities can be colonized by a sparse to dense assemblage of algae, sponges, and corals, and are often dominated by octocorals (e.g. octocoral-dominated hardgrounds, Alcyonaria-sponge communities) (Voss and Voss, 1955; Opresko, 1973). Nearshore hard-bottom communities have also been referred to as reef flats (Stephenson and Stephenson, 1950). The objectives of this field study are to inventory coral species and describe coral distribution patterns on several types of nearshore hard-bottom communities. An initial step in understanding communities is describing spatial patterns. This study included field surveys of low-relief hard-bottom communities and patch reefs at five study sites in the middle Florida Keys representing three discernible hard-bottom types: (1) algal-dominated nearshore hard-bottom, (2) octocoral-dominated nearshore hard-bottom, and (3) coral-dominated patch reef. These three discernible commu- nity types present unique combinations of vectorial and stochastic events as evident in the patterning of corals. MatTERIALS AND METHODS—Survey methods—Natural-color aerial photography of the middle Florida Keys was interpreted to identify the occurrence of shallow-water marine hard-bottom commu- nities and to select specific survey sites (Table 1). Three sites in Florida Bay were chosen, two near Long Key and one near Fiesta Key. One site on the oceanside of Craig Key and one site to the south of Lower Matecumbe Key were also surveyed. The sites are adjacent to Channel 5 and Channel 2 offshore of Long Key. All of the low-relief hard-bottom communities are at least one hectare in size. The patch reef site was the smallest of the communities surveyed ( < 400 m?). The study sites encompassed three main hard- bottom types in the Florida Keys (low-relief hard-bottom and patch reef) and consisted of: (1) algal- dominated hard-bottom (QHS, NTS, FKS), octocoral-dominated hard-bottom (CKS), and coral- dominated patch reef (CGS). These sites are part of a longer term ecological monitoring study initiated in 1989. Data presented in this report were collected from September to November of 1992. Survey sites were selected based on the abundance of hard-bottom communities in the middle Florida Keys area and proximity to the Keys Marine Laboratory on Long Key (Marszalek, 1981). Standardized checklists have been developed for coral species that occur in the middle Florida Keys based on earlier studies of Florida hard-bottom communities (Opresko, 1973; Jaap, 1984; Wheaton and Jaap, 1988). Species presence-absence surveys consisted of a 2-hour haphazard survey of all corals within each site. This qualitative method is designed to inventory coral species over the extent of a community that may not be included in quantitative surveys. Stony corals were identified in situ. Spicule analysis was used to identify octocoral colonies (Bayer, 1961; Cairns, 1977). The overlap of occurrence 110 FLORIDA SCIENTIST [VOL 57 TABLE 1. List of survey sites in the middle Florida Keys from Lower Matecumbe Key to Long Key, Florida. Sampling effort (number of 1 m? quadrats sampled for coral abundance) and depth range (m) provided for sites surveyed from September to November, 1992. Community Sampling Site Location Type LAT/LONG Depth (m) — Effort (m/) QHS Florida Bay Low-relief 24° 48.619'N 1.0-1.5 50 Hard-bottom 80° 50.104' W NTS Florida Bay Low-relief 24° 49.453'N 1.0-1.5 100 Hard-bottom 80° 49.122' W FKS Florida Bay Low-relief 24° 50.544'N 10-15 50 Hard-bottom 80° 47.785' W CKS Nearshore Low-relief 24° 49.752'N 1520 50 Oceanside Hard-bottom 80° 45.784' W CGS Nearshore Channel Patch V4e SONS IN 3.0-5.0 20 Oceanside Reef 80° 43.633' W of species among communities allows adjacent communities to be compared based on species compo- sition. This overlap can be quantified by several indices of community similarity (Hubalek, 1982). Thus, species inventory lists provide semi-quantitative data to characterize a community within a larger ecosystem and compare species occurrences among several survey areas. Belt quadrat sampling of marine benthic communities allows for the collection of detailed information on the spatial patterning of benthos (Dana, 1976; Sullivan and Chiappone, 1992). In this study, 1 m?* quadrats were surveyed contiguously along randomly placed transects. At all sites except CGS, at least two randomly placed 25-m transect lines were used to survey corals. A 20-m transect was used to survey corals at site CGS due to the geometry (20 m x 10 m) of the patch reef community. Preliminary surveys indicated that the number of quadrats surveyed was sufficient from analysis of species-area curves {cumulative number of species versus area (m’) sampled} (Gleason, 1922). Data collected from belt quadrat surveys of hard-bottom communities included colony abundance, size, and area cover. A coral colony was defined as any colony growing independently of its neighbors. In cases where a colony was clearly separated into two or more portions by the death of intervening parts, each living part was considered to be a separate individual (Loya, 1972). When branching colonies occurred in thickets, branches that could be traced to a common origin were considered to be part of a single colony. Planar dimensions of stony coral colonies were measured in situ with calipers to the nearest 0.5 cm. Colony dimensions (length, width, radii) were measured to estimate the planar area cover (cm?) of each stony coral colony. Information on coral colony sizes was used to calculate the area cover in each 1 m? quadrat. Data analyses—Coral species presence-absence information was compared among sites (Q- mode) using the coefficient of Jaccard (Hubalek, 1982; Legendre and Legendre, 1983). Similarity values were used to construct species presence-absence similarity matrices for comparison of species compo- sition among sites. A dendrogram was constructed from cluster analysis of similarity values using a group- average sorting strategy. Similarity between species (R-mode) was calculated using the coefficient of Jaccard with a group-average sorting of values to construct a dendrogram (Pielou, 1977). The purpose of this analysis was to determine whether certain coral species assemblages were characteristic of hard- bottom community types in the middle Florida Keys. Information from belt quadrat sampling included colony abundance (no. m?) and cover (cm? m- 2). Percent relative colony abundance and percent relative area cover values were computed based on each speciescontribution to the total colony numbers and cover of stony corals or octocorals, respec- tively. For relative abundance and cover data, percent similarity (Bray-Curtis Index) values were computed [Similarity P = {minimum value (p,,, p,,)} where P,, and P., are the percentage of colony abundance or area coverage contributed by species i in COnimanntes 1 and 2, respectively] (Pielou, No. 3 1994] CHIAPPONE ET AL.—PATTERNS OF CORAL ABUNDANCE 111 1977). Similarity values for colony density and cover were used to construct percentage similarity matrices and dendrograms using a group average strategy. Spatial patterning analyses and tests for inter-specific association were used as additional tools to analyze abundance pattems of corals across study sites. The method of Malatesta and co-workers (1992) was used to test for scales of aggregation of coral colonies. The patchiness of coral colonies was compared using the sample variance-to-mean ratio for consecutive quadrats {I(V) = s*, / mean,, where V = size of the patch}. Patch sizes from 1 to 5 m? were tested. In cases where aggregation of coral colonies was detected {I(V) > 1}, a randomization procedure was used to test the significance of the aggregation. Given a hypothesized value v for the true-scale aggregation V, the following was tested: H,: significant aggregation at scale v versus; H,: no significant aggregation at scale v. The null hypothesis was rejected if 95 percent of the randomized values exceeded the original value {I(v)}. Tests for inter-specific association among coral species were used following the methods of Ludwig and Reynolds (1988). A 2x2 contingency table was used to test for associations between corals and octocorals. The null hypothesis was that the occurrence of stony coral and octocoral colonies in quadrats was independent. Null hypotheses were tested over a spatial scale of 1 to 5 m? for all sites using a chi- square (X°) test. Additionally, a null association model was used to test the multi-species case of inter- specific association among stony coral and octocoral species occurrences in quadrats. The null hypothesis was that no significant associations existed among coral species in all sites. Species associations were tested using the chi-square (X°) test. ResuLts—A number of patterns were evident from coral abundance data from the five study sites. These patterns are discussed within the framework of each method, the hard-bottom community types sampled, and the spatial scales under consideration: (1) over the entire survey area (50 km”), (2) within sites (< 1 to 2.5 km’), and (3) within belt quadrats (1-100 m/’). Species presence-absence surveys—Twenty-one stony coral species and 16 octocoral species were observed on the five hard-bottom sites (Tables 2 and 3). The algal-dominated hard-bottom communities had the fewest stony corals species (3-5), while the oceanside hard-bottom communities had the greatest number of species (13). Siderastrea radians and Porites porites forma divaricata were common to the Florida Bay communities (QHS, NTS, FKS); these three sites had a subset of the species recorded on the octocoral-dominated hard-bottom community (CKS). No stony coral species were common to all five communities; only two genera, S iderastrea and Porites, were common toall sites. Similarity (coefficient of Jaccard) of stony coral species presence-absence data among sites is illustrated in Figure 2. QHS and NTS were the most similar (75%), with a distinct pattern of decreasing similarity from the 3 algal-dominated hard-bottom communities to the oceanside hard-bottom commu- nities (CKS and CGS). Octocoral species presence-absence exhibited a similar pattern to stony coral species composition (Table 3). The Florida Bay communities (QHS and NTS) had the fewest number of octocoral species (2), while the octocoral-dominated site (CKS) had the greatest number of species (14). Pterogorgia anceps was observed on all sites except the patch reef. No species or genera were common to all five sites. Only species from the Family Briareidae and Gorgoniidae were observed on the 3 Florida Bay communities. Two of the Florida Bay communities (QHS and FKS) were the most similar (50%); the patch reef had no octocoral species in common with the Florida Bay communities (QHS, NTS, and FKS) (Fig. 1). The clustering of species (R-mode) based on coral presence-absence among communities allows for biotic assemblages to be recognized over the spatial scale of 112 FLORIDA SCIENTIST [VOL 57 TABLE 2. Stony coral (Milleporina and Scleractinia) species presence-absence lists for survey sites. Species OHS NTS FKS CKS CGS Acropora cervicornis e Agaricia agaricites e Cladocora arbuscula ° Colpophyllia breveserialia ° C. natans ° Diploria clivosa ° e Favia fragum ° e ° Manicina areolata ° ° @ Meandrina meandrites ° Millepora alcicornis ° ° Montastraea annularis ° M. cavernosa e e Oculina diffusa e Porites astreoides e e P. porites divaricata ° ° e e P. porites furcata ° P. porites porites ° Siderastrea radians ° e ® e S. siderea e Solenastrea bournoni ° S. hyades e e Total Species 3 3 4 13 13 the survey area (50 km’) (Fig. 2). Of the five sites studied, a continuum of four coral assemblages on hard-bottom communities was evident: (1) species observed on both Florida Bay and oceanside hard-bottom communities, (2) species observed only on CKS, (3) species recorded on CKS and CGS, and (5) species observed only on CGS. The stony corals Manicina areolata, Solenastrea hyades, Porites porites forma divaricata, and Siderastrea radians are species that occurred on the Florida Bay No. 3 1994] CHIAPPONE ET AL.—PATTERNS OF CORAL ABUNDANCE 113 TABLE 3. Octocoral species presence-absence lists for survey sites. Octocoral Species QHS NTS FKS CKS CGS Briareum asbestinum e ° Erythropodium caribaeorum 3 Eunicea fusca ° ° E. knighti e E. palmeri @ e E. tourneforti e Gorgonia ventalina e e e Muricea elongata e e e Plexaurella dichotoma ° P. fusifera e Pseudoplexaura flaggelosa e P. porosa e Pseudopterogorgia acerosa e P. americana : P. bipinnata e ° e e Pterogorgia anceps e ° P. guadalupensis Total Species y) 2 4 14 5 hard-bottom communities and octocoral-dominated hard-bottom (CKS). Species strictly confined to the oceanside hard-bottom (CKS) included Solenastrea bournoni, Oculina difusa, and Cladocora arbuscula. Stony coral species that occurred on both the octocoral-dominated hard-bottom and patch reef were Millepora alcicornis, Diploria clivosa, and Montastraea cavernosa. Species documented only at the patch reef were Acropora cervicornis, Agaricia agaricites, Colpophyllia natans, and Meandrina meandrites. In a similar manner, octocorals grouped by genera displayed a distinct pattern among sites (Fig. 2). Species from the genera Briareum, Pterogorgia, and Pseudopterogorgia occurred on all sites except the patch reef. Species from the genera Muricea and Plexaurella were only observed at the octocoral-dominated site (CKS), while species of Eunicea and Pseudoplexaura were only observed on the 114 FLORIDA SCIENTIST [VOL 57 PERCENT SIMILARITY STONY CORALS 100 80 60 40 20 0 OCTOCORALS CGS Fic. 1. Similarity dendrograms of species presence-absence data for study sites. Similarity values calculated using the Jaccard coefficient and clustered using a group-average sorting strategy. Sites include three hard-bottom community types: (1) algal-dominated hard-bottom (QHS, NTS, FKS), (2) octocoral-dominated hard-bottom (CKS), and (3) coral-dominated patch reef (CGS). oceanside communities (CKS and CGS). Gorgonia ventalina and Erythropodium caribaeorum were observed only on the patch reef community (CGS). Belt quadrat surveys—The patterns in similarity for some quantitative at- tributes (cover, colony size) were evident among survey sites. There was no significant trend in stony coral colony abundance among sites. Colony abundance (no. m*) was quite variable among the Florida Bay communities; density was highest (11.7) at QHS and lowest at NTS (0.94) (Table 4). Site QHS had more than twice the colony numbers per sampling unit compared to other sites. Coral cover was greatest (21.8%) at the patch reef and lowest on the Florida Bay communities (0.03-0.6%). Coral cover was greater on the oceanside sites relative to the Florida Bay commu- nities. The size distribution of colonies was similar among the low-relief hard-bottom communities. Sites QHS, NTS, FKS, and CKS had similar size structures of coral colonies; there was a predominance of small ( < 5 cm diameter) colonies on these sites. Mean colony size of stony corals was greatest at CGS. The patch reef site (CGS) had a greater abundance of larger colonies (> 15 cm diameter), a lack of smaller size (< 5 cm diameter) colonies, and greater variability in colony diameter. In terms of relative colony abundance, Siderastrea radians accounted for at least 78 percent of the total stony coral colony abundance among the three Florida Bay hard-bottom communities. Porites porites forma divaricata contributed at least 8.5 percent of the total stony coral abundance at all sites except the patch reef. Similarity No. 3 1994] CHIAPPONE ET AL.—PATTERNS OF CORAL ABUNDANCE LS Stony Coral Species PERCENT SIMILARITY 100 80 60 40 20 C [_$—}—_+—_}+_+— Manicina areolata Porites porites divaricata A Siderastrea radians Solenastrea hyades Oculina diffusa B Cladocora arbuscula Porites porites furcata Solenastrea bournoni Favia fragum GC Montastraea cavernosa Millepora alcicornis Diploria clivosa Porites astreoides Siderastrea siderea Agaricia agaricites Acropora cervicornis D Colpophyllia natans Colpophyllia breveserialis Porites porites Montastraea annularis Meandrina meandrites Octocoral Genera a Pterogorgia A Pseudopterogorgia Briareum B Muricea Plexaurella = Eunicea C Pseudoplexaura Gorgonia D io Erythropodium Fic. 2. Cluster analysis of stony coral species and octocoral genera from study sites. Similarity among species and genera calculated using the Jaccard coefficient (R-mode) and clustered using an group- average sorting strategy. Assemblages occurring on distinct hard-bottom communities are: (A) Florida Bay and nearshore oceanside (QHS, NTS, FKS, CKS), (B) strictly nearshore oceanside (CKS), (C) nearshore oceanside and patch reef (CKS and CGS), and (D) strictly patch reef (CGS). based on species relative abundance was greatest among the three Florida Bay communities; similarity values were 97.3 percent between QHS and NTS and 95.6 percent between NTS and FKS (Fig. 3). The patch reef was the least similar to other sites, reflecting the dominance of reef-building species at this site. The pattern of relative species cover was similar to colony abundance (Fig. 3). 116 FLORIDA SCIENTIST [VOL 57 TaBLE 4. Stony coral (Milleporina and Scleractinia) colony abundance results. Values represent mean (1 SD) numbers of colonies m7. Species QHS NTS FKS CKS CGS Acropora cervicornis 0.45 (1.47) Agaricia agaricites 0.30 (0.57) Diploria clivosa 0.02 (0.15) Favia fragum 0.43 (0.85) 0.05 (0.22) Manicina areolata 0.02 (0.14) Millepora alcicornis 1.11 (1.17) 0.05 (0.22) Montastraea annularis 1.85 (1.37) M. cavernosa 0.10 (0.45) Oculina diffusa 0.11 (0.39) Porites astreoides (0.40 (0.68) P. porites divaricata 0.68 (1.45) 0.08 (0.34) 0.28 (0.83) ~ 0.50 (1.11) Siderastrea radians 11.02 (8.62) 0.86(1.35) 2.02(2.55) 1.68 (1.16) S. sideria 1.15 (1.39) Solenastrea bournoni 0.18 (0.39) S. hyades 0.05 (0.21) Total Species 11.72 (8.63) 0.94(1.37) 2.30(2.73) 4.06(1.74)) sasha) Siderastrea radians accounted for at least 86 percent of the coral cover on the Florida Bay communities, although the absolute abundance of this species differed remark- ably among these sites (Table 5). Coral cover on the octocoral-dominated hard-bottom (CKS) was dominated by Solenastrea bournoni, while cover at the patch reef (CGS) was dominated by S. siderea and Montastraea annularis. Similarity based on relative species cover was greatest among the three Florida Bay communities (Fig. 3). Octocoral colony abundance was low on the Florida Bay sites and high at the octocoral-dominated community (CKS) (Table 6). Site CKS had a significantly higher octocoral colony abundance than all other sites. Similarity based on relative colony abundance was greatest (90.9%) between QHS and FKS (Fig. 3). This high similarity was due to the dominance of Pterogorgia anceps at two of the three Florida Bay communities; this species accounted for at least 87.7 percent of the total octocoral colony abundance at these sites. The patch reef (CGS) was the least similar to the other four communities, reflecting the relative scarcity of octocorals and the No. 3 1994] CHIAPPONE ET AL.—PATTERNS OF CORAL ABUNDANCE 117 Octocoral Species Relative Density PERCENT SIMILARITY 100 80 60 40 20 C Po} + NTS QHS FKS CKS CGS Stony Coral Species Relative Density QHS NTS FKS CKS Stony Coral Species Relative Area Coverage po f+ QHS CGS Fic. 3. Similarity dendrograms of coral abundance data from study sites. Similarity values between sites calculated using the Percent Similarity Index and clustered using a groupaverage sorting strategy. Sites include three hard-bottom community types: (1) algal-dominated hard-bottom (QHS, NTS, FKS), (2) octocoral-dominated hard-bottom (CKS), and (3) coral-dominated patch reef (CGS). different species composition at this site. A final pattern emerged with results from the analyses of abundance patterns and interspecific association within communities (Fig. 4). This analysis attempted to detect changes in the patterning of abundance within a 1-5 m’ scale along belt quadrats. There was a trend toward decreasing patchiness with greater spatial scale, or patch size (m?), for both stony corals and octocorals. Site QHS exhibited the greatest patchiness of stony coral colonies from scales of 1 to 5 m?. The octocoral- dominated hard-bottom (CKS) differed from other communities in exhibiting a random to uniform abundance of stony coral colonies at all patch sizes examined. A slightly different pattern emerged from the analysis of octocoral abundances. Two of the three algal-dominated hard-bottom communities exhibited the greatest 118 FLORIDA SCIENTIST [VOL 57 TABLE 5. Stony coral (Milleporina and Scleractinia) area coverage results. Values represent mean (1 SD) cover (cm?) m”. Species QHS NTS FKS CKS CGS Acropora cervicornis 131.3 (491.6) Agaricia agaricites 8.5 (17.2) Diploria clivosa 0.1 (0.4) Favia fragum 2.0 (4.2) 1.0 (4.4) Manicina areolata 0.1 (0.4) Millepora alcicornis 7.6 (16.3) Montastraea annularis 586.2 (921.2) M. cavernosa 13.4 (60.0) Oculina diffusa 6.8 (23.4) Porites astreoides 65.9 (191.7) P. porites divaricata SAL) 0.3 (1.3) 1.4 (5.8) 13.230) Siderastrea radians 55.5 (48.4) 3.0 (5.2) 9.1 (116) 14.2 (22.6) S. sideria 1371 (2375) Solenastrea bournoni 58.0 (182.6) S. hyades 0.5 (2.4) Total Species 59.0 (47.8) 3.30.3) 10.5 (13.3) 90.4 (188.6) 2182 (2416) degree of patchiness at all spatial scales tested (Fig. 4). Octocoral colonies at other sites were randomly distributed at smaller scales (1-2 m?*) and uniformly distributed at larger scales (3-5 m*). Results from tests for inter-specific association suggested that coral and octocoral species were not significantly dependent (p > 0.05; X° test) at all patch sizes tested (1-5 m?). At the octocoral-dominated community, stony corals and octocorals were found to be positively associated (p < 0.05; X? test). Discussion—Spatial patterning—The spatial patterning of organisms in shal- low-water marine hard-bottom communities has received increasing attention in recent years. In a study of coral populations on a fringing reef in Panama, Dana (1976) observed that species displayed an aggregated distribution on a 10 m? patch scale. Dana (1976) concluded that the observed aggregation of corals may be linked to habitat differences resulting from irregular topography of the reef surface. The creation of micro-habitat differences resulted in the strongest influence on the observed patterning of coral species. Bradbury and Young (1983) analyzed coral No. 3 1994] CHIAPPONE ET AL.—PATTERNS OF CORAL ABUNDANCE 119 TaBLE 6. Octocoral colony abundance results. Values represent mean (1 SD) numbers of colonies m”. Species QHS NTS FKS CKS CGS Briareum asbestinum 0.02 (0.93) 0.04 (0.21) 0.72 (0,89) Eunicea palmeri 1.44 (0.86) Gorgonia ventalina 0.10 (0.31) Muricea elongata 0.22 (0.43) Plexaurella dichotoma 0.06 (0.24) Pseudoplexaura flagellosa 0.39 (0.61) P. porosa 1.11 (0.96) Pseudopterogorgia acerosa 0.06 (0.24) P. americana 0.02 (0.14) 0.13(0.40) 0.11 (0.32) P. bipinnata 0.06 (0.24) Pterogorgia anceps 0.20 (0.93) 1.24 (1.54) 0.17 (0.38) Total 02221093) C0204) 5 WA GrS2)) 42339), ON0(O381) interactions and community structure of reefs on two spatial scales. Results indicated that at a smaller scale, there was a random mingling of species along sampled transects. At a larger scale, a spatial pattern of species zonation along sampled transects was evident. In a similar study, Reichelt and Bradbury (1984) investigated two ecological scales of reef benthos: (1) the local distribution of corals (neighbor- hood patterns) and (2) large-scale patterns (among sites) of distribution. Results indicated that at the local scale, coral species and colonies were randomly distrib- uted. Interactions on the smaller scale could not explain larger-scale patterns of benthos distribution. Yoshioka and Yoshioka (1989a) investigated abundance patterns of gorgonians on shallow reefs of Puerto Rico. This study employed a multi-scale analysis of pattern with results indicating that gorgonian colonies occurred at patch sizes (mosaic scales) of 0.25 and 2 m”’. Species were found to be positively associated at patch sizes less than 2 m’. No instance of negative association among gorgonian species was observed, suggesting that competition for spatial resources may not play a critical role in the ecology of shallow-water gorgonian communities. Several patterns emerged from this study of five hard-bottom communities in the middle Florida Keys: (1) distinct species assemblages among sites, (2) low cover and small colony size on nearshore communities, (3) distinct relative abundance patterns of corals, and (4) an absence of negative associations among coral species. Results from spatial patterning analysis on a small scale (1-5 m?) indicated that stony 120 FLORIDA SCIENTIST Stony Corals 8 ¢ 6 {49} ( Patch Size (m2) Fic. 4. Spatial pattern analysis of stony corals and octocorals. Values represent sample vari- ance (s*) to mean ratios of colony abundance for different patch sizes (m?). M: QHS. L: NTS. v: FKS. # :CKS. X: CGS. Sites include three hard- bottom community types: (1) algal-dominated hard-bottom (QHS, NTS, FKS), (2) octocoral- dominated hard-bottom (CKS), and (3) coral- dominated patch reef (CGS). [VOL 57 corals tended to be patchy at 1 m*at four of the five sites. Coral colonies were randomly to uniformly distributed at a scale of 1-5 m? at the octocoral-domi- nated hard-bottom community (CKS). Octocorals were uniformly to randomly distributed on the oceanside hard-bot- tom communities. Similar to results from Yoshioka and Yoshioka (1989a), no negative asso- ciations among stony coral and octocoral species were observed. Patch sizes of stony corals were related to: (1) micro- scale ( < 1 m) topographic relief of the substratum and (2) spacing of sand- mud and Thalassia testudinum seagrass patches. Octocorals displayed different spatial patterns among sites. At QHS and FKS, octocoral colonies were patchy, but at NTS, colonies were uni- formly distributed and extremely sparse. Octocoral patterning at CKS was uni- form to random, which may be a func- tion of the size of colonies and the maximum packing of individuals on the substratum. Competition for available substratum was apparently not a limit- ing factor in the pattern of corals among the four nearshore, low-relief hard-bot- tom communities. In contrast, octo- corals at CGS were randomly distrib- uted and low in abundance, possibly reflecting the limited amount of avail- able substrata due to the high abun- dance of algae, sponges, and stony cor- als at this site. Physical influences on hard-bottom communities—Based on results from this study and other investigations of factors affecting coral distribution in south Florida, we hypothesize that nearshore hard-bottom communities in the middle Florida Keys are subjected to two principal physical forces: (1) sediment transport and deposition and (2) temperature fluctua- tions (Roberts et al., 1982; Shinn et al., 1989). The region of the Florida Keys can be divided into three areas: upper Keys, middle Keys, and lower Keys. These divisions are based on a geological perspective and the distribution of hard-bottom commu- No. 3 1994] CHIAPPONE ET AL.—PATTERNS OF CORAL ABUNDANCE 121 TaBLe 7. Life history characteristics of selected scleractinian corals from study sites. Sexual mode refers to H (hermaphroditic) or G (gononchoric). Larval development refers to E (external) or B (brooding). Sexual Larval Colony — Recruitment Species Mode Development Size Rate Sites*® Acropora cervicornis H E large low Ill Diploria clivosa H E large low Lie i Montastraea annularis H E large low Ill M. cavernosa G E large low II, Il Porites astrevides lal E small high II, Ul Siderastrea siderea G E large low Ill Agaricia agaricites G B small high Ill Favia fragum H B small high I, Il, Il Manicina areolata H B small high L,I P. porites**° (E B small low Lei, Wt S. radians (@ B small high Iain Site® designations are: I — Florida Bay hard-bottom communities, I] — oceanside hard-bottom communities, and III — patch reef communities. ¢*Also see ney and Peters (1987). °*° Information on P. porites includes data for three forma: porites, furcata, and divaricata. Information on life histories from Bak and Engel (1979), Fadlallah (1983), Rylaarsdam (1983), Van Moorsel (1983), Szmant (1986), Chornesky and Peters (1987), and Soong (1991). nities in the region (Marszalek et al., 1977; Jaap, 1984; Shinn et al., 1989). Ginsburg and Shinn (1964) determined that active three-dimensional reefs in south Florida and the northern Bahamas are restricted to the seaward side of islands along eastern- facing carbonate banks. The large tidal passes separating the Pleistocene islands in the middle Florida Keys allows sediment drainage from Florida Bay, which actively produces sediment from both biogenic and lithogenic processes. Relative to the upper and lower Florida Keys, hard-bottom communities in the middle Florida Keys are subjected to considerable turbidity corresponding to the semi-diurnal tidal cycle (Shinn et al., 1989). The effect of sediment transport on hard-bottom community structure is evident in the sparse occurrence of channel patch reefs and offshore ( > 5 km) bank-barrier reefs in the middle Florida Keys (Marszalek et al., 1977). Large tidal channels influence the distribution of nearshore (< 1 km), low-relief hard-bottom communities, which are most abundant in the middle Keys relative to other areas on the south Florida shelf. Thus, sediment transport and the presence of large tidal passes: (1) prevent active reef growth except on the seaward side of islands where communities are relatively buffered from 122 FLORIDA SCIENTIST [VOL 57 sediment deposition and turbidity and (2) scour the limestone substratum adjacent to tidal passes, thus enhancing the abundance of organisms capable of withstanding the current regime and sediment abrasion. Sediment transport may effect hard- bottom fauna as a result of sediment deposition (sediment depth), turbidity, and sediment abrasion associated with tidal currents (Opresko, 1973; Yoshioka and Yoshioka, 1989b; Rice and Hunter, 1992). The second physical force, temperature, is an important factor in considering high latitude reef systems such as Bermuda, south Florida, the northwestern Gulf of Mexico, and the northern Bahamas (Goldberg, 1973b; Opresko, 1973; Burns, 1985). The shallow bays of south Florida can exhibit hypothermal and hyperthermal conditions depending on seasonal factors (i.e. wind speed, air temperature). During periods of low wind velocity and high temperature, water in Florida Bay can become hyperthermal and hypersaline (Jaap, 1984). During the passage of cold, polar continental air masses over the south Florida shelf, water in Florida Bay may become hypothermal. Rapid cooling of shallow water is believed to occur because of the lower heat storage capacity relative to offshore water masses in the Florida Keys (Roberts et al., 1982). The implications of temperature as a factor influencing nearshore hard-bottom communities are two-fold: (1) the ability of benthic organisms to respond to rapid temperature changes and (2) the temporal extent of the thermal stress. Hudson and co-workers (1976) established a correlation between the mass mortality of her- matypic corals in south Florida and the occurrence of unusually cold water associ- ated with a cold front. In a later study, Hudson (1981) transplanted colonies of Montastraea annularis from offshore to nearshore sites adjacent to a tidal creek in the upper Florida Keys. Results indicated that colonies placed nearest to shore were most affected by thermal stress and that the nearshore limit of M. annularis was 2 km from shore. In a study of shallow-water reef corals at Dry Tortugas, Porter and co- workers (1982) documented a 96 percent mortality of Acropora cervicornis due to unusually severe winter temperatures. Our results support the contention that sediment transport and temperature are factors affecting coral distribution patterns in the middle Florida Keys. This was evident in the lack of major hermatypic fauna and small colony sizes on survey sites. Sites characterized by higher sediment coverage and variable turbidity were dominated by one or two coral species. At CKS, where sediment coverage was low and current velocities were greatest, octocoral abundances were high. Life history strategies of hard-bottom benthos—The physical forces on natural communities are reflected in the differential life histories of organisms. Table 6 presents information on the reproductive ecology and growth characteristics of selected scleractinians observed during this study. The most evident pattern that emerged from this analysis was that algal-dominated hard-bottom communities were dominated by stony corals that: (1) brood larvae, (2) attain a small colony size, and (3) have high, but variable recruitment rates. In contrast, the patch reef was dominated by stony coral species that broadcast spawn, attain large colony sizes, and have low reported recruitment rates (Bak and Engel, 1979; Szmant, 1986; Soong, 1991). Life history patterns of stony corals at the octocoral-dominated hard-bottom No. 3 1994] CHIAPPONE ET AL.—PATTERNS OF CORAL ABUNDANCE 123 community were intermediate between those observed on algal-dominated and coral-dominated communities, suggesting that both stony corals and octocorals occur along a continuum rather than in discrete communities along an environmen- tal gradient (see Yoshioka and Yoshioka, 1989b). Patterns in the life history of selected corals observed on study sites have been well-documented (Lewis, 1989). Kissling (1965), in a study of coral distribution on a nearshore shoal in the lower Florida Keys, documented only four scleractinians at a depth of 1-2 m, including Siderastrea radians, Favia fragum, Porites porites, and P. astreoides. Similarly, Zischke (1973) surveyed an alcyonarian-sponge community offshore of Pigeon Key (middle Florida Keys) and recorded four stony coral species and eleven octocoral species. In a survey of relict reefs located off eastern Florida, Goldberg (1973b) observed only one species of stony coral (S. radians) and one genus of octocoral (Pseudopterogorgia) at 1.5 m depth. These studies and other investigations provide further evidence that certain corals are able to withstand extreme conditions on some nearshore hard-bottom communities (Stephenson and Stephenson, 1950). Although scarce information exists on the life histories of octocorals (Goldberg, 1973a; Brazeau and Lasker, 1990; Lasker, 1983), a few points are worth noting concerning the species distributions and abundance patterns observed during this study. Colony morphology is believed to play a critical role in the ability of octocorals to colonize hard-bottom substrata. The absence of octocorals from the Family Plexauridae on the Florida Bay sites may be a result of high sediment transport or temperature fluctuations. The reason for the high diversity and abundance of octocorals on CKS relative to the Florida Bay sites remains to be tested. The colony morphology of two genera, Pseudopterogorgia and Pterogorgia, may be related to sediment transport processes. This is evident in the arrangement of polyps along the colony axis, which reduces the area of exposed tissue to sediment particles in the water column. Species of Muricea and Eunicea observed at CKS have well- developed calicular lips which may prevent sediment abrasion (Bayer, 1961; Opresko, 1973). Although differences in temperature fluctuation among sites were not measured, a combination of available substrata, sediment abrasion, and thermal stress may limit the ability of most octocoral taxa to settle on Florida Bay hard-bottom communities. The results from this investigation are similar to previous studies of spatial patterning of marine organisms on hard-bottom communities (Opresko, 1973; Yoshioka and Yoshioka, 1989a) . Opresko (1973) documented patterning evident in the distribution and abundance of octocoral species along an environmental gradi- ent. Results from this study suggest that the middle Florida Keys are subjected to intense environmental forces that result in distinct biotic assemblages over short (km) spatial scales. These patterns are important in comparing other ecosystems and understanding causal mechanisms of dynamics on nearshore hard-bottom commu- nities. ACKNOWLEDGMENTS—The authors gratefully acknowledge field support from Mr. Jose Levy, Mr. Robert Sluka, Mr. Gabe Delgado, Ms. Nancy Black, and Ms. 124 FLORIDA SCIENTIST [VOL 57 Yvette Paroz. The authors thank The Nature Conservancy Florida Keys Initiative and the University of Miami Marine Science Program for technical support. Special thanks to the staff of the Keys Marine Lab on Long Key, Florida and the Florida Institute of Oceanography for field and boat support. Sea and Sky Foundation provided aerial photographs of survey sites. LITERATURE CITED Bak, R.P.M. AND M.S. ENGEL. 1979. Distribution, abundance and survival of juvenile hermatypic corals (Scleractinia) and the importance of life history strategies in the parent coral community. Mar. _ Biol. 54:341-352. Bayer, F.M. 1961. The Shallow-water Octocorallia of the West Indian Region. Stud. Fauna Curacao Carib. Isla. 55:1-373. Brapsuky, R.H. anp P.C. Younc. 1983. Coral interactions and community structure: an analysis of spatial pattern. Mar. Ecol. Prog. Ser. 11:265-271. Brazeau, D.A. AND H.R. Lasker. 1990. Sexual reproduction and external brooding by the Caribbean gorgonian Briareum asbestinum. Mar. Biol. 104:465-474. Burns, T.P. 1985. Hard-coral distribution and cold-water disturbances in South Florida: variation with depth and location. Coral Reefs 4:117-124. Buss, L.W. 1986. Competition and community organization on hard surfaces in the sea. Pp. 517-536 In: Community Ecology. Diamonp, J., AND T.J. Case (eds.). Harper and Row, New York. AND J.B.C. Jackson. 1979. Competitive networks: Nontransitive competitive relation- ships in cryptic coral reef environments. Amer. Nat. 113(2):223-234. Cairns, S. 1977. Guide to the Commoner Shallow-water Gorgonians (Sea Whips, Sea Feathers and Sea Fans) of Florida, the Gulf of Mexico, and the Caribbean region. Sea Grant Field Guide Series Number 6, University of Miami Sea Grant Program. Coral Gables, FL. 74 pp. Cuapwick, N.E. 1991. Spatial distribution and the effects of competition on some temperate Scleractinia and Corallimorpharia. Mar. Ecol. Prog. Ser. 70:39-48. CuorNEsky, E.A. AND E.C. Peters. 1987. Sexual reproduction and colony growth in the scleractinian coral Porites astreoides. Biol. Bull. 172:161-177. Dana, T.F. 1976. Reef-coral dispersion patterns and environmental variables on a Caribbean coral reef. Bull. Mar. Sci. 26(1):1-13. ; FaDLALLaH, Y.H. 1983. Sexual reproduction, development and larval biology in scleractinian corals. A review. Coral Reefs 2:129-150. FaGeErstTRoM, J.A. 1987. The Evolution of Reef Communities. John Wiley & Sons, Inc., New York. Ginspurc, R.N. AND E.A. SHINN. 1964. Distribution of the reef-building community in Florida and the Bahamas. Amer. Assoc. Petr. Geol. Bull. 48:527. GueEason, H.A. 1922. On the relation between species and area. Ecology 3:158-162. GoLbBERrG, W.M. 1973a. Ecological aspects of salinity and temperature tolerances of some reef-dwelling gorgonians from Florida. Carib. J. Sci. 13:173-177. . 1973b. The ecology of the coral-octocoral communities off the southeast Florida coast: Geomorphology, species composition, and zonation. Bull. Mar. Sci. 23(3):465487. HOFFMEISTER, J.E. AND H.G. MuLTER. 1968. Geology and origin of the Florida Keys. Geol. Soc. Amer. Bull. 79:1487-1502. Husa kk, Z. 1982. Coefficients of association and similarity based on binary (presence/absence) data: An evaluation. Biol. Rev. 57:669-689. Hupson, J.H. 1981. Response of Montastrea annularis to environmental change in the Florida Keys. Pp. 233-240 In: GoMEz, E.D., BIRKELAND, C.E., BUDDEMEIR, R.W., JOHANNES, R.E., MARSH, J.A., AND R.T. Tsupa (eds.). Proc. Fourth Intern. Coral Reef Symposium, Vol. 2. Quezon City, Manila, Philippines. , E.A. SHINN, R.B. HALLey, AND B. Lipz. 1976. Sclerochronology: a tool for interpreting past environments. Geology 4:361-364. Hutcuinson, G.E. 1953. The concept of pattern in ecology. Proc. Acad. Nat. Sci. Phila. 105: 1-12. Jaap, W.C. 1984. The Ecology of South Florida Coral Reefs: A community profile. U.S. Fish and Wildlife Service, FWS/OBS-82/08, Washington, D.C. 138 pp. KissLinc, D.L. 1965. Coral distribution on a shoal in Spanish Harbor, Florida Keys. Bull. Mar. Sci. 15(3):599-611. Lasker, H.R. 1983. Vegetative reproduction in the octocoral Briareum asbestinum (Pallas). J. Exp. Mar. No. 3 1994] CHIAPPONE ET AL.—PATTERNS OF CORAL ABUNDANCE 125 Biol. Ecol. 72:157-169. LEGENDRE, L. AND P. LEGENDRE. 1983. Numerical Ecology. Elsevier, New York. Lewis, J.B. 1989. Spherical growth in the Caribbean coral Siderastrea radians (Pallas) and its survival in disturbed habitats. Coral Reefs 7:161-167. Lupwic, J.A. AND J.F. REYNOLDs. 1988. Statistical Ecology: A primer on methods and computing. John Wiley & Sons, Inc., New York. Loya, Y 1972. Community structure and species diversity of hermatypic corals at Eilat, Red Sea. Mar. Biol. 13: 100-123 ManinTyre, I.G. anp O. H. Pivkey. 1969. Tropical reef corals: Tolerance of low temperatures on the North Carolina continental shelf. Science 166:374-375. Ma.atesta, R.J., P.J. AUSTER, AND B.P. Caruin. 1992. Analysis of transect data for microhabitat correlations and faunal patchiness. Mar. Ecol. Prog. Ser. 87:189-195. Marsza.kK, D.S. 1981. Florida Reef Tract Marine Habitat and Ecosystem (Map Series). Published in cooperation with the Florida Department of Natural Resources and the U.S Department of the Interior-Bureau of Land Management. Rosenstiel School of Marine and Atmospheric Science, University of Miami, FL. , G. Bapasnorr, JR., M.R. Nok, AND D.R. Wortey. 1977. Reef distribution in south Florida. Pp. 223-229 In: Proceedings of the Third International Coral Reef Symposium, Vol. 2. Miami, FL. Opresko, D.M. 1973. Abundance and distribution of shallow-water gorgonians in the area of Miami, Florida. Bull. Mar. Sci. 23(3):535-557. Pirtou, E.C. 1977. Mathematical Ecology. John Wiley & Sons, Inc., New York. Porter, J.W., J.F. Batrey, aND G.J. Situ, JR. 1982. Perturbation and change in coral reef communities. Proc. Natl. Acad. Sci. 79:1678-1681. REICHELT, R.E. aND R.H. BrabBury. 1984. Spatial patterns in coral reef benthos: multiscale analysis of sites from three oceans. Mar. Ecol. Prog. Ser. 17:251-257. 23 Rice, S.A. anDC.L. Hunter. 1992. Effects of suspended sediment and burial on scleractinian corals from west central Florida patch reefs. Bull. Mar. Sci. 51(3):429-442. Roserts, H.H., L.J. Rouse, JR., N.D. WALKER, AND J.H. Hupson. 1982. Cold-water stress in Florida Bay and northern Bahamas: A product of winter cold-air outbreaks. J. Sed. Petr. 52(1):145-155. RobriGuEz, G. 1959. The marine communities of Margarita Island, Venezuela. Bull. Mar. Sci. Gulf. Carib. 9(3):237-280. RyaarsbaM, K. 1983. Life histories and abundance patterns of colonial corals on Jamaican reefs. Mar. Ecol. Prog. Ser. 13:249-260. SiNN, E.A., B.H. Lipz, R.B. Hau.ey, JH. Hupson, anv J.L. Kinpincer. 1989. Reefs of Florida and the Dry Tortugas. Field Trip Guidebook T176, American Geophysical Union, Washington, D.C. Soonc, K. 1991. Sexual reproductive patterns of shallow-water reef corals in Panama. Bull. Mar. Sci. 49(3):832-846. STEPHENSON, T.A. AND A. STEPHENSON. 1950. Life between tide-marks in North America. I. The Florida Keys. J. Ecol. 38:354-402. SULLIVAN, K.M. anb M. CiliAprpoNne. 1992. A comparison of belt quadrat and species presence/absence sampling of stony coral (Scleractinia and Milleporina) and sponges for evaluating species patterning on patch reefs of the central Bahamas. Bull. Mar. Sci. 50(3):464-488. SzMaNT, A.M. 1986. Reproductive ecology of Caribbean reef corals. Coral Reefs 5:43-53. VanmoorsEL, G.W.N.M. 1983. Reproductive strategies in two closely related stony corals (Agaricia, Scleractinia). Mar. Ecol. Prog. Ser. 13:273-283. Voss, G.L. AND N.A. Voss. 1955. An ecological survey of Soldier Key, Biscayne Bay, Florida. Bull. Mar. Sci. Gulf Carib. 5(3):203-229. WuatTon, J.L. anDW.t. Jaap. 1988. Corals and other prominent benthic Cnidaria of Looe Key National Marine Sanctuary, Florida. Fla. Mar. Res. Publ. 43:1-25. Yosioka, P.M. AND B.B. Yostioka. 1989a. A multispecies, multiscale analysis of spatial pattern and its application to a shallow-water gorgonian community. Mar. Ecol. Prog. Ser. 54:257-264. AND B.B. Yosiioka. 1989b. Effects of wave energy, topographic relief and sediment transport on the distribution of shallow-water gorgonians of Puerto Rico. Coral Reefs 8:145-152. Ziscukk, J.-A. 1973. An Ecological Guide to the Shallow-water Marine Communities of Pigeon Key, Florida. St. Olaf College, Northfield, MN. Florida Scient. 57(3):108-125.1994. Accepted: April 20, 1994. 126 FLORIDA SCIENTIST [VOL 57 FLORIDA ACADEMY OF SCIENCES 1994 ANNUAL MEETING OUTSTANDING STUDENT PAPER AWARDS AGRICULTURAL SCIENCES Martin Chiona, Agronomy Department, University of Florida, “Corn Yield Prediction from Ear Leaf Measurements.” Graduate Co-Award. Serge Edme, Agronomy Department, Univ of Florida, “Photosensitivity in a Set of Experimental Tropical Corn Synthetics.” Graduate Co-Award. Xiamping Qu, Division of Agriculture, Florida A&M University, “Testing Genetic Variation among Bunch and Muscadine Grapes Using RAPD Markers.” Graduate Co-Award. ANTHROPOLOGICAL SCIENCES Susan Stans, Department of Anthropology, University of Florida, “Variation, Perceptual Differ- ences, and Tanning Ranges in Reflectance Spectrophotometer Readings of Florida Undergraduate Students. Graduate Award. BIOLOGICAL SCIENCES Ginny Eckert, Department of Zoology, University of Florida, “A Novel Larval Feeding Strategy for the Tropical Sand Dollar (Encope michelini).”. Graduate Co-Award. Stephen Kajiura, Department of Biological Sciences, Florida Institute of Technology, “Sexual Heterodonty Correlates with Mating Season in the Atlantic Stingray (Dasyatis sabina).” Graduate Co-Award. James Burch, Department of Biological Sciences, Florida International University, “Love Vine Parasitism of Brazilian Pepper in Southern Florida.” Graduate Honorable Mention. Elizabeth Garland, Department of Environmental Engineering Sciences, University of Florida, “A New Microbial Test for the Specific Determination of Solids-Associated Heavy Metal Toxicity.” Graduate Honorable Mention. Don Spence, Biology Department, Stetson University, “What is the Adaptive Significance of Extracellular Accumulations of Mucilage in Watershield (Brasenia scherberi, Gmelin)?” Undergradu- ate Co-Award. Lianne Bishop, Biology Department, Stetson University, “Sexual Dimorphism in the Rattlesnake (Sistrurus miliarius barbouri).” Undergraduate Co-Award. Karen Davis, Department of Biology, University of Tampa, “Species Composition and Spatial Distribution of Hermit Crabs in Tampa Bay, Florida, and Surrounding Waters.” Undergraduate Honorable Mention. COMPUTER SCIENCE AND MATHEMATICS A.B. Vafaie, Operations Research Program, Florida Institute of Technology, “Optimal Periodic Testing for Stand-by Systems.” Graduate Award. ENVIRONMENTAL CHEMISTRY AND CHEMICAL SCIENCES Charles D. Norris, Department of Chemistry, University of South Florida, “Studies of Metal Extraction from Environmentally Significant Media Using Supported Chelating Agents.” Graduate Award. Reza Mahram, Department of Chemistry, University of South Florida, “Effects of Microwave Irradiation (2450 MHZ CW) on Ptychodiscus brevis and Nannochloris oculata Interactions.” Under- graduate Award. (GEOLOGICAL AND HYDROLOGICAL SCIENCES Kendall Fountain, Department of Geology, University of Florida, “Fe-P Cycling as a Principle Digenetic Mechanism in the Evolution of West Florida Shelf Phosphorite.” Graduate Co-Award. Rich Copeland, Department of Geology, Florida State University, “A Solution of the Solute Transport Ground-Water Flow Problem Using Time Series Analysis.” Graduate Co-Award. MEDICAL SCIENCES Patricia Cruz, Department of Molecular Biology and Microbiology, University of Central Florida, “Isolation and Classification of Airborne Molds and their Medical Significance.” Graduate Co-Award. Nydia Rodriguez, Department of Molecular Biology and Microbiology, University of Central Florida, “Isolation and Purification of the Major Allergenic Components from Planet Pollens.” No. 3 1994] 127 Graduate Co-Award. Noel Alonso, School of Natural and Health Sciences, Barry University, “Ryanodine and Verapamil: Depletion of Muscle Energy Stores.” Graduate Co-Award. RARE AND ENDANGERED BioTA (FLORIDA COMMITTEE ON RARE AND ENDANGERED PLANTS AND ANIMALS) Christine Small, Department of Biology, University of Central Florida, “The Effects of Reintro- duction on Reproduction of Gopher Tortoises (Gopherus polyphemus).” Graduate Award. Joseph Helkowski, Department of Biology, Stetson University, “The Effects of Disturbance on Growth, Flowering, and Fruiting in the Endangered Pawpaw (Deeringothamnus rugelii).” Under- graduate Award. SCIENCE TEACHING Mary Beth Potami, Valencia Community College, “Microscale Organic Chemistry Labs: A Student’s Evaluation.” Undergraduate Co-Award. Lisa Santiago, University of Central Florida, “Chemical Management and Disposal.” Undergradu- ate Co-Award. Thomas Bayes, Brevard Community College, “Enhancement of Science Education and Awareness by Voluntary Activities.” Undergraduate Co-Award. URBAN AND REGIONAL TEACHING Guy McGrane, University of Florida, “Personal Transportation Alternatives: An EMERGY Evaluation.” Graduate Award. SPECIAL AWARDS AMERICAN ASSOCIATION OF THE ADVANCEMENT OF SCIENCE AWARD UNDERGRADUATE AWARDEES: Karen Davis, Department of Biology, University of Tampa, “Species Composition and Spatial Distribution of Hermit Crabs in Tampa Bay, Florida, and Surrounding Waters.” Female Nominee. Joseph Helkowski, Department of Biology, Stetson University, “The Effects of Disturbance on Growth, Flowering, and Fruiting in the Endangered Pawpaw (Deeringothamnus rugelii).” Male Nominee. CENTRAL FLORIDA CHAPTER, THE EXPLORERS CLUB AWARD James Burch, Department of Biological Sciences, Florida International University, “Introduction to the Ethnobotany of the Chachi in an Ecuadorian Rain Forest.” FLORIDA INSTITUTE OF TECHNOLOGY CHAPTER, SIGMA XI AWARD Stephen Kajiura, Department of Biological Sciences, Florida Institute of Technology, “Sexual Heterodonty Correlates with Mating Season in the Atlantic Stingray (Dasyatis sabina).” UNIVERSITY OF FLORIDA CHAPTER, SIGMA XI AWARD Kendall Fountain, Department of Geology, University of Florida, “Fe-P Cycling as a Principle Digenetic Mechanism in the Evolution of West Florida Shelf Phosphorite.” VICE ADMIRAL WILLIAM W. BEHRENS, JR. / FLORIDA INSTITUTE OF OCEANOGRAPHY AWARD Ginny Eckert, Department of Zoology, University of Florida, “A Novel Larval Feeding Strategy for the Tropical Sand Dollar (Encope michelini).” ENVIRONMENTAL CHEMISTRY AND CHEMICAL SCIENCES AWARD Charles D. Norris, Department of Chemistry, University of South Florida, “Studies of Metal Extraction from Environmentally Significant Media Using Supported Chelating Agents.” J.-M. MonTcoMEry ENGINEERING AwarD (WATER RELATED GEOLOGY AND HyDROL- ocy PaPER) Lorene Whitecross, Department of Geology, Florida State University, “The Benefits of Cave Diver Assisted Water Sampling at Wakulla Springs, Florida.” SPIKOWSKI PLANNING ASSOCIATES AWARD Guy McGrane, University of Florida, “Personal Transportation Alternatives: An EMERGY Evaluation.” — Carl A. Luer, Chairman, Awards Committee Mote Marine Laboratory, Sarasota, FL 128 FLORIDA SCIENTIST [VOL 57 BOOK REVIEW F.A. McClure, The Bamboos, Smithsonian Institution Press, Washington and London. 1993. Pp. xxii + 345. Price: 16.95. IT is a delight to have this classic back on bookstore shelves. First published in 1966, no other work on the bamboos so completely covers the diversity and natural history of these fascinating giant grasses. Not only is this book meticulous and thorough, but it is also infused with the excitement and enthusiasm of someone who was not only a botanist but also a plant-lover. Such people and such books are rare in today’s publish-or-perish world of high-tech biology. The text is well-written and interesting, but I particularly like the numerous clear line drawings that illustrate important details of bamboo structure in their great taxonomic variety. One has here a well-illustrated text on bamboo morphology, which if studied thoroughly could enable one to go out into the field and become an expert bamboo taxonomist. There is also much information on the propagation, cultivation and utilization for bamboos, making this in addition a practical handbook for someone entering into the commercial exploitation of these plants. The key provided to the genera of bamboos cultivated in the United States is very useful, but one wishes that author could have employed vegetative characters exclusively in the key, as bamboos flower so infrequently. References to floral structures in the key become serious stumbling blocks for most field use of the key. One might also wish fora key to all bamboo genera, that could be carried into the field in all parts of the earth, but it is clear that on both counts, McClure gave us everything he had. We can only hope that the republication of this book will inspire a new generation of bamboo botanists who will fill in the remaining gaps and provide us with the complete treatment that McClure undoubtedly dreamed of —Frederick B. Essig, Department of Biology, University of South Florida, Tampa, FL 33620. CORRIGENDUM ProrFit, C. E., K. M. Jouns, C. B. Cocurang, D. J. DEVLIN, T. A. REYNOLDs, D. L. PaYNE, S. JEPPESEN, D. W. PEEL, AND D. D. Linpen. 1993. Field and laboratory experiments of the consumption of mangrove leaf litter by the macro detritivore Melampus coffeus L. (Gastropoda:pulmonata). Florida Scient. 56 (4): 221-222. The first line on page 214 should read: “ranged from 99.4 (45.8) to 130.2...” We regret any inconvenience occasioned by this error. INSTRUCTIONS TO AUTHORS Individuals who publish in the Florida Scientist must be active members in the Florida Academy of Sciences. Submit a typewritten original and two copies of the text, illustrations, and tables. All typewritten material—including the abstract, literature citations, footnotes, tables, and figure legends—shall be double-spaced. Use one side of 8'/2 x 1] inch (21'/2 cm x 28 cm) good quality bond paper for the original; the copy may be xeroxed. Margins should be at least 3 cm all around. Number the pages through the Literature Cited section. Avoid footnotes and do not use mimeo, slick, erasable, or ruled paper. Use metric units for all measurements. Assistance with production costs will be negotiated directly with authors of pages which exceed 10 printed pages. Excess over 10 pages by an author in a single volume will result in charges for additional pages. ApppreEss follows the author’s address. ApsTRACT—AIll manuscripts shall have a short, concise, single-paragraphed abstract. The abstract follows immediately the author's address. ACKNOWLEDGMENTS are given in the body of the text preceding immediately the Literature Cited section. LITERATURE CITED section follows the text. Double-space every line and follow the format in the current issue. Manuscrirts of 5 or less, double-spaced typewritten pages should conform to the short article protocol. See recent issue for proper format of both long and short articles. TABLES shall be typed on separate sheets of paper, double-spaced throughout. Each table shall contain a short heading. Do not use vertical rulings. Maximum character width, including spaces, of tables is 98.0. This includes a minimum space of 5 characters between columns. Tables are charged to authors at $13 per quarter page, $29 per half page, and $39 per full page. ILLustTRATIONS—AIll drawings shall be done in good quality India ink, on good board or drafting paper. Letter by using a lettering guide or equivalent. Typewritten letters on illustrations are unaccept- able. Drawings and photographs should be large enough to allow 1/3 to 1/2 reduction in size. Photographs shall be glossy prints of good contrast. Whenever possible, mount photographs in lots size. The author’s name and figure number should be penciled lightly on the back of each figure. Figure legends must be listed on a separate page and not on the drawing or photograph. The legend must be double spaced. Illustrations are charged to authors at $6.00 per page or fraction. We ask authors to prepare accepted manuscripts in personal computer formats. Please do not submit a manuscript on disk. Once reviewers have recommended a manuscript be accepted, you will be given directions for submitting the final manuscript on disk. PRoor must be returned promptly. Notification of address change and proofreading are the author’s responsibility. Alterations after the type has been set will be charged to the author. REPRINTS may be ordered through the FAS office from the printer on forms provided at the time that proofs are sent to authors for checking. FLORIDA ACADEMY OF SCIENCES INSTITUTIONAL MEMBERS FOR 1994 Archbold Biological Station St. John’s River Community College, Palatka Brevard Community College, Florida Advanced _ St. Petersburg (Florida) Public Library Technology Center Library, Palm Bay St. Thomas University, Miami, FL The British Library, Boston Spa, Weatherby, U.S. Department of Commerce, National UK Marine Fisheries (Panama City) Charles Courtney / Hillsborough County EPC U.S. Environmental Protection Agency Library, Florida Institute of Technology Gulf Breeze, FL Harbor Branch Oceanographic Institution, Inc. University of Central Florida, Division of Lake-Sumter Community College Sponsored Research Newfound Harbor Marine Institute University of South Florida CORPORATE MEMBERS FOR 1994 Mote Marine Laboratory, Sarasota Membership applications, subscriptions, renewals, changes of address, and orders for back numbers should be addressed to the Executive Secretary, Florida Academy of Sciences, P.O. Box 033012, Indialantic, FL 32903-0012 PUBLICATIONS FOR SALE by the Florida Academy of Sciences Complete sets. Broken sets. Individual numbers. Immediate delivery. A few num- bers reprinted by photo-offset. All prices strictly net. Special bargain: $500 for nearly complete sets. Prices quoted include domestic postage. Some issues may not be available. All are $20 per volume or $5 per issue, except for symposium issues. PROCEEDINGS OF THE FLORIDA ACADEMY OF SCIENCES (1936-1944) Volumes 1-7 QUARTERLY JOURNAL OF THE FLORIDA ACADEMY OF SCIENCE (1945-1972) Volume 8-35 FLORIDA SCIENTIST (1973) Florida’s Estuaries—Management or Mismanagement?—Academy Symposium FLORIDA SCIENTIST 37(4)—$5.00 Land Spreading of Secondary Effluent—Academy Symposium FLORIDA SCIENTIST 38(4)—$5.00 Solar Energy—Academy Symposium FLoripA SCIENTIST 39(3)—$5.00 (includes do-it-yourself instructions) Anthropology—Academy Symposium FLORIDA SCIENTIST 43(3)—$7.50 Shark Biology—Academy Symposium FLORIDA SCIENTIST 45(1)—$8.00 Future of the Indian River System—Academy Symposium FLORIDA SCIENTIST 46(3/4)—$15.00 Second Indian River Research Symposium—Academy Symposium FLORIDA SCIENTIST 53(3)—$15.00 Please send payment with order. If required, an invoice will be sent on purchases over $20 from a recognized institution. Please include Florida sales tax if shipped to a Florida address, unless a copy of a valid Florida Consumer’s Sales Tax Certificate is enclosed. The Florida Academy of Sciences, Inc. NON-PROFIT P.O. Box 033012 ORGANIZATION Indialantic, Florida 32903-0012 U.S. Postage PAID Permit #26 Melbourne, FL ISSN: 0098-4590 pF 65 H lorida Scientist Volume 57 Fall, 1994 Number 4 CONTENTS Recognition Characters and Juxtaposition of Florida and Mississippi Slimy Salamanders (Plethodon Glutinosus Complex) .........:::cccecsseseeteteeeeeteees James Lazell 129 Benthic Invertebrates and Allied Macrofauna in the Suwannee River and Biter mp glzcosy stem, MlOrid ay c.f oe. cee: cxcctveasseatecwe mducednineseouconvbuteureuteageace: William T. Mason, Jr., Robert A. Mattson and John H. Epler 141 Florida's Tomato Producers: Are They Moving Toward Sustainability? ........ M. E. Swisher and E. P. Bastidas 161 Identification of Sailfin Catfishes Introduced to Florida ...............ccccceseeeeeee: Lawrence M. Page 171 Additions to the Flora of South Florida: Four New Species of Naturalized CDT IDL TNS ar aes Ua a a a A NO rec ic ONe A e John B. Pascarella 173 Selenium Levels in the Palm River and MacKay Bay Region of Tampa Bay Robert F. Benson, Joseph A. Saraceno, David R. Lowell and David W. Brett 177 NW et a AAG ANA Ah II uct dll vadact he, 186 OZ ETS ee oe eee, Oe Ore Ca Barbara B. Martin 187 TEMPE PMTCT OR INE VICWETS (on. 00vens0222deceaencoeieccnadaconieazesdidssesavesdlasheeadenasnoness 188 aN UN SIP ce AL ia or RN Re Rt en 191 FLORIDA SCIENTIST QUARTERLY JOURNAL OF THE FLORIDA ACADEMY OF SCIENCES Copyright© by the Florida Academy of Sciences, Inc. 1994 Editor: Dr. DEAN F. MarTIN Co-Editor: Mrs. BARBARA B. MARTIN Institute for Environmental Studies Department of Chemistry University of South Florida 4202 East Fowler Avenue Tampa, Florida 33620-5250 Phone: (813) 974-2374 E-Mail: Martin@chuma.cas.usf.edu THE FLoripa SCIENTIST is published quarterly by the Florida Academy of Sciences, Inc., a non-profit scientific and educational association. Membership is open to individuals or institutions interested in supporting science in its broadest sense. Applications may be obtained from the Executive Secretary. Direct subscription is available at $40.00 per calendar year. Original articles containing new knowledge, or new interpretations of knowledge, are welcomed in any field of Science as represented by the sections of the Academy, viz., Biological Sciences, Conservation, Earth and Planetary Sciences, Medical Sciences, Physical Sciences, Science Teaching, and Social Sciences. Also, contributions will be considered which present new applications of scientific knowledge to practical prob- lems within fields of interest to the Academy. Articles must not duplicate in any substantial way material that is published elsewhere. Contributions are accepted only from members of the Academy and so papers submitted by non-members will be accepted only after the authors join the Academy. Instructions for preparation of manuscripts are inside the back cover. Officers for 1994-1995 FLORIDA ACADEMY OF SCIENCES Founded 1936 President: Dr. Patricia M. Dooris Treasurer: Dr. FREDERICK B. BUONI Institute for Environmental Studies Florida Institute of Technology Department of Chemistry 150 W. University Blvd. University of South Florida Melbourne, Florida 32901 Tampa, Florida 33620-5250 Executive Secretary: Mrs. BETTY PREECE President-Elect: Lisa B. BEEVER P.O. Box 033012 Charlotte County — Punta Gorda MPO Indialantic, Florida 32903-0012 2800 Airport Road, A-6 Tel: 407:723-6835 Punta Gorda, Florida 33982-2411 Program Chair: Dr. DEL DELUMYEA Secretary: Mrs. MARCELLA GUITERREZ-MAaYKA Millar Wilson Laboratory for Chemical Research 701 E. River Dr. Jacksonville University Temple Terrace, Florida 33617 Jacksonville, Florida 32211 Published by the The Florida Academy of Sciences, Inc. P.O. Box 033012 Indialantic, Florida 32903-0012 Printing by C & D Printing Company, St. Petersburg, FL Florida Scientist QUARTERLY JOURNAL OF THE FLORIDA ACADEMY OF SCIENCES DEAN F. Martin, Editor BARBARA B. MartTIn, Co-Editor Volume 57 Fall, 1994 Number 4 Biological Sciences RECOGNITION CHARACTERS AND JUXTAPOSITION OF FLORIDA AND MISSISSIPPI SLIMY SALAMANDERS (PLETHODON GLUTINOSUS COMPLEX) JAMES LAZELL The Conservation Agency, 6 Swinburne Street, Jamestown, RI 02835 Asstract: Plethodon of the glutinosus complex include the forms grobmani Allen and Neill (1949) and mississippi Highton (1989) which approach each other on the Alabama Gulf Coast. Size, facial/snout proportions, and light-dark pattern elements facilitate separation of the two, known to be distinct in protein electrophoresis. Remarkably, the largely Florida form grobmani extends across Mobile Bay to at least Mon Louis Island, southeastern Mobile County, Alabama. The largely Mississippi form, mississippi, occurs in central Mobile County and adjacent southeastern Mississippi. Nearly ubiquitous habitat destruction over most of Mobile County precluded determination of such alternatives as close parapatry, sympatry, or intergradation between these forms. Most grobmani have a light axilla. In most mississippi the axilla is darkly pigmented. VaRIATION, be it individual, ontogenetic, or geographic, in characteristics of slimy salamanders of the Plethodon glutinosus complex has long been bewildering (Bishop, 1943; Highton, 1962). In a sweeping analysis of variation in some proteins and enzymes, the readily and regularly electrophoresed gene products (RREGS), Highton (1989) was able to divide this extraordinary complex into 16 forms. Highton (1989) provided scant clues as to other sorts of characters helpful in separating those forms. Except for cases involving sympatry and therefore unequivocal evidence of full biological species distinctness (e.g. Plethodon glutinosus sensu stricto vs. P. kentucki or P. aureolus), Highton (1989) cited no evidence for taxonomic status of the parapatric or dichopatric forms. Many of these Highton (1989) called semispecies; Tilley (1990), in reviewing Highton’s work, called them semispecies or subspecies. Grobman (1944) stated his belief that definition of the forms would necessitate study of living material. To this end, in 1990 I began field study of the form named Plethodon mississippi Highton (1989) and its interface with P. glutinosus sensu stricto. That study has expanded from the edges of the state of Mississippi north and east into Tennessee, across Alabama, and to far-flung portions of the range in New England, the Carolinas, and Texas (Lazell, MS). 130 FLORIDA SCIENTIST [VOL 57 SVL 15 25 35 45 SCORE Fic. 1. Quantitative differentiation of salamanders of the Plethodon glutinosus complex in southeast- ern Mississippi and coastal Alabama. Solid squares, mississippi. Open squares, grobmani. Score is explained in “methods,” see text. No. 4 1994] LAZELL—RECOGNITION CHARACTERS OF SLIMY SALAMANDERS 131 Through the efforts of my colleague Thomas Sinclair, who collaborated in the present study, I am now able to report on the unexpected juxtaposition of the Mississippi form and the Florida slimy salamander described as Plethodon glutinosus grobmani Allen and Neill (1949) where these forms approach each other on the Alabama Gulf Coast. MetHops—All salamanders referred to herein are members of the genus Plethodon and the Plethodon glutinosus complex. So as not to prejudice the open question of their biological relationships, which will prove complicated to finally determine because of the combination of morphological, color, and biochemical characters involved, I hereafter refer to the forms in question by their trivial names only, avoiding binomens or trinomens: the Florida slimy salamander, grobmani, and the Mississippi slimy salamander, mississippt. Highton (1989: 69) described grobmani as a “small species with large brassy dorsal spots and abundant lateral white or yellow spotting,” and stated: “Morphologically, it is not detectably different from...” mississippi, which Highton (1989: 65) described in identical terms. I have examined more than a thousand specimens from the central South, including 765 from Mississippi, of which 503 are from the coastal zone. It was receipt of living specimens of grobmani from Ware and Brantley Counties, southeast Georgia, now in the Museum of Comparative Zoology (MCZ A-11673-5), that inspired me to look more closely into the differences between grobmani and mississippi. Accordingly, Sinclair and I set out 11- 14 November 1993 to examine the interface region. A character scoring system was devised utilizing facial/snout proportions and several aspects of pattern. This score was plotted against size. Colors in life were recorded in field notes and color photographs were taken by Sinclair. Size is the least tractable character; both forms start life at less than or about 20 mm snout-vent length (SVL) and grow to about 65-70 mm SVL (Highton, 1956; present study, Fig. 1). I measure SVL to the posterior edge of the vent instead of the anterior edge. The difference is about six percent, with my measurements herein to posterior edge consistently longer. Examination of the large coastal series of mississippi noted above indicated four size classes present in November, with hatchlings (the smallest at ca 19 mm SVL) being extremely rare. Nevertheless, for the sake of simultaneous comparison, we collected 34 fresh specimens of mississippi in Perry and Greene Counties, Mississippi, 13 November 1993. These form the basis of comparison to Alabama specimens of each taxon. We collected 31 specimens from five localities in Mobile and Baldwin Counties, Alabama, 11-12 and 14 November 1993. (See Fig. 1). Snout length difference is shown in Fig. 2. The difference, easily seen in comparison of actual specimens, is difficult to quantify and score. I devised the following measurement method: a plastic microscope slide cover slip was applied to the salamander’s snout and stopped posteriorly by the orbits. The cover slip was held in place with index finger, and the salamander with other fingers, on damp paper towel to resist slippage. With the other hand, I bought a vernier caliper tip to a stop on edge of the slip at position A (Fig. 2). The other caliper tip was then brought to a position on the slip corresponding to position B (Fig. 2), directly over the tip of snout. The resulting snout length, along the midline, was used in ratios. Head width, as a maximum measurement, seems prone to some sexual dimorphism: obvious adult males with enlarged, prominent mental glands have broader and more jowly heads than females (equally large but without mental gland development). There is also great variation in head width across the orbits resulting from differential orbital depression in the salamanders when fixed in fluid. Therefore, facial width was measured immediately in front of the orbits: distance C in Fig. 2. Snout width/length ratio, measured as described above, was calculated for all specimens. This is multiplied by 10, rounded to the nearest whole number, and added to the overall score. A scoring method was used for light or dark color and certain pattern elements. Light lateral blotches that are scattered and cover less than one half the lateral skin surfaces score 0. When the light and dark pattern elements are about equal in coverage the score is 1. When more than half the lateral skin is light but about 20 percent dark ground color remains the score is 2. When light pattern elements amalgamate reducing the dark ground color to a broken reticulation the score is 3. Pigmentation of the gular fold edge requires examination under a dissecting microscope. An entirely dark edge scores 0. A single unpigmented spot on an otherwise dark edge scores 1. The presence of a mix of pigmented and clear areas along the median edge scores 2. An unpigmented gular fold edge across more than one half of the midventral area scores 3. (See Fig. 3). 132 FLORIDA SCIENTIST [VOL 57 Fic. 2. Heads in dorsal view of (left) Mississippi slimy salamander and (right) Florida slimy salamander showing points A and B and measurement C described in “methods.” Traced from photographs of adult males MCZ A-11837 and A-11832, both 64 mm SVL. Fic. 3. Semi-diagramatic ventral views of Plethodon glutinosus complex salamanders. Numbers are scores for character states described in “methods,” see text. No. 4 1994] LAZELL—RECOGNITION CHARACTERS OF SLIMY SALAMANDERS 133 Axillary pigmentation is the most easily seen difference between the forms. It can be estimated (light or dark) in living animals and quantified more tightly under the microscope. A dark, fully pigmented axilla scores 0. A narrow reticulum of light and dark just along the skin fold ventrally scores 1. Patchy pigmentation along both sides of the fold with some clear areas scores 2. A wide zone of unpigmented skin on the ventral axillary fold scores at least 3; it scores only 3 if the zone is narrower than the tip of the third digit of the manus, but 4 if the unpigmented zone is equal to or greater than the width of the third digit (Fig. 3). Each axilla is scored separately and each score is added to the sum. The character score thus consists of five factors: facial/snout ratio times ten, lateral pattern, gular fold edge pigmentation, and pigmentation of the left and right axillae. The score may then be plotted against size. Differences between sample scores were evaluated by Student’s t-test; “significance” implies the means were different at 95 percent level of confidence. Coloration in life was not quantified, but mississippi is often more colorful than any grobmani seen. REsuLTs—At least where they are juxtaposed in coastal Alabama, the forms grobmani and mississippi seem well differentiated: Fig. 1. At the time of Highton’s (1989) revision based on RREGS, the known geographic relationships of the forms were as depicted in Fig. 4. Itseemed reasonable to assume that Mobile Bay separated the forms along the coast. The results of our study are shown in Fig. 5. The occurrence of a grobmani population on Mon Louis Island, just north of Alabama Port, west of Mobile Bay was totally unexpected. The strength of the differences between the two forms and our failure to find intermediate populations were also unexpected. Analyzed separately, the character states that differentiate the two forms show broad or even complete overlap (Table 1). Summed, the set of differences provide a significant quantitative difference. The only single characters showing significant differences are axillary pigmentation and head proportions (Table 1). Axillary pigmentation may show character approach in Mobile mississippi to grobmani, but given the small sample sizes this cannot be tested statistically. If the forms intergrade, it may prove easiest to demonstrate this north of the latitude of Mobile Bay, outside the coastal zone herein studied. I recognized the Mon Louis Island grobmani as such, in the field. The scoring method was developed a posteriori, back in the lab. Certainly field familiarity with living salamanders facilitates identification, as predicted by Grobman (1944). Labo- ratory identification is far more tedious. Coloration in life augments the differences. Usually, mississippi are rather colorful and grobmani less so. While grobmani may appear as described by Highton (1989; quoted above), many have near-white to ashy dorsal flecks and ash-gray, or light grayish-brown, lateral blotches. Most mississippi have yellowish to golden dorsal flecks with variation including bright gold, red-gold, copper, and green-gold or copper tarnish hues. Lateral blotching may be buffy to light yellow-green (e.g. MCZ A-11836, Greene County, Mississippi). Within Mobile County, the populations of the two forms we were able to sample are disjunct by about 38.5 km. On 13-14 November 1993, we attempted to locate intermediate populations in the vicinities of Seven Hills, Dees, Dawes, Theodore, Irvington, and Fowl] River - allin Mobile County. We extended our efforts to southeastern George and northeastern Jackson Counties, Mississippi, as well. We were unsuccessful. [VOL 57 FLORIDA SCIENTIST 134 ‘UD OOT staeg ‘Apnys EG6T JoquIsAoN oy} 0} 101d aur Aq eater uses suautteds 107 are soport9 uedG *(686T) uo wstP Aq ATfeoyoe10ydo.da[e pourutrejop suoneindod 0} ore sonteooT ‘yoroidde Aay} a1aym (sarenbs) iepueuryes Aus tddisstsstpy oy} yo a8ues oy} Jo Wed pur (sjop) 1opuvureyes AULS EPLLOT] G4] JO aduKI oY} SUIMOYS S9}e}S POPU) UIOIsvOyINOS oY} JO Weg “f “Oly No. 4 1994] LAZELL—RECOGNITION CHARACTERS OF SLIMY SALAMANDERS 135 Fic. 5. Present study area, coastal Alabama and adjacent southeastern Mississippi. Dots are localities for Florida slimy salamanders. Squares are localities for Mississippi slimy salamanders. Open circles and squares are for museum specimens not seen alive. Stippling indicates the approximate area originally covered by slash pine (Pinus elliotii) savannahs and wire grass (Aristida stricta) prairies. Counties mentioned in the text are: P, Perry. G, Greene. E, George. J, Jackson. M, Mobile. B, Baldwin. 1 is Mon Louis Island. 2 is Dauphin Island. 3 is Pelican-Sand Island. Bar is 20 km. We did successfully locate populations of slimy salamanders at Millers Park, Mobile, and in Perry and Greene Counties, Mississippi. The habitats were a mixture of pines and hardwoods with a high percentage of oaks. White oak, Quercus alba, was a good indicator species. Much or most of Mobile County, especially along its western border with Mississippi, was originally wiregrass prairie (Hodgkins, et al. 1979; see below), avery poor habitat for slimy salamanders. Much of this has been drained, planted for pulpwood, and repeatedly harvested; this constitutes even less suitable habitat. Almost all the rest of the County, at least south of Millers Park at the extreme northwest corner of the Mobile city limits, is pine pulpwood acreage, or vast pecan groves, or cotton fields, or residential and commercial development. Even at Millers Park, mere remnants of woodland remain. Most of the Park is now a golf course. On Mon Louis Island, east of Route 193 (the Dauphin Island Parkway), house lots extend to the Bay. West of this road, a modest two-lane blacktop in this stretch, the island is typically cleared or so cut over that no suitable habitat seems to remain. [VOL 57 FLORIDA SCIENTIST 136 (8'SFP BS )0E-ST (FI 0F06 I)PT'Z-€S'T (I: 1+ 1)P-0 (0'1+8'0)3-0 (9'0FF'0)3-0 OF iddississtur [Ty (8°SFE'SS)OE-ST (ST 0468 'I)P1'S-ES'T (0'1+0'T)F-0 (0'1+8'0)3-0 (L'0FS'0)3-0 Pe (9‘d) tddississtur (L°G¥9'€S)8ZS-1Z (L0'0+06'1)00°G-6L'T (CT¥6'T)F-1 (O'T#L°0)3-0 (F'0F6'0)T-0 9 (IN) tddississtux (I SFT 'SE)EF-0€ (€6'0F0S'S)60'E-€0'S (S'O0F9'€)F-€ (6 OFL'T)€-0 (O'LFT T)€-0 GG juRUIqoIs [[V (FSFE 'SE)EP-TE (ZS OFSS'S)60'E-ES'S (GOFF €)F-€ (6 0FS'T)3-0 (ST +P 1)€-0 ra (q) rueurqors (6'G¥6 PE)6E-0F (PS OFOP'S)ES'S-E0'S (F'OF8'E)F-€ (O'T+6'T)€-0 (S'0#8'0)3-0 €1 (JA) Fueurqors AWYOOS 1S/MH oRT[EXV qejns pero] u (AjuNOD) UOXe], 1X9} UL _SpOYJOUL,, 9s :4{SUa] NOUS SITS ‘YPM pesy st AAP ‘Setoos us9}3ed 0} Jofor VET[IXY pur ‘IE[Nd ‘[e19}e'] “G “BI Vas :oUsaIg=5 pue Aliog= J oe saiqunory Iddississi ‘uLMpreg=q puke a[Iqop= ee saqUNOD eueqery ‘2ddrssisstwu st rapueuepes Aus Iddississtyy ay} ‘2uDWGoss si JapueUTEes AUIS EPPLOL A OYL ‘tddississty yuooe(pe pure eureqely Wisy NOs UI EGET JEquIOAON] UT aATTe paydaT[Oo s1apueUreyes xofduroo snsournjs uopoyja}g Jo suonsodoad pur sasoog *T ATAV |, No. 4 1994] LAZELL—RECOGNITION CHARACTERS OF SLIMY SALAMANDERS 137 Some of the house lots, on high ground between small fresh water swales entering the Bay, remain wooded. On one of these, four miles north of Alabama Port, slimy salamanders were abundant. We did not attempt to investigate others for lack of time and owner permission. Three specimens of grobmani in the U.S. National Museum of Natural History (USNM 156756-8) were collected 27 March 1964 “1.2 miles West of D’Olive Creek on Route 90,” Baldwin County. This locality would be in the Bridgehead section of Spanish Fort. All appear to be very typical grobmani with unpigmented gular fold edges and very broadly unpigmented axillae. They seem to have been fixed in strong formalin which has apparently artificially obscured much of their dorsal and lateral pattern. They are strikingly different from the mississippi specimens we collected in Millers Park, Mobile, just 32 km farther west. Two additional specimens are relevant. MCZ 4745 was collected at Mobile by J. Hurter. The data read 2 June 1910, but it is not certain if this was date of collection or receipt at Cambridge. The specimen is a juvenile, 34 mm SVL. It is nearly uniform tan-brown, obviously much faded. Nevertheless, the axillae and gular fold edge seem to be as fully pigmented as the surrounding ventral surfaces. The facial/snout ratio is 2.0, quite typical of mississippi. In 1910 Mobile was a relatively small town. I have thus taken the locality datum quite literally and mapped this specimen at the highway confluences apparently corresponding to the position of the town as shown by Mohr (1901). If this specimen is really mississippi, as it appears to be, it narrows the gap between the forms to ca 16 km (Fig. 5). The second specimen, MCZ 26360, was collected by George Nelson between 26-30 November 1945 at Silver Hill, Baldwin Co. This locality, ca 13 km east of our new Montrose site, is interesting because it is well within the wiregrass prairie zone (Hodgkins et al. 1979): see below. The specimen is badly deformed and virtually black; it appears to have been desiccated. It is aca 60 mm SVL female. The short facial/snout ratio is 2.9, nearing the extreme for grobmani. The axillae and gular fold edge seem to have been relatively little pigmented compared to the rest of the ventral surfaces. Thus, I cannot judge this specimen to be different from typical grobmani. Because of the typical difficulties involved in trying to identify old museum speci- mens, it could be any sort of slimy salamander. Discussion—Onur goal in this study was to facilitate identification, in the field and in museum collections, of Florida and Mississippi slimy salamanders. One hopes these data will enable more extensive determination of precise ranges, biological relationships, and survival refugia for the forms. More proximate populations should be sought. An attempt to determine concordance or discordance of the morphologi- cal characters and the biochemical differences described by Highton (1989) should be made. Additional biochemical and molecular approaches, especially involving mitochondrial DNA or introns (see Hillis and Moritz, 1990) may be especially appropriate. All of these avenues will be best followed from a local base, most auspiciously in a university setting with laboratory facilities. The remarkable juxtaposition of the forms, with mississippi, determined by RREGS, occurring in Clarke County east of the westernmost grobmani localities farther south (Fig. 4), and with apparent grobmani occurring west of Mobile Bay 138 FLORIDA SCIENTIST [VOL 57 (Fig. 5) begs explanation. Considering the Recent configuration of land, terrestrial ecosystems, sea, and river drainages, one can envision land barge transport across Mobile Bay. Mount (1975) invokes this basic method to explain the presence of kingsnakes and mudsnakes of eastern genetic influence west of the Bay on Dauphin Island. Rosemary or bottlebrush, Ceratiola ericoides (Empetraceae), similarly seems to have crossed the Bay on moving sand barrier islands and has managed to colonize bits of the adjacent mainland as well (Williams, 1981; W. McDearman, 1994). I investigated the Pelican-Sand Island complex in 1984 and found evidence of snakes, rodents, and mink (unpublished data). This island (or, variably, pair of islands) has moved westward and curved northward dramatically in recorded history. Indeed, the pattern of arcuate spits proceeding from the eastern Alabama and Florida coastal peninsulas westward to form or amalgamate with Gulf barrier islands seems well documented (Otvos, 1981, 1985). Pelican-Sand Island will predictably someday join Dauphin Island, bringing along terrestrial organisms whose point of departure was east of Mobile Bay. Salamanders are not usual inhabitants of barrier islands (Lazell, 1979), but Dauphin Island is an exception to this rule, harboring an unusually diverse amphib- ian fauna for a barrier island, including salamanders (Mount, 1975). However, no Plethodon is known from Dauphin Island (Jackson and Jackson, 1970; Mount, 1975; much field time spread over two decades by me and colleagues, including several hours on 12 November 1993). If slimy salamanders crossed Mobile Bay on a land barge, they either missed Dauphin Island or a transient population there has been extirpated in the meantime. Going back somewhat further in time, to Pleistocene glacial maxima, most recently about twelve thousand years ago, one need envision no land movement. The sea was then more than 100 m below present level. There was no Mobile Bay. The coast of the Gulf of Mexico was about 110 km south of the present site of Mon Louis Island (Dolan, 1970). At this time, slimy salamanders could certainly have dispersed to their present positions over land. Within Mobile County, the present populations of grobmani and mississippi appear to be separated by what was originally a savannah or plains ecosystem once dominated by wiregrass, Aristida stricta (Hodgkins et al. 1979). Wiregrass drops out quickly west of the Alabama state line and Andropogon spp. become the dominant savannah grasses in Mississippi (Peet, 1993). Hodgkins and coworkers (1979) mapped extensive coastal flatwoods east of the Pascagoula estuary, but a much narrower strip west of it. Wolfe (1971) delineated the northern edge of slash pine (Pinus elliottii) savannah where it meets the forest of the pine belt dominated by longleaf pine (Pinus palustris). Ground truthing convinces me that Wolfe’s (1971) line more accurately depicts the original extent of savannah or plains in western Jackson (and adjacent Harrison) County than does the line of Hodgkins and coworkers (1979) in that limited area (Fig. 5). Slimy salamanders seem to be absent or very rare in this savannah zone. This zone is open country, seasonally marshland, dominated by grasses, studded with pitcher plants (Sarracenia spp.), with scattered pine trees and shrub thickets. It is No. 4 1994] LAZELL—RECOGNITION CHARACTERS OF SLIMY SALAMANDERS 139 sandwiched between the original longleaf pine forest to the north and a maritime forest right along the coast: the littoral belt of Mohr (1901). The maritime forest is a mix dominated by slash pine, magnolia (M. grandiflora), and various oaks, most obviously live oak, Quercus virginianus. Maritime forest flanks Mobile Bay and the Pascagoula estuary, and merges into the river bottom hardwood forests that extend northward into the pine belt (Hodgkins et al. 1979). Maritime forests and river bottom hardwoods are especially rich habitats for slimy salamanders (Fig. 5). An alternative explanation for present slimy salamander deployment, deriving from physiographic conditions in the Tertiary, is suggested by Fitzpatrick (1986). In this scenario, the Tennessee River cut southwestward from its present elbow in Marshall County, northern Alabama, traversed present-day southern Mississippi, to exit into the Gulf of Mexico. At such time, riverine flood plains exiting to the Gulf in the Mobile area would have been narrow because most drainage was via the great river to the northwest. The forms mississippi and grobmani might then have been isolated, with grobmani occurring westward to the floodplain of that ancient Tennessee River, and mississippi occurring eastward to it. If this situation obtained, one may speculate that mississippi has expanded its range east and south across the ancient valley while grobmani has merely held its ground, or actually lost range. This would be puzzling because one might expect grobmani, occurring much farther south (into peninsular Florida) than mississippi, to be the better adapted of the two at the southern edge of the range. These hypotheses invoke very different temporal sequences. The first calls for a very recent, post-glacial colonization. The second involves a glacial maximum dispersal, some 60 to 12 thousand years ago. The third involves retention of a far more ancient, pre-Pleistocene deployment with an existing relictual grobmani population west of Mobile Bay, going back more than 700,000 years. Biochemical methods could be used to test these hypotheses, alluded to above. Slimy salamanders are tough, resilient organisms. They do not require pristine habitats or old growth forests. Everywhere we found them was disturbed to some degree, at best second growth. We found them under discarded rubbish such as lumber, roofing, sheet metal, and concrete blocks, as well as under logs in the woods. There may be a minimum size wood lot necessary for slimy salamander survival. We did locate and search several seemingly florally suitable habitats in the areas where we failed to find salamanders, but all were very small. Our inability to find slimy salamanders over wide areas, in contrast to the ease with which we found them elsewhere, signals the pervasive extent of habitat destruction in much of this portion of the Gulf Coast region. The implications for species less durable than slimy salamanders are depressing. ACKNOWLEDGMENTS—This project was made possible by the efforts of Thomas Sinclair and funded by The Conservation Agency. Many landowners gave us permission to root through their property, most notably Charles Stewart, four miles north of Alabama Port, who was visited repeatedly. Dorothy Allard, The Nature Conservancy; Anne Bradburn, Tulane University; Will McDearman, Mississippi Museum of Natural Science; Sidney McDaniel, Mississippi State University; and Robert Peet, University of North Carolina, supplied critical botanical records and references. The curatorical staffs at the Museum of Comparative Zoology, Mississippi Museum of Natural Science, U.S. National Museum of Natural History, and the University of Georgia Department of Zoology cooperated in making specimens available. 140 FLORIDA SCIENTIST [VOL 57 LITERATURE CITED ALLEN, E.R. AND W.T. NEILL. 1949. A new subspecies of salamander (genus Plethodon) from Florida and Georgia. Herpetologica 5: 112-114. BisHop, S.C. 1943. Handbook of Salamanders. Comstock, Ithaca, NY. Dotan, R. 1970. Coastal landforms. The National Atlas of the United States of America. U.S. Department of the Interior, Geological Survey: 79, Washington, D.C. FITZPATRICK, J.F. 1986. The pre-Pliocene Tennessee River and its bearing on crawfish distribution (Decopoda: Cambaridae). Brimleyana 12: 123-146. GropMaN, A.B. 1944. The distribution of the salamanders of the genus Plethodon in eastern United States and Canada. Ann. New York Acad. Sci. 45(7): 261-316. HicuTon, R. 1956. The life history of the slimy salamander, Plethodon glutinosus, in Florida. Copeia 1956(2): 75-93. . 1962. Revision of North American Salamanders of the genus Plethodon. Bulletin of the Florida State museum 6: 597-613. . 1989. Biochemical evolution in the slimy salamanders of the Plethodon glutinosus complex in the eastern United States. Part 1. Geographic protein pattern. Illinois Biological Monographs Dies. Hituis, D.M., AND C. Moritz. 1990. Molecular Systematics. Sinauer Associates, Sunderland, MA. Hopckins, E.J., M.S. GotpEN, AND W.F. MILLER. 1979. Forest habitat regions and types on a photomorphic-physiographic basis: a guide to forest site classification in Alabama-Mississippi. Southern Cooperative Series 210, Alabama Agricultural Experiment Station, Auburn, AL. Jackson, C.G. and M.M. Jackson. 1970. The herpetofauna of Dauphin Island, Alabama. Quarterly Journal of the Florida Academy of Sciences 33: 281-287. LazELL, J. 1979. Deployment, dispersal, and adaptive strategies of land vertebrates on Atlantic and Gulf barrier islands. Proc. First Conference Sci. Res. National Parks 1: 415-419. McDearman, W. 1994. Mississippi Museum of Natural Science, Jackson, Pers. Commun. Monr, C. 1901. Plant life of Alabama, an Account of the Distribution, Modes of Association, and Adaptations of the Flora of Alabama, together with a Systematic Catalog of the Plants growing in the State. Alabama edition, Brown Printing, Montgomery, AL. Mount, R.H. 1975. Reptiles and Amphibians of Alabama. Agricultural Experiment Station, Auburn, AL. Orvos, E.G. 1981. Barrier island formation through nearshore aggradation — stratigraphic and field evidence. Marine Geo. 43: 195-243. . 1985. Coastal evolution — Louisiana to northwest Florida. Guidebook, Am. Asso. Petrol. Geologists Annual Meeting, New Orleans, Geological Survey: 1-91. PEET, R.K. 1993. A taxonomic study of Aristida stricta and A. beyrichiana. Rhodora 95(881): 25-37. TILLEY, S.G. 1990. Review: Biochemical evolution in the slimy salamanders of the Plethodon glutinosus complex in the eastern United States. Herp. Rev. 21(4): 99-100. Wi.uiaMs, D.L. 1981. Scrub oak-saw palmetto vegetation of southern Mississippi. J. Miss. Acad. Sci. 36:46. WoL FFE, J.L. 1971. Mississippi Land Mammals. Mississippi Museum of Natural Science, Jackson, MS. Florida Scient. 57 (3):129-140. 1994. Accepted: March 18, 1994 No. 4 1994] MASON ET AL.—BENTHIC INVERTEBRATES IN THE SUWANNEE RIVER 14] Biological Sciences BENTHIC INVERTEBRATES AND ALLIED MACROFAUNA IN THE SUWANNEE RIVER AND ESTUARY ECOSYSTEM, FLORIDA WiiuraM T. Mason, Jr. , Ropert A. Mattson ”, AND JOHN H. EpLer® ® Southeastern Biological Science Center, National Biological Survey, U. S. Department of the Interior, 7920 N. W. 71st Street, Gainesville, Florida 32653 ®) Suwannee River Water Management District, Route 3, Box 64, Live Oak, Florida 32060 ®) Environmental/Biological Consulting, Route 3, Box 5485, Crawfordville, Florida 32327 Asstract: A total of 707 freshwater and estuarine benthic invertebrate taxa and allied macrofauna was collected from the Suwannee River and estuary during five separate surveys between 1979 and 1993. Relative percentages of taxa occupying seven vertical habitat zones were as follows: surface-dwelling, 8%; hyperbenthos, 7%; epibenthos-section I, 14%; epibenthos-section I, 48%; embenthos, 18%; hypobenthos, 1%; and epizoos/parasites, 4%. Arthropoda numbered about 70% of the total taxa, of which, 56% were Insecta of the orders (in descending order of prominence) Diptera, Coleoptera, Hemiptera, Odonata, Ephemeroptera, Trichoptera, and Plecoptera. The arthropod Crustacea composed 11% of the total taxa, and about one-half of these were Decapoda. Molluscan taxa composed 12% of the total and, of these, the number of Gastropoda taxa was about three times the Bivalvia. Annelid taxa contributed 15% of the total taxonomic richness. The exclusively estuarine taxa, 28% of the total, were split almost evenly among arthropods, molluscs, and annelids; however, in fresh waters, arthropods far exceeded the diversity of the other two phyla. Of the 3% other taxa, one-third were allied macrofauna such as Hemichordata and Chordata. The taxonomic composition is indicative of clean water conditions throughout most of the ecosystem. BENTHIC invertebrates and allied macrofauna are intermediate in the aquatic food web between the primary producer plants and vertebrate consumers and are essential to the overall health of ecosystems. The diversity (taxonomic richness) and composition of invertebrates are thus useful indicators of the overall capacity of aquatic ecosystems to support life. Information on the presence and distribution of some benthic invertebrates, such as the shellfish, is important to commerce in the region. This checklist is a composite developed from five separate surveys of the Suwannee River and estuary, Florida, between 1979 and 1993. Information from the surveys was used for (1) river basin status assessments under Sections 305(b) and 106(e) of the Clean Water Act (Florida Department of Environmental Regulation, 1985), (2) status of food resources of the threatened Gulf sturgeon Acipenser oxyrinchus desotoi in the system (Mason, 1991), (3) information for the Surface Water Improvement and Management by the state (Mattson, 1992), (4) effects of freshwater inflows on shellfish grounds in the Suwannee estuary (Wolf and Wolf, 1985), and (5) asurvey of food resources of the spotted seatrout Cynoscion nebulosus and other fish at the Cedar Keys National Wildlife Refuge in the Gulf of Mexico (Mason and Zengel, unpublished data). The list serves as a benchmark on the 142 FLORIDA SCIENTIST [VOL 57 structure of the benthic invertebrate community, and the vertical habitat zone classification will hopefully further the understanding of ecological and biotic association relationships. Basin description—The land area draining to the brown water Suwannee River has been variously cited as 25,770 km? (Heath and Conover, 1981) and 28,542 km? (Florida Department of Natural Resources, 1989). These differences are due to the ever-changing boundaries of the Okefenokee Swamp headwaters in southern Georgia. The Highlands subbasin is separated from the Lowlands by the Cody Scarp (Ceryak et al., 1983), a thumblike plateau that extends from Georgia into northern Florida. The river (Fig. 1) flows on an S-shaped course for about 400 km from the headwaters to the Gulf of Mexico and has two major tributaries: the Withlacoochee River enters at km 200, and the Santa Fe River confluence is at km 109. At km 233 during low flow, the average depth is about 0.25 to 0.5 m. However, heavy rains in the headwaters can cause the river at km 126 to rise quickly 6-8 m above normal stage. | Zz 7 a a S QO ° | 3 Pages! t w Se e = ! km 200 SS =) Su Kane’ | We ee | | ete | | i} 1 2 Lo», = or a aoc | . | 4t | | | GE RIV. | N | al | o Re | : Cie oe Gulf of L km 109 24 <% Mexico | 8 J | : | i | Sees li z km 55 ayy ees 5 ees Suwannee Sound \ y West Pass By East Pass i} e Pas} Cedar Keys ? NWR : 10 5 O 5 10 kilometers Seahorse Key Fic. 1. Map of the Suwannee River and Suwannee Estuary, Florida. No. 4 1994 | MASON ET AL.—BENTHIC INVERTEBRATES IN THE SUWANNEE RIVER 143 At km 55, the 37 year (1940-1977) average flow was 10,624 cfs, and average water level was 1.6 m (Source: U.S. Geologic Survey, Tallahassee, Florida). In the lower 5 km of the East Pass and West Pass (Fig. 1), the river is mostly fresh water and oligohaline (0.5-5 ppt) salinity to within 3-5 km from the passes at the Gulf of Mexico, and there quickly grades to mesohaline (5-18 ppt) salinity (Mason, 1991). The headwaters of the basin are acidic (Table 1), tannin stained, and low in nutrients (Florida Department of Natural Resources, 1989; Mason, 1991; Mattson, 1992). In the upper reach, several tributaries and the main stem receive small amounts of toxic, organic waste and thermal discharges from point and nonpoint sources. Generally, though, the mainstem meets the state water-use standards (Florida Department of Environmental Regulation, 1985; Mattson, 1992). In a survey of the mainstem, Mason (1991) found that 83% of the benthic invertebrate taxa were indicators of “clean water” conditions. Seahorse Key is one of five islands of the Cedar Keys National Wildlife Refuge that lie about 5 nautical miles west and 10 nautical miles south of the mouth of the Suwannee River at the Gulf of Mexico (Fig. 1). The Cedar Keys are in the north to south circulation flow of the Gulf loop that sweeps southward from the Big Bend area along northwestern coastal Florida and toward the Caribbean. At Seahorse Key, sand is predominant on the shallow coastal shelf (average depth at low tide is about 1 m), and several species of seagrasses—turtle grass Thalassia testudinium, shoal grass Halodule wrightii, and manatee grass Syringodium filiforme—grow in the intertidal zone. The water is well mixed by winds, and storms erode high sand embankments. The annual range of water temperatures was 12-31 °C (average 23 °C), and total salinity was 23-28 ppt (average 26 ppt). MetHops—The macrofauna was collected during all seasons and from a variety of different substrates. Organisms were collected from very fine sand near the Florida-Georgia border with a petite Ponar grab (15 x 15 cm square; 225 cm2) and, along shore, collections were taken from pockets of leaf litter and on wild celery (or eelgrass) Vallisneria americana with a modified bottom dip net (45 cm X 20 cm metal frame with cutting bar and 595 um-mesh bag net). Epiphytes on limestone outcroppings and snag communities, common in the Suwannee River, were collected with a modified Larsen (1974) benthic suction sampler. Either of two interchangeable rings were attached to the 18-cm-diameter suction head; serrated for sand substrates and convave for submerged logs, for vacuuming organisms. Suction occurs when a strong stream of water is injected through the center of the bag-end of the elbow by hose from an onboard 3-hp gas-powered centrifugal pump. Artificial substrate samplers—EPA hardboard multiplate samplers and rock-filled basket samplers (APHA et al., 1989; Klemm et al., 1990)—were installed at 0.25-0.5-m depths at 100-km increments in the mainstem, from the East Pass to the Florida border. These devices installed for sequential 4-6-week colonization by organisms mimicked snag habitat. The coarse sand, marl, old shell and oyster beds, coarse particulate organic matter, and silt substrates at the river’s mouth and estuary were sampled using a Ponar grab and a 9.5-cm-diameter steel corer (Wolf and Wolf, 1985). Mason and Zengel (unpublished data) collected the shallow-water estuarine hyperbenthos in sea grass beds at Seahorse Key in 10-m hauls using Pullen and co-worker’s (1968) steel sled trawl (17 x 25-cm opening and 563-um-mesh bag). Estuarine core samples were washed in a 500-um sieve, and a 595-um sieve was used on riverine hardboard multiplate samples. Grab samples and other artificial substrate samples were washed ina U.S. Standard No. 40, 425-um sieve. Samples were preserved in either 70% ethanol or 5% buffered formalin. Rose bengal, an animal tissue differential stain, was added (0.18 g/L) to the preservatives to aid in sample sorting (Mason and Yevich, 1967). Organism identification—The benthic invertebrates were identified by standard taxonomic references (APHA et al., 1989; Klemm et al., 1990). Questionable identifications were referred to other 144 FLORIDA SCIENTIST [VOL 57 TABLE 1. Average and range of physicochemical measurements for February 1989 to September 1991 in the Suwannee River, Suwannee Sound, and Seahorse Key, Florida. Average Values (Range) Habitat Reach/ km Flow WaterTemp. pH Color Salinity Area m3/s =; SUs Pt/Co 0/oo I 263-333 42 22 3.9 459 Banks steep, eroded (<1-518) (11-31) (3.1-7.4) (200-1,200) very fine sand; veg.— oak-gum-hickory-pine river channel sand porous limestone. II 205-262 54 29 6.2 272 Banks with Tulepo- (1-852) (12-27) (5.1-8.0) (30-1,000) gum some some cypress; cypress; bottom pockets fine sand on bedrock; springs numerous; snags III 110-204 199 ot 6.7 170 Banks moderately (43-2,376) (11-28) (5.6-8.2) (10-750) sloped; veg.-willow- gum-cypress; springs SAV sparse along shore; snags numerous IV 10-109 300 22 Th 150 Banks low, lined with (84-2399) (13-29) (6.1-8.3) (12-750) tangled web of cypress- willow-palmetto; SAV lines banks; snags com- mon; tidal to km 25. V 0-9 No data 23 ee 3 Typical marsh delta; (West Pass) (13-29) (6.5-8) (0-14) palmetto, saw grass- reeds; SAV lines shore, abundant; bottom sub- strate sand/mud/clay overlain with CPOM. VI Suwannee 23 8.0 24 Shallow coastal sand Sound? (12230>)) A7.128:5) (5-31) flats interspersed with oyster bars; shallow 1-2 m, but holes 3 fa deep VI Seahorse 23 8.24 26 Island with high eroded Key (12-31) (8.1-8.4) (23-28) sandy banks that slope steeply onto the shallows (0.5-1.5 m depth); seagrasses abundant No. 4 1994] MASON ET AL.—BENTHIC INVERTEBRATES IN THE SUWANNEE RIVER 145 taxonomists (listed in the Acknowledgments) for assistance. Unresolved identifications were noted in the checklist with a question mark after the original author. The taxonomy and systematics of Insecta are based on adults, and not the immature aquatic stages, but, unfortunately, the “immatures” of aquatic insects often have not been associated with the adults. Identifications to species require possession of the adults through capture, rearing, or other methods. Eventually the specimens should be compared to museum material. We collected adults with insect aerial nets from around lighted docks and other structures on shore, and from emergent vegetation at the water’s edge. Larval and pupal exuviae (cast skins after molts), also helpful in identification of many dipteran species, were skimmed from the water’s surface with a fine-mesh net. Collection of two or more life history stages at the same site and adults that keyed to the same species sufficed for species identification by indirect association. Voucher specimens are stored at the National Fisheries Research Center, National Biological Survey, USDI, Gainesville, Florida, the Suwannee River Water Manage- ment District, Live Oak, Florida, or Florida Department of Environmental Protection, Jacksonville, Florida. Vertical habitat zone classification—The benthic macrofauna was categorized into seven vertical zones, depending on their mode of life (Fig. 2; Table 2). Surface-dwelling—Organisms living on or suspended from the water’s surface or associated with emergent parts of plants wholly or part of the time (= true bugs, spiders, and other semi-aquatic invertebrates); Hyperbenthos—Living free in the water column but close to bottom substrates, aquatic plants, and other submerged objects and may temporarily feed, rest, or hide on the substrate’s surface (= drift macrofauna): Epibenthos-section I—Attached or relatively immobile organisms living in colonies or singly in tubes and shells attached to submerged substrate materials that protrude from the bottom (rocks, aquatic vegetation, sticks and snags), or associated with algae or moss on these objects (Aufwuchs, biofilm, epiphytes, epiphyton, and macroinvertebrates, in part); Epibenthos-section II—Mobile organisms, free-living or in self-constructed tubes and shells on or within flocculent bottom materials that overlay firm bottom sediments (macrofauna, macroinfauna, infauna, and macroinvertebrates, in part; meiofauna or meiobenthos, and microbenthos, in part); Embenthos—Burrowing, tunneling, and tube-dwelling organisms that live within interstices of firm bottom sediments but may feed at the surface (macrofauna, macroinfauna, infauna, and macroinvertebrates, in part; interstitial dwelling meiofauna or meiobenthos, and microbenthos, in part); Hypobenthos Living deep (generally >10 cm) within the bottom substrate in tubes open to the substrate’s surface or freely mobile within the substrate; Epizoos/parasites—Living as symbionts on or in other biota. Definitions for the zones are combinations of terms from standard freshwater and estuarine references (e.g., Holme and McIntyre, 1971). We mostly adhered to the definition of Mees and Hamerlynck (1992) for “hyperbenthos,” and divided Barnes’ (1987) “epibenthos” into two sections to increase resolution. We did not attempt to classify the small metazoans and other interstitial substrate- dwelling “meiofauna” nor the “microbenthos” because only a portion of these organisms were collected due to the relatively coarse mesh sieves used in sample washing. However, when collected in the future they could be placed in Epibenthos-Section II, Embenthos, and Hypobenthos zones. RESULTS AND Discussion—During 1979 to 1993, 707 benthic invertebrate taxa were collected in the five surveys (Appendix). In many ways the Suwannee River macrofauna resembled those communities from other brown-water, coastal rivers of the lower Southeast (Roback, 1953; Subrahmanyam et al., 1976; Soponis, 1980; Benke et al., 1984; Caldwell and Parrish, 1987). Generally, the composition of the Suwannee River estuarine species was similar to other benthic studies in the northeast Gulf of Mexico (Stoner, 1979; Lewis, 1984). In addition, we found many of the same amphipod species as reported for the Gulf of Mexico and mid-Atlantic to Florida Coast (Fox and Bynum, 1975). We caution, however, that most estuarine amphipod identifications in the Suwannee River Estuary surveys were based on several keys intended for the coastal mid-Atlantic and Northeast. Apparently there are many undescribed species of Amphipoda in the Gulf of Mexico (Farrell, 1993), and, therefore, our list is incomplete. 146 FLORIDA SCIENTIST [VOL 57 The taxonomic composition of phyla in the Suwannee River and estuary consisted of 70% Arthropoda, 15% Annelida, 12% Mollusca, and 3% other phyla. Insecta (Arthropoda) was by far the most diverse group, with 398 taxa or 56% of the total taxa. We estimate that our checklist (Appendix) contains about 60-75% of the actual total diversity of benthic invertebrates in the ecosystem. Combined crustacean arthropod diversity for the ecosystem was 11% of the total taxa. Crustacea were represented by 5% Decapoda and 3% Amphipoda. The remaining crustacean taxa were split almost evenly among Isopoda, Mysidacea, Cumacea, and other crustacea. The exclusively estuarine crustacean taxa accounted for about 9% of the total taxa of the river and estuarine ecosystem, but for the estuary alone, these taxa accounted for 98% of the macrobenthic arthropod diversity (See Hyperbenthos and Epibenthos-Section II, Appendix). The Polychaeta, all exclusively estuarine taxa, contributed 8.5% to the total diversity and composed 58% of the annelid taxa. Hirudinea was <1% of the total taxa. The percentages for Annelida; however, may be somewhat understated because the sampling effort in the river exceeded that in the estuary and some small forms passed through the sieve during the sieving process. Molluscs, especially snails, were most diverse in the Santa Fe River and in the middle reach of the mainstem river. Gastropods (42 taxa) living among the sub- merged aquatic vegetation at Seahorse key represented 6% of the total diversity and 51% of the total molluscan diversity in the ecosystem. This gastropod diversity at Seahorse Key is only a small fraction, perhaps 7-10%, of the total gastropod taxa of the Northeast Gulf Coast (Auffenberg, 1993). Bivalves were less than 3% of the total diversity in the ecosystem. The Asian clam Corbicula fluminea was ubiquitous in the Suwannee River and was the most ecologically significant bivalve in the lower and middle river (Mason, 1991; Mattson, 1992). Upriver from km 200, the Asian clam was also found mixed with native mussels Quincuncina infucata, Elliptio icterina, and a few sphaeriid species (Embenthos-Table 2). Environmental conditions there (Table 1), i.e., low pH and calcium carbonate and unstable very fine sand and “flashy” river flows, provided less favorable conditions for the Asian clam (Bass and Hitt, 1973). Diptera, Coleoptera, and Hemiptera were the most diverse riverine arthropods. The number of freshwater arthropod taxa outnumbered the exclusively estuarine taxa by a factor of 2.6 (512 or 72% to 195 or 28%). The number of exclusively estuarine Arthropoda, Mollusca, and Annelida taxa was split almost evenly at 8% of the total diversity, while the diversity of these phyla in the river was 60%, 6%, and 5%, respectively. The greater diversity of freshwater invertebrate fauna compared with estuarine is partly due to better taxonomic definition, greater sampling frequency, and greater number of samples collected in the river. The mosaic substrates with overlapping ecotones of the river in contrast to the relative monotypic substrates of the estuary create more niches for the freshwater fauna and increased taxonomic richness. There were 143 dipteran taxa (20% of the total), and the group accounted for 24% of the total arthropod diversity. Alone the Chironomidae accounted for 121 No. 4 1994] MASON ET AL.—BENTHIC INVERTEBRATES IN THE SUWANNEE RIVER 147 (17%) of the total taxa. Our identifications of Chironomidae were enhanced by a recent key (Epler, 1992); however, we were still unable to effectively identify many larval and pupal Orthocladiinae and Tanytarsini. Coleoptera and Hemiptera were the second and third most diverse arthropods to the Diptera. Collection of adult aquatic beetles enabled us a high degree of species-level identifications. Among the major invertebrate macrofaunal groups, the freshwater and estuarine Amphipoda (Crustacea) and Ceratopogonidae (Diptera), and estuarine Gastropoda (Mollusca), are in urgent need of systematics attention. Vertical habitat zone classification—Most invertebrates transition through a series of different life history stages as they mature into adults. Each stage, some lasting for months or years (mollusks) and some lasting just a few days (Diptera), may require a different habitat. Day and co-workers (1989) recognized this problem by classifying the juveniles stages of macrobenthos as “temporary meiofauna.” Thus, for ecological studies, habitat zone classifications for benthic organisms like ours (Fig. 2; Appendix) should be viewed as changing and dynamic. For purposes of this checklist the taxa were grouped according to the dwelling place of the most mature aquatic stage. Most mussel species are under Embenthos although their life history progresses through a hyperbenthic larval stage, epizoos glochidial stage on a fish host (Fig. 2), and, as juveniles, an Epibenthic-Section I existence. Some adult mussels can also move between vertical zones when necessary. For example, depending on environmental conditions, adult Asian clams may move to the loose materials of Epibenthos-Section II, and they may even release and float downstream. However, typically Asian clams anchor to the firm bottom sediments of the Embenthos. Surface Dwelling { %*% [ | Epizoos/Parasites ) At ee rRMIGLY af ee Cas c. Melson Soph C¢L2 “EN Mae - Epibenthos-Section | t Hyperbenthos a UT RE 7) Fee. no o = nar pete RP se ett arnt ‘pibenthos-Section T oe ea ae re _” Embenthos .o SIFT Y¢ . _, Hypobenthos tae i ae Fic. 2. Sketch of a fictional littoral area showing the seven vertical habitat zones. 148 FLORIDA SCIENTIST [VOL 57 Taxonomic notes—Diagnostic characteristics of immature biting midges (Ceratopogonidae: Diptera) are poorly known throughout the Nearctic region, and, as immatures, there are several undescribed species in the Suwannee River. Under Hyperbenthos, Conchapelopia, Hayesomyia, Helopelopia, Meropelopia, and Rheopelopia (Chironomidae: Diptera) in Epler (1992) were previously assigned to Thienemannimyia (Fittkau). In addition, the larval and pupal stages of 8 of the 10 new Tanytarsus species reported in Mattson (1992) have not been associated with the adult. There are perplexing taxonomic dilemmas that offer future challenges, e.g., Chironomus (s.s) in the Appendix; Epibenthos-Section II, is the plumosus group (Diptera: Chironomidae) in North America. This group contains about 8 common species that are difficult to differentiate as larvae. We are uncertain about the identification of several other species in our checklist. Gammarid amphipods occur in both fresh water and brackish water in the Suwannee River basin. The most prevalent one in the freshwater Suwannee River closely fits the description of Gammarus fasciatus Say (Appendix; Hyperbenthos). However, its known range was previously limited to the northeast and therefore its presence in the Suwannee River is questionable. The identity of G. nr. tigrinus that inhabits the brackish water mixing zones of the Suwannee Riveris likewise uncertain. Under Epibenthos-Section IJ, the trichopteran Oecetis (Leptoceridae) is prob- ably composed of seven species, including O. inconspicua (Walker). It is impossible to distinguish these species with existing keys. Similarly, the beetle Desmopachria (Dytiscidae: Coleoptera) keys to convexa. This species, however, is not known in Florida, and our specimens are most likely D. grana (Le Conte)—a complex of subspecies. Some semi-aquatic organisms, e.g., Collembola (7 spp.) and Acarina (18 spp.), were included in the Appendix, although these groups are often omitted from faunal inventories. Collembola venture in and out of the water’s edge to feed on other invertebrates and Acarina prey singly on smaller invertebrates or in packs on larger animals. Thus, these often overlooked organisms are important links in the aquatic food web. According to Thompson (1984), the Gulf Coast pebblesnail Somatogyrus walkerianus (Aldrich) occurs only in western Florida. He noted that “an undescribed species of Somatogyrus occurs in the Apalachicola River.” This undescribed species might be the pebblesnail we found in the Suwannee River. We encourage continued monitoring of the benthic fauna in the ecosystem, and especially along the coastal shelf between the mouth of the Suwannee River and the Cedar Keys. This area is prime habitat for fish foods and spawning grounds for commercial and recreational fish and shellfish. Secondly, better taxonomic defini- tion of benthic invertebrate species of the Northeast Gulf Coast are prerequisite to expanded knowledge about the ecology and life histories of these important organ- isms. ACKNOWLEDGMENTS—We are grateful for the taxonomic assistance of Kurt Auffenberg (Gastropoda), Lewis Berner (Ephemeroptera), Ralph O. Brinkhurst (Oligochaeta), Paul H. Carlson (general tax- onomy), John R. Cox (Nematoda), David L. Evans (Coleoptera), Douglas A. Farrell (Amphipoda), Michael R. Milligan (Polychaeta), William Heard (Decapoda), Donald J. Klemm (Hirudinea), Raymond No. 4 1994] MASON ET AL.—BENTHIC INVERTEBRATES IN THE SUWANNEE RIVER 149 B. Manning (Decapoda), Manuel L. Pescador (Ephemeroptera), James D. Williams (Bivalvia), and F. N. Young (Coleoptera). Also, we thank the many dedicated benthic taxonomists of the region who provided descriptions and keys for the invertebrates. Our special thanks go to James N. Bradner, James P. Clugston, Bruce C. Cowell, James L. Hulbert, Michael R. Milligan, and Manuel L. Pescador for manuscript reviews. LITERATURE CITED AMERICAN PuBLic HEALTH ASSOCIATION (APHA), AMERICAN WATER Works ASSOCIATION, AND WATER POLLUTION CONTROL FEDERATION. 1989. Standard Methods for the Examination of Water and Wastewater. 17th ed. Washington, DC. AUFFENBERG, K. 1993. Malacology Section, Florida Museum of Natural History, Gainesville, FL, Pers. Commun. Barnes, R. D. 1987. Invertebrate Zoology. Saunders College Publishing. 5th ed. Philadelphia, PA. 893 Bass, ial AND V. G. Hirt. 1973. Sport fishery ecology of the Suwannee and Santa Fe Rivers, Florida. Report I. Study 3. Benthic ecology of the Lower Santa Fe River and Lower Suwannee River. Unpublished Report. Florida Game and Fresh Water Fish Commission, Lake City, FL. Benke, A. C., T. C. Van Arspa.t, D. M. GILLEsPiF, AND F. K. Parrisit. 1984. Invertebrate productivity in a subtopical blackwater river. The importance of habitat and life history. Ecol. Monogr. 54:25- 63. CaLpWELL, B. A. AND F. K. Parris. 1987. Chironomidae (Diptera) of small Georgia streams. Entomol. Scand. Suppl. 29:375-379. Ceryak, R., M. S. Knapp, aND T. Q. Burson. 1983. The geology and water resources of the upper Suwannee River Basin, Florida. Rep. Investig. No. 87. Bureau of Geology, Florida Depart. of Natural Resources, Tallahassee, FL. 165 pp. Day, J. W., Jk.,C. A. S. HALL, W. M. Kemp, AND A. YANEZ-ARANCIBIA. 1989. Estuarine Ecology. John Wiley & Sons, New York, NY. 558 pp. EPLeER, J. H. 1992. Identification manual for the larval Chironomidae (Diptera) of Florida. Florida Department of Environmental Regulation, Orlando, FL. 302 pp. FarrELL, D. A. 1993. Florida Department of Environmental Protection, Surface Water management, Tampa, FL., Pers. Commun. FLORIDA DEPARTMENT OF ENVIRONMENTAL REGULATION. 1985. Limnology of The Suwannee River, Florida. Division of Environmental Programs, Tallahassee, FL. 330 pp. FLORIDA DEPARTMENT OF NATURAL REsourcEs. 1989. Florida Rivers Assessment. Division of Recreation and Parks, Tallahassee, FL. 452 pp. Fox, R. S. anp K. H. Bynum. 1975. The amphipod crustaceans of North Carolina estuarine waters. Chesapeake Sci. 16:223-237. Heatu, R. C. anp C. S. Conover. 1981. Hydrologic Almanac of Florida. Open file Report 81-1107. U. S. Geological Survey, Tallahassee, FL. HoimeE, N. A. AND A. D. Mc Intyre. 1971. Methods of Study of Marine Benthos. IBP Handbook 16. Blackwell Scientific Publications, Oxford, England; distributed by F. A. Davis Co., Philadelphia, PA. 334 pp. Kem, D. J., P. A. Lewis, F. FULK, AND J? M. Lazorcuak. 1990. Macroinvertebrate field and laboratory methods for evaluating the biological integrity of surface waters. EPA/600/4-90/030. Environ- mental Monitoring Systems Laboratory, U. S. Environmental Protection Agency, Cincinnati, OH. 256 pp. LarsEN, P. F. 1974. A remotely operated shallow water benthic suction sampler. Chesapeake Sci. 15:176- 178. Lewis, F. G., III. 1984. Distribution of macrobenthic crustaceans associated with Thalassia, Halodule, and bare sand substrata. Marine Ecol.-Prog. Ser. 19: 101-113. Mason, W.T., JR. AND P. P. Yevicu. 1967. The use of phloxine B and rose bengal stains to facilitate sorting benthic samples. Trans. Amer. Microsc. Soc. 86:221-223. . 1991. A survey of benthic invertebrates in the Suwannee River, Florida. Environ. Monit. and Assess. 16:163-187. AND S. A. ZENGEL. Unpublished Data. Benthic invertebrates in seagrasses at the Cedar Keys National Wildlife Refuge, Florida. National Biological Survey, USDI, Gainesville, FL. Mattson, R. A. 1992. Characteristics of benthic macroinvertebrate communities of the Suwannee River Drainage. Unpublished Report. Suwannee River Water Management District, Live Oak, FL. 188 pp. + Appendix. 150 FLORIDA SCIENTIST [VOL 57 MEES, J. AND O. HAMERLYNCK. 1992. Spatial community structure of the winter hyperbenthos of the Schelde Estuary, The Netherlands, and adjacent coastal waters. Netherlands J. Sea Res. 29:357- 370. PULLEN, E. J.,C. R. Mock, anp R. D. Rinco. 1968. A net for sampling the intertidal zone of an estuary. Limnol. Oceanogr. 13:200-202. Rosack, S. S. 1953. Savannah River tendipedid larvae (Diptera: Tendipedidae= Chironomidae). Acad. Nat. Sci. Philadelphia 55:91-132. Soponis, A. R. 1980. Taxonomic composition of Chironomidae (Diptera) in a sand-bottomed stream of Northern Florida. Pp. 163-169. In: Murry, D. A. (ed.) Chironomidae Ecology, Systematics, Cytology, and Physiology. Pergamon Press, New York, NY. STONER, A. W. 1979. The macrobenthos of seagrass meadows in Apalachee Bay, Florida, and the feeding ecology of Lagondon rhombiodes (Pices: Sparidae). Thesis. The Florida State Univ., Tallahassee, FL. 175 pp. a PRON id B., W. L. Kruczynsk1, AND S. H. Drake. 1976. Studies on the animal communities in two North Florida salt marshes. Part II. Macroinvertebrtate communities. Bull. Mar. Sci. 26:172-195. Tuompson, F. G. 1984. Freshwater Snails of Florida: A Manual for Identification. University of Florida Press, Gainesville, FL. 94 pp. Wo r, S. H. anp L. E. Wo-r. 1985. The ecology of the Suwannee River Estuary: An analysis of ecological data from the Suwannee River Water Management District study of the Suwannee River Estuary 1982-1983. Unpublished Report. Suwannee River Water Management District, Live Oak, FL. 118 pp. Florida Scient: 57(4):141-160. 1994. Accepted: May 13, 1994 APPENDIX Benthic invertebrate taxa and allied macrofauna in vertical habitat zones of the Suwannee River and estuary, Florida. Source references follow the author of taxa as follows: 1- Florida Department of Environmental Regulation, 1985; 2- Mason, 1991; 3- Mattson, 1992; 4- Wolf and Wolf, 1985; and 5- Mason and Zengel, unpublished data. An asterisk (*) precedes the scientific name of the exclusively estuarine taxa. SURFACE-DWELLING FAUNA Cnidaria - Coelenterates Entomobryidae Entomobrya Rondani 3 Sipl hora - Port se Man-of-W:z iphonophora - Portuguese Man-of-War Salina MacCillivary 3 Physaliidae eee Sinella Brook 3 *Physalia physalia (Linnaeus) 5 Willowsia Shoebotham 3 Scyphozoa - Scyphozoans Isotomidae Isotoma Bourlet 3 Isotomurus Borner 3 Poduridae Podura aquatica Linnaeus 3 Rhizostomeae - Jellyfishes Stomolophidae - Cannonball jellyfish *Stomolophus meleagris L. Agassiz 5 Arthropoda Sminthuridae Araneae - Semiaquatic Spiders Bourletiella Banks 3 Pisamiidae Sminthurus Borner 3 Dolomedes okefinokensis Bishop 3 Diptera - True Flies D. triton (Walckenaer) 3 Culicidae - Mosquitoes Aedes vexans (Meigen) 3 Insecta - Insects E Anopheles Meigen 3 Collembola - Springtail ollembola - Springtails Ceratopogonidae Alluaudomyia Kieffer 3 No. 4 1994] | MASON ET AL—BENTHIC INVERTEBRATES IN THE SUWANNEE RIVER 151 Coleoptera - Beetles Gyrinidae - Whirligig Beetles Dineutus angustus LeConte 3 D. carolinus LeConte 3 D. ciliatus (Forsberg) 3 D. discolor Aube 3 D. serrulatus LeConte 3 Gyrinus analis Say 3 G. elevatus LeConte 3 G. pachysomus Fall 3 G. woodruffi Fall 3 Hemiptera - True Bugs Corixidae - Water Boatmen Hesperocorixa Kirkaldy 3 Sigara zimmermanni (Fieber) 3 Trichocorixa kanza Sailer 3 T. louisianae Jaczewski 3 T. minima (Abbott) 3 T. sexcincta (Champion) 3 Gerridae - Water Striders Gerris conformis (Uhler) 3 G. nebularis Drake and Hottes 3 Limnoporus canaliculatus (Say) 3 Metrobates hesperius Uhler 3 Neogerris hesione Kirkaldy 3 Rheumatobates tenuipes Meinert 3 Hebridae - Velvet Water Bugs Merragata brunnea Drake 3 Hydrometridae - Marsh Treaders Hydrometra australis Say 3 Mesoveliidae - Water Treaders Mesovelia amoena Uhler 3 M. mulsanti White 3 Notonectidae - Back Swimmers Buenoa margaritacea Torre-Bueno 3 Notonecta irrorata Uhler 3 Nepidae - Water Scorpions Ranatra australis Hungerford 3 R. buenoi Hungerford 3 R. drakei Hungerford 3 R. fusca Palisot de Beauvois 3 R. kirkaldyi Torre-Bueno 3 R. nigra Herrich-Schaeffer 3 Pleidae - Pygmy Backswimmers Neoplea striola Fieber 3 Saldidae - Shore Bugs Micracanthia Reuter 3 Saldula Van Duzee 3 Veliidae - Broad-Shouldered Water Striders Microvelia Westwood 3 Paravelia brachialis (Stal) 3 Rhagovelia choreutes Hussey 3 R. obesa Uhler 3 HyYPERBENTHOS Arthropoda Crustacea Amphipoda - Free Living Amphipods Ampeliscidae *Ampelisca vadorum Mills 3, 4 °A. verrilli Mills 3, 4 Ampithoidae *Ampithoe longimanna Smith 5 *Cymadusa compta (Smith) 4, 5 Aoridae *Grandidierella bonnieroides Myers 4 *Microdeutopus Costa 4 *Unicola dissimilis Shoemaker 2 Crangonycidae Crangonyx serratus (Embody) 3 Dexaminidae *Polycheria 5 Gammaridae °G. mucronatus Say 2, 4,5 *G. nr. tigrinus Sexton 2, 5 G. cf. fasciatus Say 1, 2,5 Haustoriidae *Parahaustorius cf. longimerus Bousfield 2,5 Hyalidae *Hyale plumosa (Stimpson)? 3 *Lyanassidae 2 Melitidae * Melita nitida Smith 3, 5 Oedicerotidae *Monoculodes nyei Shoemaker 5 Talitridae Hyallela azteca (Saussure) 2 Orchestia uhleri Shoemaker? 3 Decapoda - Decapods, Shellfish, Shrimps Dendrobranchiata: Penaeidae - Penaeid Shrimps * Penaeus duorarum Burkenroad - Pink Shrimp 4, 5 *P. setiferus (Linnaeus) - White Shrimp 45 Sergestidae * Acetes carolinae 4 Pleocyemata: Caridea Alpheidae - Snapping Shrimps *Alpheus heterochaelis Say 4, 5 Hippolytidae - Grass shrimps *Hippolyte pleuracanthus (Stimpson) 4, 5 *Latreutes fucorum (Fabricus) 4, 5 °L. parvulus (Stimpson) 4, 5 *Tozeuma caroliense Kingsley 4, 5 *Thor dobkini Chace 5 152 FLORIDA SCIENTIST Palaemonidae - Prawns, Grass Shrimps *Palaemon floridanus Chace 5 * Palaemonetes pugio Holthuis 4, 5 P. vulgaris Say 3,5 *Periclimenes iridescens Lebour 4, 5 *P. longicaudatus (Stimpson) 4, 5 Processidae - Deep-Water Shrimps *Ambidexter symmetricus - Manning and Chace 2, 5 Pleocyemata: Brachyura Portunidae - Swimming Crabs *Callinectes sapidus Rathbun - Blue Crab 4,5 *Portunus Weber 3, 5 Mysidacea - Grass Shrimps Mysidae - Mysids *Mysidopsis almyra Bowman 4, 5 *M. bahia Molenock 4,5 Taphromysis bowmani Bacescu 4, 5 °T. louisianae Banner 2 Insecta Coleoptera Curculionidae - Weevils Lissorhoptrus LeConte 3 Listronotus Jekel 1, 3 Onychylis nigrirostris (Boheman) complex 3 Stenopelmus rufinasus Gyllenhal 3 Tanysphyrus lemnae (Fabricus) 3 Hemiptera Naucoridae - Creeping Water Bugs Pelocoris carolinensis Torre-Bueno 3 P. femoratus Palisot de Beauvois 3 EPIBENTHOS-SECTION I Bryozoa - Bryozoans, Moss Animals Pectinatellidae Pectinatella magnifica Leidy 2 Cnidaria - Coelenterates Hydrozoa - Hydras Hydroidea Clavidae Cordylophora lacustris Allman 5 Hydridae Hydra americana Hyman? 2 Porifera - Sponges *Demospongiae 5 *Calcarea - Purse Sponges 5 [VOL 57 Tunicata (=Urochordata) - Tunicates Didemnidae *Didemnum duplicatum F. Minniot 5 Perophoridae * Ecteinascidia turbanata Herdman 5 Polyclinidae *Aplidium constellatum Verrill 5 Polycitoridae *Clavelina 5 *Distaplia bermudensis van Name 5 *Eudistoma hepaticum 5 *E. carolinense van Name 5 Nemertea - Nemerteans Prostoma Duges 1, 4 Mollusca Bivalvia - Mussels and Oysters (part) Dreissenidae *Ischadium recurvum (Rafinesque) 5 Mytilopsis leucophaeata (Conrad) 5 *Parastarte triquetra Conrad 5 Ostreidae *Crassostrea virginica (Gmelin) 5 - Eastern oyster Arthropoda Insecta Coleoptera - Beetles Psephenidae - Riffle Beetles, Water Pennies Ectopria LeConte 3 Diptera - True Flies Chironomidae - Chironomids, Non- biting Midges Chironominae: Chironomini Apedilum elachistus Townes 1, 3 Asheum beckae (Sublette) 3 Axarus Roback 3 Beardius truncatus Reiss and Sublette 3 Dicrotendipes Kieffer 1 D. lucifer (Johannsen) 3 D. modestus (Say) 2 D. neomodestus (Malloch) 2 D. nervosus (Staeger)? 2 D. simpsoni Epler 3 D. tritomus Kieffer? 3 Endochironomus nigricans (Johannsen) 2 E. subtendens (Townes) 3 Endotribelos hesperium (Sublette) 3 Glyptotendipes Kieffer 3 Harnischia Kieffer 1 Microtendipes pedellus (de Geer) group 1 Nilothauma Kieffer 3 Omisus Townes 3 No. 4 1994] Pagastiella Brundin 3 Parachironomus carinatus (Townes) 2 P. frequens (Johannsen) 1 P. schneideri Beck and Beck 1 Paralauterborniella nigrohalterale (Malloch) 1, 2 Phaenospectra punctipes (Wiedemann) group 3 P. obediens (Johannsen) group 1 Polypedilum aviceps Townes 3 P. convictum (Walker) 2 P. fallax (Johannsen) 3 P. halterale (Coquillett) 2 P. illinoense (Malloch) 2 P. scalaenum (Schrank) 2 P. trigonus (Townes) 3 P. tritum (Walker) 3 . Stelechomyia perpulchra (Mitchell) 2, 3 Stenochironomus Kieffer 2 Tribelos fuscicorne (Malloch) 3 T. jucundum (Walker) 2 Chironominae: Pseudochironomini Pseudochironomus richardsoni Malloch 2 Chironominae: Tanytarsini Cladotanytarsus Kieffer 2 Micropsectra Kieffer 2 Nimbocera Reiss 2 Paratanytarsus Thienemann and Bause 2 Rheotanytarsus Thienemann and Bause 2 Stempellina Thienemann and Bause 2 Stempellinella Brundin 2 Tanytarsus v.d. Wulp (8 spp.) 2, 3 T. cf. buckleyi Sublette 3 T. quadratus Sublette 3 Diamesinae Potthastia longimana Kieffer group 3 Orthocladiinae Cardiocladius Kieffer 1 Corynoneura Winnertz 2 Cricotopus/Orthocladius complex. v.d. Wulp 3 C. bicinctus (Meigen) 2 C. nostocicola Wirth? 2 C. politus Coquillett 3 Eukiefferiella claripennis (Lundbeck) group 3 Hydrobaenus Fries 3 Lopescladius Oliveria 3 Nanocladius balticus (Palmen) group Laz N. crassicornus Saether 3 N. cf. rectinervis (Kieffer) 3 Orthocladius (Orthocladius) annectens Saether 2 Parakiefferiella Thienemann 3 Parametriocnemus Goetghebuer 3 Psectrocladius elatus Roback 3 MASON ET AL.—BENTHIC INVERTEBRATES IN THE SUWANNEE RIVER 153 Rheocricotopus robacki (Beck and Beck) 3 Rheosmittia Brundin 3 Stilocladius Rossaro? 3 Synorthocladius Thienemann 3 Tvetenia discoloripes (Goetghebuer) group 1, 2 Unniella multivirga Saether 2, 3 Zalutschia Lipina 3 Lepidoptera - Moths/Butterflies Pyralidae Parapoynx Hubner 3 Trichoptera Hydropsychidae Cheumatopsyche Wallengren 3 Hydropsyche simulans Ross 3 Macrostemum carolina (Banks) 3 Polycentropidae Neureclipsis McLachlan 2 Nyctiophylax Brauer 3 Phylocentropus placidus (Banks) 2 Polycentropus Curtis 2 Psychomyiidae Cyrnellus fraternus (Banks) 2 Lype diversa (Banks) 2 EPIBENTHOS-SECTION II Echinodermata Holothuroidea - Sea Cucumbers *Leptosynapya parvipatina 4 *Sclerodastyla brairens 5 Stelleroidea - Starfishes Ophiurida Ophiactidae *Ophiactis rubropoda Singletary 5 Amphiuridae *Amphioplus abditus 4 °A. pulchella 4 *A. thrombiodes 4 *Ophiothrix angulata (Say) 5 Mollusca Bivalvia - clams, mussels Sphaeriidae Sphaerium Scopoli 2 Musculium lacustre (Muller) 2 Pisidium dubium (Say) 3 Gastropoda - Snails Acteonidae *Acteon punctostriatus (C. B. Adams) 5 Ancylidae - Limpets Ferrissia hendersoni Walker 2 Hebetancylus excentricus (Morelet) 3 Laevapex Walker 3 154 FLORIDA SCIENTIST Atyidae ® Haminoea succinea (Conrad) 5 Bullidae *Bulla striata Bruguiere 5 Caecidae *Caecum pulchellum Stimpson 5 °C. vesatitum Folin 5 Cerithiidae *Cerithium muscarum Say 5 *Cerithiopsis greeni (C. B. Adams) 5 Columbellidae *Anachis semiplicata Abbott 5 *Mitrella lunata (Say) 5 Conidae *Conus stearnsi Conrad 5 Crepidulidae *Crepidula maculosa Conrad 5 Ellobiidae *Detracia floridana (Pheiffer) 5 Hydrobiidae Amnicola Gould 2 Cincinnatia floridana (Frauenfeld) 2 Notogillia wetherbyi (Dall) 2 Somatogyrus walkerianus (Aldrich)? 9 y) Spilochlamys conica Thompson 2 Epitoniidae - Wentletraps *Epitonium angulatum (Say) ) Fasciolariidae - Tulip shells *Fasciolaria hunteria 5 Littorinidae - Periwinkles *Littorina irrorata (Say) 3 Lymnaeidae Fossaria modicella (Say) 1 Pseudosuccinea columella (Say) 3 Marginellidae *Granulina ovuliformis (d’Orbigny) 5 *Hyalina veliei (Donovan) 3 *Marginella apicina Menke 5 °M. lavalleeana Orbingy 5 Melongenidae - Conchs, Whelks *Busycon sp. Roding - Whelk 5 *B. spiratum (Lam.) - Fig Whelk 5 *Melongena corona (Gmelin) - Crown Conch 5 Modulidae - Button Snails * Modulus modulus (Linnaeus) 5 Nassariidae - Mudsnails, Nassas *Nassarius albus (Say) 5 °N. vibex (Say) 5, Naticidae - Moonsnails *Polinices duplicatus (Say) 5 Olividae - Olive Snails *Olivella mutica (Say) 5 Neritidae - Nerites *Neritina reclivata (Say) 2 Physidae - Pond Snails Physella Haldeman 2 Pilidae - Apple Snails Pomacea paludosa (Say) 2 Planorbidae - Orb Snails Gyraulus parvus (Say) 2 Micromenetus dilatatus avus (Pilsbry) 2 M. dilatatus dilatatus (Gould) 2 M. floridensis (Baker) 3 Planorbella duryi (Weatherby) 2 P. scalaris (Jay) 3 P. trivolvis intertexta (Jeffreys) 3 Pleuroceridae - River Snails Elimia floridensis (Reeve) 2 E. athearni (Clench and Tumer) 2 Pyramidellidae *Sayella hemphilli (Dall) 5 Retusidae *Retusa sulcata (Orbingy) 5 Siphonodentaliidae *Cadulus carolinensis Bush 5 Siphonariidae - False Limpets *Siphonaria pectinata (Linnaeus) 5 °S. alternata (Say) 5 Terebridae - Auger Snails (part) °Terebra floridana Dall? 5 Turbinidae - Turban Snails *Turbo castanea Gmelin 5 Turridae - Auger Snails (part) *Mangelia biconica C. B. Adams 5 *M. cf. ceroplasta Bush 5 Viviparidae Campeloma floridense Call 2 C. limum (Anthony) 2 Lioplax pilsbryi choctawhatchensis Vanatta 2 Viviparus georgianus (Lea) 3 *Nudibranchia - Sea Slugs 5 Platyhelminthes - Flatworms Dugesia polychroa (O. Schmidt) 3 D. tigrina (Girard) 3 Arthropoda Arachnoidea Xiphosura - Horseshoe Crabs *Limulus polyphemus Linnaeus 5 Crustacea Malacostraca Amphipoda - Amphipods, Scuds Corophiidae - Tube Dwelling Amphi- pods [VOL 57 No. 4 1994] MASON ET AL.—BENTHIC INVERTEBRATES IN THE SUWANNEE RIVER 155 *Corophium cf. insidiosum Crawford 2 °C. cf. tuberculatum Shoemaker 2 °C. cf. lacustre Vanhoffen 2 *Cerapus cf. tubularis Say 2 Ischyroceridae - Tube Dwelling Amphi- pods *Erichthonius brasiliensis 4 *Jassa falcata Smith 2, 4 Isopoda - Isopods, Aquatic Sow Bugs Asellota ‘Asellidae Asellus racovitzai Williams 2 Lirceus Rafinesque 3 Gnathiidea Anthuridae *Cyathura polita Stimpson 2, 4,5 Valvifera Idoteidae *Chiridotea Harger 3 *Edotea cf. montosa (Stimpson) 3, 4 *Erichsonella attenuata 4 *Idotea baltica (Pallas) 4 Flabellifera Sphaeromidae °*Sphaeroma quadridentatum Say 3 *Cassidinidea ovalis (Say) 3 Tanaidacea (= Chelifera) Paratanaidae *Hargeria rapax (Harger) 4,5 Decapoda - Shellfish, Decapods Cambaridae - Crayfishes Procambarus fallax (Hagen) 5 P. paeninsulanus (Faxon) 2 P. youngi Hobbs 3 Grapsidae - Warf Crabs *Sesarma cinereum (Bosc) 3, 5 Ocypodidae - Fiddler Crabs *Uca minax LeConte 3,5 Paguridae - Hermit Crabs *Pagurus Say 4,5 Pinnotheridae - Commensal Crabs *Pinnixa chaetopterana Stimpson 4 °P. cylindrica (Say) 4 °P. pearsi Wass 4 *P. retinens Rathbun 4 *P. sayana Stimpson 4 Xanthidae - Mud Crabs *Panopeus texana (Stimpson) 5 *Rhithropanopeus depressus 5 *R. harrisii (Gould) 2 Cumacea - Cumaceans Nannasticidae *Almyracuma sp. A Heard 4, 5 *Oxyurostylis smithi Calman 4, 5 Cirripedia - Barnacles Thoracica - Acorn Barnacles Balanidae *Balanus eburneus Gould? 5 Insecta Coleoptera Chrysomelidae - Leaf Beetles Agasicles hygrophila Selman and Vogt 3 Dryopidae - Dryopid Beetles Helichus lithophilus (Germar) 3 Pelonomus obscurus LeConte 3 Dytiscidae - Diving Beetles Acilius fraternus (Harris) 3 Agabus johannis Fall 3 Bidessonotus longovalis (Blatchley) 3 B. pulicarius (Aube) 3 Celina Aube 3 Coptotomus interrogatus (Fabricius) | C opelatus caelatipennis princeps Young 3 C. chevrolati chevrolati Aube 3 Cybister fimbriolatus (Say) 3 Desmopachria grana (LeConte) complex 3 Hydaticus bimarginatus (Say) 3 Hydroporus carolinus Fall? 3 H. clypealis Sharp 3 H. dixianus Fall 3 H. hybridus Aube 3 H. lobatus Sharp 3 H. undulatus Say? 3 H. venustus LeConte 3 H. vittatipennis Gemminger and v. Harold 3 Hydrovatus pustulatus compressus Sharp 3 Laccophilus fasciatus Aube 3 L. gentilis LeConte 3 L. proximus Say 3 Lioporius pilatei (Fall) 3 Matus ovatus blatchleyi Leech 3 Thermonectes basillaris (Harris) 3 Elmidae - Elmid Beetles Ancyronyx variegatus (Germar) 3 Dubiraphia vittata (Melsheimer) 1, 2 Macronychus glabratus Say 3 Microcylloepus pusillus lodingi (Musgrave) 2 Stenelmis antennalis Sanderson 3 S. convexula Sanderson 3 S. crenata (Say) 3 S. fuscata Blatchley 2 S. hungerfordi Sanderson 2 S. lignicola Schmude and Brown 3 S. sinuata LeConte 3 Hydraenidae Hydraena Kugelmann 3 FLORIDA SCIENTIST Haliplidae - Crawling Water Beetles Haliplus annulatus Roberts 3 H. fasciatus Aube 3 H. punctatus Aube 3 Peltodytes floridensis Matheson 3 P. lengi Roberts 3 P. oppositus Roberts 3 P. sexmaculatus Roberts 3 Hydrochidae Hydrochus foveatus Haldeman? 3 H. inaequalis LeConte 3 H. rugosus Mulsant? 3 Hydrophilidae - Water Scavenger Beetles Berosus aculeatus LeConte 3 B. exiguus (Say) 3 B. infuscatus LeConte 3 B. striatus (Say) 3 B. peregrinus (Herbst) 3 Cymbiodyta vindicata Fall 3 Derallus altus (LeConte) 3 Enochurus blatchleyi (Fall) 3 E. cinctus (Say) 3 E. consors (LeConte) 3 E. ochraceus (Melscheimer) 3 E. pygmaeus nebulosus (Say) 3 Helobata striata (Brulle) 3 Helochares maculicollis Mulsant 3 Hydrobiomorpha casta (Say) 3 Paracymus nanus (Fall) 3 Sperchopsis tessellatus (Ziegler) 3 Tropisternis blatchleyi D’Orchymont 3 T. collaris striolatus (LeConte) 3 T. lateralis nimbatus (Say) 3 Noteridae - Burrowing Water Beetles Hydrocanthus oblongus Sharp 3 H. regius Young 3 Notomicrus nanulus (LeConte) 3 Suphis inflatus (LeConte) 3 Suphisellus gibbulus (Aube) 3 S. puncticollis (Crotch) 3 Scirtidae (= Helodidae) Cyphon Paykull 3 Prionocyphon Redtenbacher 3 Scirtes Illiger 3 Diptera Athericidae - Snipe Flies Atherix lantha Webb 3 Ceratopogonidae - Biting Midges, No- See-Ums Atrichopogon Kieffer 3 Bezzia Kieffer/Palpomyia Meigen complex 2 Culicoides Latreille 3 Dasyhelea Kieffer 3 Forcipomyia Meigen 3 Mallochohelea Wirth? 3 Probezzia Kieffer 3 Phaenobezzia opaca (Loew) 3 Sphaeromias Curtis 3 [VOL 57 Stilobezzia Kieffer 3 Chironomidae - Non-biting Midges Chironominae: Chernovskiia Saether 2 Chironomus (s.s.) Meigen 3 C. decorus Johannsen group 2, 3 Cladopelma Kieffer 3 Cryptochironomus fulvus (Johannsen) group 2, 3 C. blarina Townes 2 Demicryptochironomus cuneatus (Townes) 2 C ryptotendipes casuarius (Townes) 2 Goeldichironomus carus (Townes) 1 G. holoprasinus (Goeldi) 3 Kiefferulus dux (Johannsen) 2 Lauterborniella agrayloides (Kieffer) 2 Zavreliella marmorata (v.d. Wulp) 1, 3 Orthocladiinae Cricotopus sylvestris (Fabricus) group sp. 3 Tanypodinae: Ablabesmyia annulata (Say) 3 A. cinctipes (Johannsen) 3 A. mallochi (Walley ) 2 A. peleensis (Walley) 3 A. rhamphe Sublette group 3 Clinotanypus Kieffer 2 Coelotanypus concinnus (Coquillett) 3 C. scapularis (Loew) 2 Conchapelopia Fittkau 2 Djalmabatista pulcher (Johannsen) 1 Hayesomyia senata (Walley) 3 Labrundinia becki Roback 3 L. johannseni Beck 3 L. neopilosella Beck and Beck 3 L. pilosella (Loew) 2, 3 Larsia berneri Beck and Beck 2, 3 Meropelopia Roback 3 Natarsia sp. A Roback 3 Nilotanypus fimbriatus (Walker) 3 Pentaneura inconspicua (Malloch) 2 Procladius Skuse 2 Rheopelopia Fittkau 3 Tanypus carinatus Sublette 3 Empididae - Shore Flies Hemerodromia Meigen 3 Muscidae Limnophora Robineau-Desvoidy is Psychodidae - Sewage Flies Psychoda Latreille 1 Simuliidae - Black Flies Simulium Latreille 2 Stratiomyidae - Soldier Flies Odontomyia Meigen 3 Tabanidae - Horse Flies Chlorotabanus crepuscularis (Bequaert) 3 Chrysops Meigen 3 Tabanus Linnaeus 3 Tipulidae - Crane Flies No. 4 1994] Geranomyia Haliday 3 Limonia Meigen 3 Ormosia Rondani 3 Prionocera Loew 3 Tipula Linnaeus 2 Corydalidae - Hellgrammites Chauliodes rastricornis Rambur 2 Corydalus cornutus (Linnaeus) 3 Sialidae Sialis Latreille 3 Neuroptera - Lacewings Sisyridae - Spongilla Flies Climacia areolaris (Hagen) 3 Trichoptera - Caddisflies Helicopsychidae Helicopsyche von Siebold 3 Hydroptilidae Hydroptila Dalman 2 Mayatrichia ayama Mosely 3 Neotrichia Morton 2 Ochrotrichia Mosely 2 Orthotrichia Eaton 3 Oxyethira Eaton 2 Leptoceridae Ceraclea Stephens 3 Mystacides Berthold 3 Nectopsyche candida (Hagen) 3 N. exquisita (Walker) 3 N. pavida (Hagen) 3 Oecetis inconspicua complex JL 22) Triaenodes McLachlan 3 Limnephilidae Hydatophylax argus (Harris) 3 Pycnopsyche Banks 3 Molannidae Molanna tryphena Betten 3 Philopotamidae Chimarra florida Ross 3 C. moselyi Denning 3 Rhyacophillidae Rhyacophila Pictet 3 Ephemeroptera - Mayflies Baetidae Acerpenna pygmaea (Hagen) 2, 3 Baetis alachua (Bemer) 3 B. armillatus Waltz and McCafferty 3 B. ephippiatus Traver 3 B. frondalis McDunnough 3 B. intercalaris McDunnough 2 B. punctiventris (mCDunnough) 3 B. propinquus (Walsh) 2 Callibaetis floridanus Banks 2 C. pretiosus Banks 3 Procloeon hobbsi (Berner) 3 P. viridocularis (Berner) 2 P. rubropictum (McDunnough) 1 Baetiscidae Baetisca obesa (Say) 3 B. rogersi Berner | MASON ET AL.—BENTHIC INVERTEBRATES IN THE SUWANNEE RIVER 157 Caenidae Brachycerus maculatus Berner 2 Caenis amica Hagen 3 C. diminuta Walker 3 Cercobrachys etowah Soldan 3 Ephemerellidae Eurylophella temporalis (McDunnough) 1,3 Heptageniidae Stenonema exiguum Traver 2 S. mexicanum integrum (McDunnough) 2, S. smithae Traver 2 Stenacron interpunctatum (Say) 2 S. floridense (Lewis) 2 Isonychiidae Isonychia arida (Say) 1 Leptophlebiidae Choroterpes hubbelli Berner 2 Leptophlebia bradleyi Needham 3 L. intermedia (Traver) 3 Paraleptophlebia volitans (McDunnough) 3 Metretopodidae Siphloplecton Clemens 3 Neoephemeridae Neoephemera compressa Berner 1, 3 Tricorythidae Leptohyphes dolani Allen 3 Tricorythodes albilineatus Berner 2 Plecoptera - Stoneflies Acroneuridae Acroneuria arenosa (Pictet) 2 Attaneuria ruralis (Hagen) 3 Perlidae Neoperla clymene (Newman) 2 Paragnetina kansensis (Banks) 3 Perlesta placida (Hagen) cmplx. 2 Perlinella Banks 3 Perlodidae Hydroperla phormidia Ray and Stark 3 Taeniopterygidae Taeniopteryx lita Frison 3 Odonata: Anisoptera - Dragonflies Aeshnidae Anax junius (Drury) 3, 4 Boyeria vinosa (Say) 3 Nasiaeschna pentacantha (Rambur) 3 Cordulidae Epitheca Burmeister 4 E. princeps Hagen 4 Neurocordulia virginiensis Davis 2 Gomphidae Aphylla willimsoni Gloyd 4 Arigomphus pallidus (Rambur) 2, 3 Dromogomphus armatus Selys 3 D. spinosus Selys 2 Gomphus cavillaris Needham 3 G. dilatatus Rambur 3 Progomphus obscurus Rambur 3 158 FLORIDA SCIENTIST [VOL 57 Hagenius brevistylus Selys 4 Libellulidae Libellula Linnaeus 3 Perithemis tenera Say 3 Erythemis simplicicollis (Say) 3 Macromiidae Didymops transversa (Say) 3 Macromia georgina (Selys) 1, 2 M. illinoiensis Walsh 3 Odonata: Zygoptera - Damselflies Calopterygidae Calopteryx maculata (Beauvois) 1 Hetaerina titia (Drury) 3 Coenagrionidae Argia apicalis (Say) 3 A. fumipennis (Burmeister) 3 A. moesta (Hagen) 3 A. sedula (Hagen) 2 A. tibialis (Rambur) 3 Enallagma Charpentier 3 Ischnura Charpentier 3 Lestidae Lestes vigilax Hagen 2, 3 Hemiptera Belostomatidae - Giant Water Bugs Belostoma lutarium (Stal) 3 B. testaceum (Leidy) 3 Lethocerus uhleri (Montandon) 3 EMBENTHOS Brachiopoda (Lamp Shells) Inarticulata Lingulida Lingulidae *Glottidia pyramidata (Stimpson) 4 Annelida - Segmented Worms Aphanoneura Aelosomatidae Aelosoma Ehrenberg 3 Oligochaeta - Oligochaetes Naididae Allonais pectinata (Stephenson) 3 Bratislavia unidentata (Harman) 3 Chaetogaster diaphanus (Gruithuisen) 3 Dero furcata (Muller) 2 D. lodeni Brinkhurst 3 D. nivea Aiyer 3 D. pectinata Aiyer 3 D. trifida Loden 3 D. vaga (Leidy) 3 Haemonais waldvogeli Bretscher 2 Nais barbata Muller 1 N. behningi Michaelsen 3 N. communis Piguet 2 N. elinguis Muller 3 N. pardalis Piguet 1 N. simplex Piguet 1 N. variabilis Piguet 3 Pristina leidyi Smith 3 P. synclites Stephenson 3 Pristinella osborni (Walton) 1 Slavina appendiculata (d’ Udekem) 1 Stylaria lacustris (Linnaeus) 2 Stephensoniana Cernosvitov 2 Branchiobdellidae 2 Enchytraeidae 3 Opistocystidae Crustipellis tribranchiata (Harman) 3 Tubificidae - Tubificids, Sludgeworms Aulodrilus americanus Brinkhurst and Cook 2 A. pigueti Kowalewski 3 Haber cf. speciosus (Hrabe) 3 Ilyodrilus templetoni (Southern) 3 Isochaetides freyi (Brinkhurst) 3 Limnodrilus hoffmeisteri Claparede 2 Psammoryctides convolutus Loden 3 Spirosperma ferox Eisen 3 Lumbriculidae - Lumbriculids, Earthworms Eclipidrilus palustris (Smith) 4 Stylodrilus heringianus Claparede? 3 Lumbriculus variegatus (Muller) 2 Polychaeta - Polychaetes, Marine Bloodworms Ampharetidae *Hobsonia florida (Hartman) 2 *Isolda pulchella Muller 4 *Melinna Malmgren 4 * Sabellides Milne 4 Capitellidae *Capitella capitata (Fabricus) 4 *Heteromastus filiformis (Claparede) 4 * Mediomastus Hartman 4 *Notomastus Sars 4 *Polydora Bosc 4 Cirratulidae *Tharyx Webster and Benedict 4 Dorvilleidae *Schistomeringos rudolphii (delle Chiaje) 4 *Eunicidae 4 Glyceridae *Glycera Savigny 4 *Glycinde Muller 4 Hesionidae *Gyptis Marion and Bobretzky 4 *Hesione Savigny 4 *Parahesione Pettibone 4 Lumbrineridae *Tumbrineris Blainville 4 Magelonidae *Magelona Muller 4 No. 4 1994] | MASON ET AL.—BENTHIC INVERTEBRATES IN THE SUWANNEE RIVER 159 Maldanidae Mollusca *Maldane Grube 4 Bivalvia - Clams and Mussels *Axiothella Verill 4 Arablemidac *Clymenella torquatus (Leidy) 4 Quincuncina infucata (Conrad) 3 Nephtyidae Carditidae ‘Aglaophamus Kinberg 4 *Carditamera floridana Conrad 4 Nephtys Cuvier 4 Corbiculidae - Marsh clams Nereidae Corbicula fluminea (Muller) - Asian *Kinbergonuphis Pettibone 4 Abra one *Laeonereis culveri (Webster) 2, 4 *Polymesoda caroliniana (Bosc) - °Nereis succinea (Frey and Leuckart) @aolineuaashiclanc! 2, + Donacidae N. occidentalis 2, 4 *Donax variabilis Say 4 °Platynereis Kinberg 4 Mactrdae Onuphidae *Rangia cuneata 4,5 *Diopatra Audouin and Milne Nuculanidae Edwards 4 *Nuculana acuta (Conrad) 4, 5 *Onuphis eremita Audouin and Milne . Tellindae Edwards 4 °Macoma tenata (Say) 4 Opheliidae Unionidae °*Polyophthalmus pictus 4 Elliptio icterina (Lea) 2 Orbintidae Lampsilis claibornensis (Lea) 2 Haploscoloplos Monro 4 Villosa villosa (Wright) 3 °Orbinia Quatrefages 4 Veneridae - Quahogs *Scoloplos Blainville 4 ° Mercenaria campechiensis (Gmelin) 4 Oweniidae °*Myriochele Malmgren 4 Arthropoda Paraonidae Tasecta *Aricidea Webster 4 Phyllodocidae OVEMGUES °Eteone Savigny 4 Chironomidae Chironominae: Chironomini Paracladopelma undine (Townes) 2, 3 Paratendipes basidens Townes 2 Robackia claviger (Townes) 2 Saetheria tylus (Townes) 2 °Eumida sanguinea (Orsted) 4 *Phyllodoce Savigny 4 Pilargidae *Loandalia Monro 4 *Parandalia Emerson and Fauchild 4 °Sigambra Muller 4 Stictochironomus devinctus (Say) 2 Polynoidae S. caffrarius (Kieffer) group 3 Orthocladiinae: *Halosydna Kinberg 4 °Harmothoe Kinberg 4 Pseudosmittia Goetghebuer 3 *Lepidonotus Leach 4 Ephemeroptera °Phyllohartmania taylori Pettibone 4 Ephemeridae - Burrowing Mayflies Sabellidae Hexagenia bilineata (Say) 3 °Fabrcicia sabella (Ehrenberg) 4 H. limbata (Serville) 1, 2 *Jasmineira Langerhans 4 Cephalochordata - Lancelets Serpulidae Branchiostomidae *Hydroides dianthus (Verrill) 4 *Branchiostoma caribaeum Sundevall 4 Spionidae °Paraprionospio pinnata (Ehlers) 4 Hemichordata *Prionospio Malmgren 4 Enteropneusta - Acorn Worms °Scolelepis squamatus (O. F. Muller) 4 Harrimaniidae °Spiophanes Grube 4 *Saccoglossus kowalevuskii (A. Agassiz) 4 °Streblospio benedicti Webster 4 *Sipuncula - Peanut Worms 5 Spionidae *Polydora Bosc/Boccardia Carazzi complex 4 °Pseudopolydora Czerniavsky 4 *Spirorbidae 5 Syllidae *Syllis Savigny 4 160 FLORIDA SCIENTIST HyYPOBENTHOS Arthropoda Crustacea Decapoda Callianassidae - Mud Shrimps, Ghost Shrimps *Lepidophthalmus louisianensis (Schmitt) 5 EP1zZ00S/PARASITES Acanthocephala Echinorhynchidea - Fish Parasites Polymorphidae 5 Annelida Hirudinea - Leeches Glossophoniidae Desserobdella phalera (Graf) 3 Gloiobdella elongata (Castle) 3 Helobdella fusca (Castle) 3 H. stagnalis (Linnaeus) - Snail Leech 3 H. triserialis (E. Blanchard) 3 Placobdella papillifera (Verrill) 1, 2 P. parasitica (Say) 3 Arthropoda Arachnoidea Acariformes (=Arachnida) - Water Mites Parasitengona Aturidae Albia Thon 3 Arrenuridae Arrenurus Duges 3 Eylaidae Eylais Latreille 3 Hydrodromidae Hydrodroma Koch 3 Hygrobatidae Atractides Koch 2 Hygrobates Koch 3 Krendowskiidae Geayia Thor 3 Krendowskia Piersia 3 Lebertiidae Lebertia cf. porosa Thor 3 Mideopsidae Mideopsis Neuman 1 Momoniidae Momonia Halbert 3 Oxidae Fromtipoda Koenike 3 Pionidae Piona cf. rotunda (Kramer) 3 Sperchonidae Sperchon Kramer 3 [VOL 57 Sperchonopsis echphyma Prasad and Cook 1 Torrenticolidae Torrenticola Piersig 1 Unionicolidae Koenikea Wolcott 3 Neumania Lebert 3 Unionicola Haldemann 3 Crustacea Branchiura - Fish Lice Arguloida Argulus japonicus Thiele 5 Insecta Diptera Chironomidae Orthocladiinae Epoicocladius Zavrel 3 Chironominae: Chironomini Xenochironomus xenolabis (Kieffer) - Sponge Miner 2 No. 4 1994] SWISHER ET. AL.—FLORIDA’S TOMATO PRODUCERS 161 Biological Sciences FLORIDA’S TOMATO PRODUCERS: ARE THEY MOVING TOWARD SUSTAINABILITY? M. E. SwisHER” AND E. P. Bastrpas”) “Home Economics Department, University of Florida, Gainesville, Fl 32611-0310 Agricultural Education and Communications Department, University of Florida, Gainesville, Fl] 32611. ABsTrAcT: Commercial tomato producers were surveyed throughout the state of Florida to determine the degree to which they have adopted production practices which minimize undesirable environmental impacts. The su rvey focused on water, pest, and nutrient management practices. Environmentally sound production practices and changes in production practices that have occurred over the past decade are described. We examine the relationships between scale of production (size of production unit) and adoption of recommended practices and examine the degree to which scale serves as a predictor of adoption. The debate over the degree to which farmers are moving toward more environmentally sound systems of agricultural production is an important one. The debate is particularly important in Florida. Much of our agriculture is large scale and input intensive. Further, many major production areas, particularly for vegetable crops, occupy regions of the state with particularly sensitive natural ecosys- tems. The on-going controversy over natural resource use in the Everglades Agricultural Area provides a good example. Special care must be taken by growers in the EAA to prevent deterioration of the natural Everglades ecosystem. Excessive concentrations of nutrients or pesticides in the water system, for example, are issues of major concern. SIZE, PRODUCTION SYSTEM, AND SUSTAINABILITY—One school of thought regarding sustainable agriculture argues that large scale, input-intensive agricultural produc- tion systems are inherently “unsustainable” (Schumacher, 1973). This viewpoint has become so associated with the “sustainable agriculture movement” that some regard the term sustainable agriculture as virtually synonymous with a call for the return to smaller scale, less input intensive farming (Buttel, 1990; Caporali, 1992; Trenbath et al., 1990). Despite this widely held viewpoint, virtually no research has been conducted which actually compares the use of environmentally protective produc- tion practices on large and small farms. One exception is the work of Hefferman and Green (1986). They estimated soil loss on a randomly drawn sample of farmers in a midwestern US county. Results proved contrary to the popular opinion that “larger farmers are less concerned about the environment and therefore less likely than small scale farmers to employ environmentally sound methods and practices” (1986: 31). Large farms were found to have lower soil loss than smaller farmers. However, Hefferman and Green believe that this result reflects the fact that larger farms occupy sites that are less prone to erosion than smaller farms. While interesting, these results are site-specific and may have little applicability to Florida agriculture where, for the most part, soil loss is not a critical issue and where there is often relatively little variation in bio-physical conditions between the sites occupied by small and larger commercial producers. 162 FLORIDA SCIENTIST [VOL 57 Crops produced in the midwest are less energy-, management-, and input-intensive than many of the major Florida crops, especially horticultural crops. Nowak’s (1987) conceptual model of adoption behavior is more comprehensive than most. Nowak compares economic and diffusion models for explaining the adoption of agricultural conservation technologies in Iowa. He concludes that the two models are complementary, not competing, and that both personal character- istics, such as information source, and economic characteristics, such as land tenure and farm income, contribute to explaining why farmers adopt new technologies. His study also deals with the adoption of conservation tillage techniques rather than a wider spectrum of environmentally protective technologies differing in manage- ment and capital demand. Grieshop and Raj (1992) conducted a survey of California farmers with the intent of measuring the degree to which these farmers are moving toward sustainability. Their study did include vegetable producers and a range of technologies. They showed positive results and were able to divide farmers into three groups, based on the degree to which they have altered their farming practices to meet environmental concerns. However, their sample was not representative of the entire farm popula- tion, but rather consisted of individuals attending conferences dealing with sustain- able agriculture. Size was not included as an independent factor predicting adoption and they did not compare practices among producers of a specific commodity. Nowak’s and Grieshop and Raj’s studies also fail to address the question of the appropriateness of different technologies for different size classes of farms. The implicit assumption in many studies is that the technologies available to farmers to enhance environmental protection are scale neutral; that is, that there are no barriers to adoption of these technologies that are related to the size of the farm operation. Yet, many of the technologies that farmers need to use to protect natural resources are either capital and/or management intensive. Injection of fertilizer can reduce the potential for nutrient pollution of surface and ground water; however, injection systems require high outlays of capital, which small farmers may not have. Use of integrated pest management techniques can reduce total pesticide use, thereby reducing production costs. This reduction could be expected to be particularly important to capital-limited, smaller growers. Rahm and Huffman (1984) addressed the issue of farm size. They included both farm size and cropping system as variables in their model of technology adoption behavior. Other factors include human capital variables, such as educational level and source of information. However, their study was concerned with the adoption of reduced tillage techniques rather than a range of technologies of differing labor, management, and capital demands. Faeth (1993) considered net farm income and net farm operating income in his discussion of the adoption of environmentally protective farming practices in Pennsylvania. Once again, however, the farmers included in the survey included grain producers, a relatively low value crop. Further, as in the case of Rahm and Huffman’s study, Faeth’s work focuses on the adoption of reduced tillage tech- niques. Shapiro and others (1993) considered risk as a factor affecting adoption of new No. 4 1994] SWISHER ET. AL.—FLORIDA’S TOMATO PRODUCERS 163 technologies in their work. However, the risk faced by the farmers who participated in this study conducted in Niger are primarily climatic in nature. Prolonged drought during the growing season is the major source of risk for these growers. For Florida tomato producers, risk is also high, but is associated largely with market and price factors, not natural phenomena. We conducted a survey of tomato producers in Florida. We tested two hypoth- eses. The first was that farm size is a predictor of adoption of recommended practices. The second was that there is a positive relationship between farm size, adoption, and capital and/or management intensiveness of recommended technolo- gies. We predict, contrary to much popular belief, that larger growers will show higher adoption rates of desirable technologies and that part of the difference in adoption rates between small and large farms can be explained by the nature of the technologies that are offered for adoption. We focused on the adoption of Best Management Practices, a term used by the United States Department of Agriculture’s Extension Service to identify manage- ment practices based on research findings to be the most sound from both a bio- physical and economic perspective. These practices include, for example, integrated pest management, adoption of irrigation systems which reduce water use, and plant management programs based on plant nutrient demand. We include seven variables in our discussion here. Three of the variables, frequency of soil testing, use of tissue tests, and change in method of fertilizer application since 1982, are nutrient management practices. Two are water management practices, method of irrigation and records of water quality test results, and two are pest management practices, basing pest management decisions on samples of pests actually present in the field and the presence of mixing and loading facilities. MetTHoDs—We used County Extension producer lists for our sampling frame. We chose County Extension producer lists, rather than lists of such organizations as the Florida Fruit and Vegetable Association, because the latter lists might be biased in favor of larger producers because there are Association fees associated with membership in FFVA. There were a total of 101 tomato growers included on the County producer lists. We selected three independent samples, based on the size of the operation. Size Class | included tomato producers with less than 67 ha of tomatoes. Size Class 2 included producers with 67 to 267 ha, and Size Class 3 included growers with over 267 ha of tomatoes. Our total sample, including all size classes, was 64. The response rate to the survey was 91%, and included 57% of all tomato producers in the sampling frame. Our results included information about 50,591 acres of tomatoes. We conducted personal interviews with the growers during the period May to July, 1993. Our survey instrument was developed with the assistance of a large number of Extension Agents and Extension Specialists in the Institute of Food and Agricultural Sciences. It focused on three major areas of environmental concern in Florida: pest, water, and nutrient management. Characteristics of the sample—Tomato acreage—Mean tomato acreage for the entire sample was 363 ha. The median was 167 ha, as was the mode. For Size Class 1 producers, mean tomato acreage was 32, with a median acreage of 31. Size Class 2 producers showed a mean of 160 ha anda median and mode of 167, the same as for the population as a whole. For the largest producers, Size Class 3, mean tomato acreage was 881 ha, with a median of 646 ha and a mode of 458 ha. The close correspondence of the median and mode for Size Class 2 with that of the population as a whole supports our original size class definitions, indicating that they were well defined. Age of the farm—We asked how long the business had been producing tomatoes at the current geographic location. We found that Size Class 2 farms are, compared to Size Classes 1 and 3, relatively new farms. Two-thirds (66%) began tomato production at the current location since 1980. For Size Class 164 FLORIDA SCIENTIST [VOL 57 1 and 3 farms, the age of the farm is much more uniformly distributed over time, with the earliest farms being established in the 1920's and the newest since 1990. All questions regarding personal character- istics were asked for the farm operator or the individual designated as the primary decision-maker for tomato production. In some cases this individual is the owner, but in other cases is an employee or, commonly, a relative of the actual land owner. Age—Mean operator age for the entire sample was 46 years. There was little variation among size classes, although Size Class 1 producers, with a mean age of 51 years, were slightly older than Size Class 2 and 3 producers, each of which showed a mean age of 44 years. Educational background—These farm operators had relatively high educational levels, compared to the Florida population as a whole (Florida Statistical Abstract, 1993). Of the farm operators, only 9% had less than 12 years of education. Twenty-eight percent had a high school diploma. Over half, 63% had some college (45%) or a college degree (19%). There were no individuals with less than 12 years of education in Size Class 2. The percentage of individuals with some college education or a college degree was nearly the same in all size classes, meaning that Size Class 2 has a higher percentage of individuals with a high school diploma than Size Classes 1 and 3. Trade-association membership—This information included both the farm as a business and/or the farm manager. In most cases, the two were synonymous in terms of trade association membership. There were differences in trade association membership by size class. All Size Class 3 farms belonged to one or more trade associations. However, 40% of Size Class 1 farms did not and a smaller percentage, only 8%, of Size Class 2 farms had no trade association membership. Of all farms, only 13% did not belong to an association. ResuLts—Overall, results indicate that Florida’s tomato producers have adopted many practices which can help protect the environment. Our results show high adoption levels (over 60% of all producers) of soil testing, use of low volume irrigation systems, records of water test results, and scouting for pests, for example. The practices discussed here are only a few of the practices that were studied and initial analysis shows high adoption rates in other cases as well. In each case, we hypothesized that smaller (Size Class 1) producers would be more apt to adopt technologies requiring low management and low capital than they would technologies with high management and capital demands. Conversely, we expected the largest producers (Size Class 3) to be more apt to adopt technologies characterized by high management and capital demand than would either Size Class 2 or Size Class 1 producers. We define capital-intensive technologies as those that require high initial capital investment, but offer no immediate return to investment. Installation of a mixing and loading facility requires a high investment and increases the grower’s protection of natural resources, but it does not provide any immediate monetary benefit to the grower. Soil testing, on the other hand, requires little capital investment and potentially offers immediate monetary benefits if fertilizer applica- tion rates can be reduced. Frequency of soil testing—The Best Management Practice is for the grower to take a soil test every production season. Soil testing is, by our definition, a low-capital, low-management technology. We would predict that this technology would be attractive to all growers, but particularly to Size Class 1 growers because of its potential for reducing production costs. Results show that there were no significant differences in adoption rates of seasonal soil testing among tomato growers. We asked growers how frequently they took soil tests. If we combine growers who tested seasonally with those who tested at least once a year, as opposed to growers who test less than once a year or never, a higher percentage of Size Class 1 growers did adopt this technology, although the differences were not statistically 165 FLORIDA’S TOMATO PRODUCERS SWISHER ET. AL. No. 4 1994] 7 SYAWaVS JO % Size 3 Size 2 SIZE CLASS io is) o = @ i} © ° 0 w 1) _j Once a year Ed Every season used for this and ) il -ta 2 *s exact test ( isher [NB. F 1Ze ing, by class si Frequency of soil test subsequent figures]; P=0.45. ] + Fe Fic ~~ 4 IN) ebay & =? dg 2&0 Ss hs s CH =| co eS Sy es Shas Oo) 6 Ss ers a oe de ee eee SS 4) gS Sp LO eS SS So ee aS ie. 2 aes Sours aa © oO Fe en O- s eioee Seo wy CO -=s CRMs meas, oS) Sr Si) Se a Og Bg wb'o gn fe Ae e se zo 0S bow 8 OF -OlaS = ea 2. Ae Of foe GS Sie) ah ees fel lO Ss GS SO = SESECES E Bn wee RE af aon04 870 - Seete . 2S F459 Sea oe = Sak EBA» O 4) Sg (ey at SS oT ODOT is} (Ob fa oman So aslee Sane OS oe Ot Ao ese Oe a 4 eno Oce DD Ay Oo € oD : nO ar 04 v a0 ie ie es eo Ne _ ee Oe OS aS eee Ss 4 ae @ © OB ASS ro oO a 2 oS tal Ce el @ eS) eS BMSoR oko Ss 10), Og Se + © 8 a aby 1 Se Secs —BRonw Aaa) Oo oNe 2 ces Te Scene cceo le oe Bee hits ee i oo DO RR 7 - One ce eG performs the tissue test him/herself and must be able to interpret the results. Pee eeeepenee SYAWYVS JO % Size 3 Size 2 SIZE CLASS 1 Ize Ss YES # NO OO aSS SIZE ;. 2. Tissue testing, by cl Fic 166 FLORIDA SCIENTIST [VOL 57 100 80 60 % OF FARMERS Size 1 Size 2 Size 3 SIZE CLASS YES & NO Fic. 3. Change in method of application; P=0.04. Statistically, there were no significant differences among growers in the adoption of this technology. However, as Figure 2 shows, there was some tendency for Size Class 2 growers to show a higher adoption rate than other size classes. This may be explained by the educational levels characteristic of Size Class 2 growers, the only group who had no members with fewer than 12 years of formal education. Also, slightly over half of all Size Class 2 growers are relatively new to tomato production, having begun to produce tomatoes since 1980. Change in method of application of fertilizer—This question gauges only changes in method of application of fertilizer since 1982, referring specifically to the installation of fertigation or injection systems, both of which are relatively high-cost technologies. We also questioned growers about method of application of fertilizer in general, but those results are not included here. More growers than those shown in Figure 3 use fertigation and/or injection. However, in this case, we were concerned largely with the capital demand associated with the technology and only secondarily with the management demand. Fertigation and injection could be considered relatively high-management technologies, but the most significant characteristic of these technologies is the high capital investment required for their use. Therefore, we would expect higher adoption rates among larger (Size Class 3) growers. Data supported our hypothesis; Size Class 3 growers showed significantly higher adoption rates than Size Class 2 growers, and Size Class 2 growers showed significantly higher adoption rates than Size Class 1 growers (Fig. 3). Change in method of irrigation—These data refer only to changes in type of irrigation systems. A higher percentage of farmers use low-volume systems than the number who have adopted low-volume irrigation since 1982. Changing irrigation systems requires a significant capital outlay. We therefore considered this a high- capital technology. Management of low-volume irrigation also could be considered 167 SWISHER ET. AL.—FLORIDA’S TOMATO PRODUCERS No. 4 1994] SYAWeVS JO % Size 3 Size 2 Size 1 SIZE CLASS YES tJ NO =0.47. Fic. 4. Change in method of irrigation (to low-volume system) since 1982, by class size; P | | | raed yy Mf MJJE=X=—_,] JJ» | | | bai aM | | I YYI=|q=——MJJjyJ@@v—-”’ WHMMIWJJIJISWWWYWW#¥V. WMMW=WWWJ|(JJl—Jl|''00#«— 50% scouted acres Wis Fic. 6. Acreage scouted for pest-management decisions, by class size; P=0.19. No. 4 1994 | SWISHER ET. AL.—FLORIDA’S TOMATO PRODUCERS 169 % OF FARMERS Size 1 Size 2 Size 3 SIZE CLASS YES tJ NO Fic. 7. Mixing and loading facilities, by class size; P=0.24. in Size Class 1. Our data do not support this hypothesis. No significant differences appeared among groups based on the Fisher’s exact test statistic, although we did find higher adoption rates among Size Class 2 growers than either Size Class 1 or Size Class 3 growers (Fig. 7). This may be due to the fact that Size Class 2 growers have a high percentage of relatively new production units. ConcLusions—The data show that there is no reason to believe that size class alone will serve as a good predictor of a farmer's. adoption of environmentally protective production practices. On the other hand, our general hypothesis that the management, and particularly the characteristics, of the technology would serve as a predictor of adoption by size class was relatively well-supported. Capital require- ment was a better predictor of adoption by size class than was management intensity. This is especially well shown by the relationships between size class and change in method of fertilizer application. The results also show that it is not necessarily true that smaller growers are more protective of the environment than larger growers. These results may reflect peculiarities of tomato production, and may not be applicable to other farming enterprises. Tomato production is much more capital-, management-, and labor- intensive than many types of farming. Therefore, the relative infrequence of differences in technology adoption according to size class may largely reflect the fact that tomato producers as a group use high-management, high-capital production practices. Furthermore, tomato production is a relatively high risk agricultural enterprise, although one that offers equally high earning potential. Production costs are high, and prices are subject to rapid, large fluctuations. This is not an enterprise that farmers with high capital restrictions are apt to use. Further, farmers spread the risk 170 FLORIDA SCIENTIST [VOL 57 associated with tomato production by planting other crops as well. Because of the risk associated with an investment in tomato production, producers must “do it right.” The result is that there is relatively little variation in the production practices employed by tomato-producers, regardless of location or size. We have also con- ducted surveys of sweet corn and potato producers. Both of these enterprises are lower risk than tomato production. Comparison of the results of those surveys with those of the tomato producer survey may permit us to draw more general conclusions in the future. ACKNOWLEDGMENTS—We wish to thank the commercial tomato producers of Florida and thé many County Extension faculty members of the University of Florida who assisted us with this work. We appreciate their valuable time and resources. LITERATURE CITED ButTEL, F.H. 1990. Social relations and the growth of modern agriculture. Pp. 113-145. In: Carrou C.R., J.H. VANDERMEER AND P. Rosset (eds.), Agroecology. McGraw-Hill, New York, NY. Caporaul, F. 1992. Validity of rotation as an effective agroecological principle for a sustainable agriculture. Agric. Ecosys. Environ. 41: 101-113. FLORIDA STATISTICAL ABSTRACT, 1993. University of Flrodia Press, Gainesville, FL, 27th Ed. Pp 696. FaetTu, P. 1993. Evaluating agricultural policy and thesustainability of production systems: an economic framework. J. Soil Water Conserv. 48(2): 94-99. GrieEsHop, J.I. AND A.K. Raj. 1992. Are California’s farmers headed toward sustainable agriculture. Cal. Agric. 46 (2): 4-7. HAFFERMAN, W.D. ANDG.P. GREEN. 1986. Farm size and soil loss: prospects for a sustainable agriculture. Rural Soc. 51 (1): 31-42. Nowak, P.J. 1987. The adoption of agricultural conservation technologies: economic and diffusion explanations. Rural Soc. 52 (2): 208-220. RauM, M.R. AND W.E. HurrMan. 1984. The adoption of reduced tillage: the role of human capital and other variables. Amer. J. Agric. Econ. 66: 405-412. SCHUMACHER, E.F. 1973. Small Is Beautiful: Economics as if People Mattered. Harper and Row, New York, NY. SHapiro, B.I., J.H. SANDERS, K.C. Reppy, AND T.G. Baker. 1993. Evaluating and adapting new technologies in a high-risk agricultural system — Niger. Agric. Sys. 42: 153-171. TRENBATH, B.R., G.R. Conway, AND I.A. Craic. 1990. Threats to sustainability in intensified agricultural systems: analysis and implications for management. Pp. 337-365. In: GLEISsMAN, S.R. (ed.), Agroecology: Researching the Ecological Basis for Sustainable Agriculture. Springer- Verlag, New York, NY. Florida Scient. 57(4):161-170. 1994. Accepted: July 5, 1994. No. 4 1994] PAGE—IDENTIFICATION OF SAILFIN CATFISHES IN FLORIDA 1 what field? IDENTIFICATION OF SAILFIN CATFISHES INTRODUCED TO FLORIDA LAWRENCE M. PAGE Illinois Natural History Survey, Champaign, IL 61820 Abstract: Two species of South American sailfin catfishes are established in Florida. Liposarcus multiradiatus is established in several canals and lakes in Dade and Palm Beach counties. Liposarcus disjunctivus is established in Hillsborough County in Hillsborough River and in Lake Thonotosassa. Both species are common in the tropical fish trade and almost certainly escaped from fish farms or were released by aquarists. SAILFIN catfishes (loricariids with 10 or more dorsal fin rays) established in Florida were tentatively identified by Ludlow and Walsh (1991) as Pterygoplichthys cf. multiradiatus. Recent taxonomic revisions by Weber (1991, 1992) allow more specific identifications of these fishes. Weber (1991, 1992) assigned sailfin catfishes to three genera. All the specimens collected in Florida clearly are assignable to Liposarcus. The specimens lack the elevated supraoccipital process of Glyptoperichthys and differ from Pterygoplichthys in having the supraoccipital bone bordered posteriorly by three scutes rather than by one large scute. Among the four species of Liposarcus recognized by Weber (1992), only L. anisiti has light spots on a dark background, and only L. multiradiatus has a pattern of uncoalesced dark spots on a light background. L. disjunctivus and L. pardalis have a dorsal pattern of coalesced dark spots on a light background. L. disjunctivus differs from L. pardalis in having the dark spots on the venter coalesced and forming a vermiculate pattern; in L. pardalis this pattern consists of discrete spots. Specimens in the fish collection of the Florida Museum of Natural History document the presence of two species of sailfin catfishes reproducing in Florida. Oneis Liposarcus multiradiatus, and the other is Liposarcus disjunctivus. Liposarcus multiradiatus is established in several canals and lakes on the Atlantic Slope in Dade and Palm Beach counties (UF 35606, 42194, 42195, 42196, 92143, and 92147). Liposarcus disjunctivus is known only on the Gulf Slope in Hillsborough County in Hillsborough River (UF 79623) and in Lake Thonotosassa (UF 89896). The six specimens of L. multiradiatus range in size from 231 to 349 mm. The 12 specimens of L. disjunctivus range in size from 244 to 364 mm. Loricariids are native to South and Central America. Liposarcus multiradiatus occurs naturally only in the Orinoco River basin of Venezuela; L. disjunctivus is endemic to the Rio Madeira drainage (Amazon River basin) of Brazil and Bolivia. Both species established in Florida are commonin the tropical fish trade and almost certainly escaped from fish farms or were released by aquarists. The 172 FLORIDA SCIENTIST [VOL 57 accepted common name for Liposarcus multiradiatus is the sailfin catfish (Robins, et al. 1991) although it often is sold in pet stores as the butterfly pleco. Liposarcus disjunctivus does not have acommon name (Robins, et al. 1991); I suggest that, in view of its vermiculated venter, it be called the vermiculated sailfin catfish. Other, similar-looking species of loricariids established in Florida have fewer than 10 dorsal rays and are assignable to the genus Hypostomus. Species of Hypostomus cannot presently be identified to species. ACKNOWLEDGMENTS—George H. Burgess assisted with examination of specimens in the Ichthyology Collection of the Florida Museum of Natural History. Michael E. Retzer, Jonathan W. Armbruster and Paul L. Shafland read and improved the manuscript. LITERATURE CITED Lup.ow, M. E. anp S. J. WatsH. 1991. Occurrence of a South American armored catfish in the Hillsborough River, Florida. Florida Scient. 54:48-50. Rosins, C. R., R. M. BatLey, C. E. Bonp, J. R. BROOKER, E. A. LACHNER, R. N. Lea, AND W. B. SCOTT. 1991. Common and scientific names of fishes from the United States and Canada. American Fisheries Society Special Publication 20. WEBER, C. 1991. Nouveaux taxa dans Pterygoplichthys sensu lato (Pisces, Siluriformes, Loricari- idae). Revue suisse Zool. 98:637-643. 1992. Revision du genre Pterygoplichthys sensu lato (Pisces, Siluriformes, Loricariidae). Revue fr. Aquariol. 19 (1&2):1-36. Florida Scient. 57(4):171-172. 1994. Accepted: July 5, 1994. No. 4 1994] PASCARELLA—ADDITIONS TO SOUTH FLORIDA FLORA We Biological Sciences ADDITIONS TO THE FLORA OF SOUTH FLORIDA: FOUR NEW SPECIES OF NATURALIZED TROPICAL TREES JOHN B. PASCARELLA University of Miami, Department of Biology, Coral Gables, FL 33124-0421 Apstract: Four additional species of introduced tropical trees have naturalized in Dade County, Florida. The species are Alstonia macrophylla (Apocynaceae), Pittosporum pentandrum (Pittosporaceae), Ixora arborea (Rubiaceae), and Harpullia arborea (Sapindaceae). Brief descriptions, herbaria records, ornamental use, and potential spread are discussed. FIELD investigations of the changing flora of a subtropical forest in south Florida (Matheson Hammock Park, Coral Gables, FL) and examination of regional herbaria (FTG, USF, FAU, and FLAS) have revealed four new species of naturalized exotic trees that have achieved sustained naturalization and spread in Dade County, Florida since the publication of the Flora of Tropical Florida (Long and Lakela, 1971). Exotic species are alien organisms either purposefully or accidentally brought from their native ranges into Florida where they have since escaped into the wild and reproduce either asexually or sexually (Anonymous, 1993a). For this paper, natural- ization is defined as a wild population having reproductive adults, juveniles, and seedlings in either disturbed or undisturbed habitats. To facilitate identification of these tree species if they continue to spread and invade other natural areas in south Florida, this paper provides a brief morphological description, examines reproduc- tive ecology, and discusses the potential for continued spread. 1. Apocynaceae-Alstonia macrophylla Wall. (Ex G. Don) Description: Tree (3-8 m), leaves whorled, petioles 1-3 cm, leaf blade 10-50 cm long, cuneate base, short-acuminate apex, glabrous above but densely pubescent below when young; lax umbellate cymes, flowers white, fragrant, nocturnal dehis- cence, corolla glabrous except for ciliate lobes, tube and lobes .5 cm long. Pendulous merocarps 30-45 cm long. Native of Malaysia (Dassayanake and Fosberg, 1983). First noted in the late 1950s, Alstonia macrophylla is now very abundant in open areas west of Matheson Hammock near Snapper Creek Canal. Other collections have been made in hammocks, pinelands, and disturbed areas. Morton (1979) erroneously reported that Alstonia scholaris R. Br. was also naturalized in Dade County yet all herbaria specimens examined fit the description of A. macrophylla in Dassayanake and Fosberg (1983). The white fragrant flowers suggest butterfly or moth pollination but the mating system and actual pollinators in south Florida have not yet been studied. This species is not listed in the horticultural trade (Anon., 1990) and has likely 174 FLORIDA SCIENTIST [VOL 57 spread through the wind dispersed seeds from specimens at Fairchild Tropical Gardens and the USDA Plant Introduction Center at Chapman Field. Dassayanake and Fosberg (1983) reported that this species rapidly became naturalized in Ceylon in moist forest regions up to 1200 m after its introduction as a timber tree. It is now a prominent member of secondary forests communities on that island. Invasion routes in south Florida may be limited to open disturbed areas for successful colonization. Specimens examined—Dade Co. Coral Gables, Matheson Hammock Park, near Snapper Creek Canal. 1993. Pascarella 110 (FTG). Dade Co., Miami. In remnants of Hattie Bauer Hammock at Orchid Jungle near Homestead, 7 Aug 1982. Stern s.n. (FLAS). Dade Co. Miami, Escape from cultivation, growing in pine woods on USDA Plant Introduction Station. 1970. Gillis 9659 (FTG). Dade Co., Coral Gables, Volunteer tree 15 m tall with a cylindrical bole, Fairchild Tropical Garden. 1970. Gillis 9038 (FTG). Dade Co., Miami. Brickell Hammock, vicinity of US #1 and Rickenbacker Causeway, fruiting. 12 April 1969. Long and Andorfer 2821 (USF). Dade Co., Miami. Tree to 35 ft. on newly cleared lot, S.R. 25 and Brickell Avenue. 29 Jan. 1961, fruiting. Will s.n. (FLAS). Dade Co., Miami. Escape from U.S.D.A., Hammock along Old Cutler Rd. 12 Dec. 1959. Atwater M-162 (FLAS). 2. Pittosporaceae-Pittosporum pentandrum (Blanco) Merr. Description: Small tree up to 6 m, leaves alternate appearing whorled, simple, wavy-marginate, lacking stipules; flowers in cymose panicles, perfect, regular, sepals 5, imbricate, petals 5, stamens 5, alternate with the petals, gynoecium of unilocular ovary, simple style, lobed stigma; fruit a dehiscent orange berry with several seeds immersed in viscid red pulp. Native of the Philippines. Recently added to the Exotic Pest Plant Council’s List of Florida’s Most Invasive Species (Anon., 1993a), this attractive small tree is beginning to spread in Dade County, FL. No native members of the Pittosporaceae exist in the local flora although a number of other species have been introduced for ornamental use (Morton, 1981). The white flowers are strongly fragrant and are visited by honeybees at the University of Miami Gifford Arboretum (UMGA)(Pascarella, pers. obs.) Both self-pollinated flowers and insect-excluded flowers did not set seed compared to successful seed set in unmanipulated open flowers visited by honeybees indicating that insect visitation and outcrossed pollen appears necessary for seed production (Pascarella, unpub. data). Seed production is high and fruit dispersal is presumably by birds, based on fruit morphology and color. This species is not currently grown by local nurseries (Anon., 1990). The population at Matheson Hammock is restricted to the lower elevation open ground east of Snapper Creek Canal. A second population has also been established in a small pineland remnant across from UMGA. In Australia, a related species, Pittosporum undulatum, has invaded remnant Eucalypt forest outside of its native range, resulting in changes occurring in species composition, fire frequency, and Eucalypt regeneration patterns (Gleadow and Ashton, 1981). This species may pose a threat to south Florida Slash Pine rockland forest preserves. No. 4 1994] PASCARELLA—ADDITIONS TO SOUTH FLORIDA FLORA 175 Specimens examined—Dade Co. Coral Gables, Small tree, 3.5 m, Matheson Hammock Park, east of Snapper Creek Canal. 1993. Pascarella 111 (FTG). Dade Co., Miami. Disturbed pinelands on oolitic limestone on grounds of U.S.D.A. Subtropical Research Station, Chapman Field. Common shrub or small tree, escaped from cultivation. 24 July 1986. Judd 5241 (FLAS). 3. Rubiaceae-Ixora arborea Roxb. ex sn. (I. parviflora Vahl). Description: Tree to 3 m, spreading crown, leaves opposite, dark green above; inflorescence cymose, 12 cm long; flowers four merous, calyx 1 mm, green, pubes- cent, corolla white, strongly fragrant, tube 8 mm, lobes 3 mm long and reflexed, anthers 1.5 mm long reflexing; bilobed green stigma exerted 2 mm above corolla lobes; fruits green turning black. Native of India. Although two Ixora species have been reported as persisting from cultivation (I. coccinea L. and I. grandiflora (Blume) Zoll. & Morr. in Long and Lakela, 1971), only I. arborea has apparently naturalized. I. arborea may be distinguished from other Rubiaceous shrubs by the very strong fragrance of the flowers and the relatively large size of adult individuals. Flowers of a specimen at UMGA are visited by various butterflies, diptera, and wasps and honey bees (Pascarella, pers. obs.) although the mating system is unknown. Fruits are likely dispersed by birds and frugivorous mammals. This species is sold by flowering tree nurseries (Anon., 1993b) for its strong jasmine fragrance. Escaped populations of this small tree have only been noted in Matheson Hammock, a small remnant pineland across from UMGA, and Simpson Park. Based on the density of individuals found in the wild, this species does not appear to be highly invasive. Specimen examined— Dade Co., Coral Gables. Small tree 2.5 m high, Matheson Hammock. 1993. Pascarella 104 (FTG). 4. Sapindaceae-Harpullia arborea (Blanco) Radlk. Description: Tree, 3-15 m, leaves alternate, compound with 9-12 oblong leaflets, acute, entire, glossy, glabrous, 10-15 cm long. Flowers small, green-yellow, anthers brown. Fruit red-orange, dehiscent, 2-3 lobed, hollow except for the non- arillate black seeds. Native of the Philippines (Morton, 1981). This species is common in Matheson Hammock with large reproductives and small juveniles in the upland hammock. One nursery was selling this plant (as Harpullia sp., Anon., 1990) but many large individuals are planted along roadsides in Dade County (Pascarella, pers. obs.). Reproductive biology has not been studied but fruit dispersal is likely by birds or frugivorous mammals. Invasion potential is high based on the high density of individuals found in Matheson Hammock and the apparent ability to recruit juveniles in undisturbed forest. Specimens examined—Dade Co., Coral Gables, Matheson Hammock. 1993. Pascarella 105 (FTG). Dade Co., Coral Gables, Volunteer seedling of tree, Fairchild Tropical Garden. Year ?. Fantz 3259 (FTG). 176 FLORIDA SCIENTIST [VOL 57 ACKNOWLEDGMENTS—I wish to thank Edwin Bridges from Fairchild Tropical Garden (FTG) for assistance in procuring specimen loans from the University of South Florida (USF) and the University of Florida (FLAS). Dan Austin (1994) did not find any specimens of these species in the Florida Atlantic University (FAU) herbarium. This material is based upon work supported under a National Science Foundation Graduate Research Fellowship. This is contribution number 438 from the University of Miami Program in Tropical Biology, Ecology, and Behavior. LITERATURE CITED Anonymous. 1990. Plantfinder: Wholesale Guide to Foliage and Ornamental Plants. Betrock Infor- mation Systems, Inc. Cooper City, F'L. . 1993a. Exotic Pest Plant Council’s 1993 List of Florida’s Most Invasive Species. 3205 College Ave., Ft. Lauderdale, FL. 33314. . 1993b. Tropical Flowering Tree Society: September 1993 Fifth Annual Show and Sale Issue. Miami, FL. AustIN, D. F. 1994. Personal communication. Professor of Biology, Florida Atlantic University, Boca Raton, FL. Dassanayak&, M.D. AND F.R. FosBerc (eds.). 1983. A Revised Handbook to the Flora of Ceylon. Volume IV. Pp. 1-488. Amerind Pub. Co., New Delhi, India. GLEADOW, R.M. AnD D.H. AsuTon. 1981. Invasion by Pittosporum undulatum of the forests of Central Victoria. I. Invasion patterns and plant morphology. Austr. J. Bot. 29:705-720. Lone, R.W. anp O. LakELa. 1971. A Flora of Tropical Florida. University of Miami Press, Coral Gables, FL. Morton, J.F. 1979. Pestiferous spread of many ornamental and fruit species in south Florida. Proc. Fla. Hort. Soc. 89:348-353. . 1981. 500 Plants of South Florida. Fairchild Tropical Garden. Miami, FL. Florida Scient. 57(4):173-176. 1994. Accepted: July 18, 1994. No. 4 1994] BENSON ET.AL.—SELENIUM LEVELS IN REGIONS OF TAMPA BAY ‘Wi Environmental Chemistry SELENIUM LEVELS IN THE PALM RIVER AND MACKAY BAY REGION OF TAMPA BAY Rosert F. BENson!, JOSEPH A. SARACENO, DaviD R. LOWELL AND Davip W. BretTr Institute for Environmental Studies, Department of Chemistry, University of South Florida, Tampa, Florida, 33620 ABsTRACT: Selenium is both an essential trace element in the diet, and it is also toxic at low levels (5 ppm). Anthropogenic activity, such as burning coal for electric power generation, or manufacture of semiconductor electronic parts release selenium into the surrounding environment. This study correlates trends in selenium concentrations with some of the industrial activity near the waters of McKay Bay and Palm River, Hillsborough County, Florida. A previous study carried out in 1976 provides a comparison. Similar trends were found in both studies. An area near the mouth of Palm river and the north end of McKay Bay was found to have the highest levels of selenium in both studies. SELENIUM is found in fresh and salt waters due to environmental cycling. In an aqueous media the equilibrium oxidation states for selenium are highly pH depen- dent. The dominant species at pH 6 is Se(IV) as H,SeO,, and at a pH of 6-9, the prevalent species is Se(VI) as SeO,”. Unpolluted open salt waters have selenium levels of <<1 ppb (Robberecht and Van Greiken, 1982). Although seasonal variations in climate cause fluctuations in selenium levels in water and soil (Haygarth et al., 1993) fresh water values show considerable variations due to the degree of anthro- pogenic activity in the area. Activities such as burning coal, the burning of large areas of forest and metal refining are major sources of selenium in the atmosphere, which may result in higher levels of selenium in nearby water (Cutter, 1989). Other industrial uses of selenium that can add to environmental selenium levels include the manufacturing of electronics, pharmaceuticals, glass, ceramics, and pigments, as well as releases from agriculture and metallurgy. As the toxicological and physiological significance of selenium as a trace element become better defined, the distribution and concentration of environmental sele- nium have become recognized as important to human and animal health. Selenium has been identified as a necessary trace element in a normal diet for prevention of disease, as well as a highly toxic substance when consumed in an excess amount. Grains, meat, fish, and dairy products are recognized as excellent sources of dietary selenium. Selenium deficiencies in diets have been identified as a factor in cardiomyopathy—heart death in people (Yang et al., 1982), white muscle disease in animals (Muth et al., 1959), and as a factor in digestive system cancer (Jackson et al.,1973). The beneficial role of selenium as a deterrent to heart death and digestive system cancer has been suggested as due to the action of an enzyme, selenoglutathioneperoxidase, to destroy oxygen radicals. As a trace element in the ' Author to whom correspondence should be addressed. 178 FLORIDA SCIENTIST [VOL 57 diet, selenium is a factor in maintaining elasticity in body tissues thereby slowing the aging process, in production of prostaglandins, and is a factor in oxygen supply to the heart (Clayman, 1993). Distribution of selenium in diets has been correlated with land factors (Jackson et al., 1986). Selenium can also be quite toxic even in small amounts: 5 to 10 ppm is considered toxic. Chronic selenium toxicity has been observed at 1.89 ppm (Oldfield, 1991). Selenium replaces sulfur in some of the metabolic processes and inhibits the action of some enzymes (Kirschman, 1978). Numerous plants accumulate selenium and transfer it at excessive levels into the food chain of animals and humans (Mertz, 1981). Adverse effects caused by mildly excessive selenium consumption can be characterized by numerous physical impairments such as baldness, loss of nails and teeth, fatigue and with overly excessive levels eventually leading to death (Clayman, 1993). As a result of the narrow range of selenium levels necessary to maintain satisfactory health in humans and animals, some activity toward monitoring the background levels of selenium in the environment could become necessary where anthropogenic activity influences selenium emissions or discharge into the environ- ment. The McKay Bay area in Tampa Bay is an example of an area in need for selenium monitoring. Considerable levels of industrial pollutants have been found in the sediments from the northern area of Tampa Bay (Long, 1993) and the opportunity for selenium emissions or discharges is significant. A large portion of the present day shoreline has been formed by dredging and filling in former mangrove wetlands. Selenium emissions or discharges have been associated with coal fly ash and flue gas desulfurization (Niss et al., 1993; Sayre, 1980). Several coal- fired electrical power generating plants are located near the shoreline in addition to railroad and ship yards. Nelson (1976) studied the selenium concentration around Tampa Bay. At that time the McKay Bay area was found to have the highest selenium concentration in the region. The development of analytical methods for the determination of trace ele- ments, such as selenium, is becoming increasingly important in light of the recent focus on trace element environmental pollution. Niss and coworkers (1993) and Karlson and Frankenberger (1986) have developed methods around ion chroma- tography that can determine selenium at low concentrations suitable for environ- mental monitoring. Several methods for analysis of selenium were considered for this study. Uria and coworkers (1990) compared selenium analyses from stripping voltammetry and spectrochemical methods in order to determine the optimum conditions and capabilities. Electrochemical methods of analysis such as stripping voltammetry offer some of the lowest detection limits available (0.4 ng/ml) with minimum sample pretreatment (Uriaetal., 1990). This method has often been used employing mercury electrodes, but innovative research into use of gold and other metal film electrodes shows great promise in further lowering the detection limits for selenium determinations (Andrews and Johnson, 1975). Posey and Andrews (1981) used anodic stripping voltammetry at an in situ gold-plated rotating glassy carbon electrode (RAuGCDE) to detect selenium at levels as low as 0.15 ppb when gold is codeposited with selenium. The anodic current at a stripping potential of No. 4 1994] BENSON ET.AL.—SELENIUM LEVELS IN REGIONS OF TAMPA BAY 179 1.1vshows a dramatic increase in response when 0.000001M Au(III) is present in the selenium samples during the 4 minute preconcentration at a potential of -0.4v. The purpose of this work was to monitor the selenium concentrations in the Palm River and McKay Bay areas of Tampa Bay, to compare these results with the Nelson study and to further correlate trends in selenium concentrations with local anthropogenic activity. MetHops—The technique for analysis of selenium by anodic stripping voltammetry using an in situ gold-plated glassy carbon electrode (Andrews and Johnson, 1975; Posey and Andrews, 1981) was adapted to carry out this study. A CV-27 Voltammograph equipped with an Analog X-Y Recorder was employed in the experiment in conjunction with a C-18 Cell Stand with stirring and degassing capabilities. The reference electrode employed was a Ag/AgCl electrode, the auxiliary electrode was a platinum wire electrode and the working electrode was a glassy carbon electrode as supplied by the manufacturer of the potentiostat. Polishing of the glassy carbon electrode was accomplished using the alumina and polishing felt supplied bythe manufacturer andas per manufacturer's instructions. The carbon electrode was polished between each analysis of the standards used in the calibration and the environmental samples being tested. Commercial grade nitrogen was used for solution degassing The solutions in the experiment were all prepared as follows using deionized water. Glassware was cleaned with concentrated nitric acid prior to the experiment. A stock solution of 1.777 x 10° M in Au(II1) was prepared by dissolving an accurately weighed amount of Puretronic grade gold metal foil (99.9999%) in a minimum amount of aqua regia and then diluting to volume with 0.1M nitric acid. One milliliter of the gold stock solution was pipetted into each 10-ml aliquot of sample being analyzed. A standard stock solution of 4.82 x 10% M in Se(IV) was prepared by dissolving an accurately weighted portion of SeO, (99.5%) and diluting to volume with 0.1 M nitric acid. Dilutions were made from this stock to prepare the standard solutions. The standard stock solution was stored in a polyethylene bottle. The water site samples as collected were acidified to pH~1 in the field with concentrated nitric acid and stored in polyethylene bottles for the analysis. The selenium analyses of the site samples were carried out as follows. Ten milliliter aliquots of sample were pipetted into a 15 mL cell and 1 mL of the Au (IIJ) solution added to result in a 1.777 x 10° °M gold solution. After polishing the glassy carbon electrode and thoroughly rinsing with deionized water, the working, reference and auxiliary electrodes were placed in the cell stand apparatus and immersed in the cell. For each analysis nitrogen was used to purge the solution for 5 min. and then diverted to blanket the solution during the experiment. A potential of -0.4v was applied to the solution for 4 min. for deposition of the selenium, and then the cell was allowed to rest for a period of 1 min. The potential was then set to cycle between -0.3v and +1.5v at a rate of 0.8v/s in order to obtain voltammagrams for the samples. Blank voltammagrams were also obtained by this procedure using the nitric acid matrix alone and with gold added. The data for the blanks provided the baseline corrections in the standards and the site samples. The reversibility of the reaction Se © Se(IV) +4e (1) was confirmed by comparing the slope of plots of the square roots of potential scanning rates versus the anodic and cathodic peak areas. Equal current data for the anodic and cathodic scan directions, respectively, confirmed reversibility of the electrode process. Measurements were made from the voltammetry traces to yield values for the stripping potential, anodic current and quantity of charge used in the analysis of the data. The regression equation used to determine the selenium concentration of the site samples was developed from the calibration curve of anodic stripping charge versus standard selenium concentration. RESULTS AND Discussion—The sites chosen for sampling and analysis of sele- nium were based upon their proximity to several important land features. These sites are identified on the map (Fig. 1) and features of these sites are summarized (Table 1). Selenium concentrations are included for each site. Several site locations and selenium values from the Nelson study are also included. While the two methods of analysis are quite different, the trends in selenium values should be comparable. 180 FLORIDA SCIENTIST [VOL 57 22nd St 60 Adamo Dr. rosstown Expy co . > Pak , ‘“-_. ettieel - ¥ Pie hedie ou Lo Causewa 7 Blyd. “aT. Port Sutton Fe Wires eae gee re Fic. 1. Map showing the location of sampling sites. Sites from Nelson study are designated with an N prefix. Stripping charge was calculated from the area under the anodic peak current- applied voltage data at an applied potential of 1.05 volts. Overlapping interference anodic peak currents were sometimes found at applied potentials of 0.95, 1.10, and 1.16 volts. Charge from the anodic current at 1.05 volts was determined from among these peaks by using curve fitting methods to isolate the peak at 1.05 volts. An example of the peak current-applied voltage data for sample seven and the blank as a baseline is shown (Fig. 3). At concentrations of Se(IV) above 133 ppb, several layers of selenium are deposited on the working electrode. For concentrations of selenium less than 133 ppb the selenium is codeposited as a monolayer on the electrode in conjunction with gold. This monolayer produces a peak in the anodic current- applied voltage data at 1.05 volts, which corresponds to the standard potential for the Se(IV)/Se couple versus Ag/AgCl electrode with a contribution due to selenium adsorption on gold. Several other potentials of high anodic current flow are sometimes observed. Sometimes an overlapping anodic current peak is observed at a potential of 1.10v, No. 4 1994] BENSON ET.AL.—SELENIUM LEVELS IN REGIONS OF TAMPA BAY 181 TABLE 1. Site characteristics and measured selenium concentrations. Sample Charge, Q, Selenium site pH Location Description millicoulombs conc., ppb 1 6,5 near solid waste plant and 0.052 35 crosstown exit #10 2 7.5 mouth of estuary containing 0.165 (191)** solid waste plant 3 7.0 at crosstown bridge 0,046 2 4 5 south bend in Palm River 0.118 125 5 1.5 north bend in Palm River 0.196 (233)22 6 5 adjacent to radio towers and 0.240 (294)** 56th Street bridge i Uo mouth of Palm River 0.148 167 8 7.8 under power lines crossing 0.082 (1:1 dilution) 132) McKay Bay 10 7.8 Port Sutton power plant 0.068 74 0.048 (1:1 dilution) 0.020 (1:3 dilution) 0.007 (1:10 dilution) N-0O* = Fresh water region of Palm River — ly, — Off map N= I = Brackish water region of a ~250 Palm River N-2* == Salt water region of Palm River = 23 =o: = McKay Bay — salt water —= <0.2 N-4* == McKay Bay — salt water a2 <0.2 *Indicates sample from Nelson dissertation cited in text. *°Values in parenthesis are extrapolated beyond calibration. probably due to adsorbed impurities at the electrode surface. At high selenium concentrations, an additional peak in the current at 0.63v is observed. A peak shows up at 0.95v in the dilutions. These additional peaks are indicative of either several species active at applied potentials near 1.05 volts or multiple states of selenium interaction at the electrode surface such as the bulk oxidation of multilayer selenium. This type of selenium comes off the electrode at slightly lower potentials than the monolayer selenium. Multiple types of selenium oxidations at the electrode distort the shape and reproducibility of the gold/selenium monolayer peak. These multiple peaks were the source of the problem in developing and extending the standard curve to more concentrated samples. In order to determine the concentration of Se(IV) in the collected water samples, a standard calibration curve of cumulative charge from the Anodic potential peak vs. Se(IV) concentration was created. A calibration curve was developed from standard solutions of differing Se(IV) concentrations over the range of 1 — 200 ppb and is shown (Fig. 2). A linear range of 1 — 133 ppb correlates well with the linear range observed by Posey and Andrews (1981). The data between 1 — 133 ppb selenium (IV) concentration were fitted to the following linear equation [Se(IV)] = 1373.4 Q - 35.9 (2) 182 FLORIDA SCIENTIST [VOL 57 0.15 0.1 0.05 Q (millicoulombs) Boon Oe bool) =o ay 200 Selenium Concentration (ppb) Fic. 2. Calibration curve for stripping charge versus selenium concentration. (0) represents calculated points using regression equation while (©) represents experimental data. where Q represents the charge in millicoulombs required to oxidize the selenium on the electrode. A correlation factor of 0.96 was obtained for this equation and a standard deviation of the regression line was found to be +12.4 ppb. The resultant selenium concentration is given in parts per billion. Error analysis of the regression data shows a relative error of 18% based upon the median value over the range of the regression. This large value has No. 4 1994 | BENSON ET.AL.—SELENIUM LEVELS IN REGIONS OF TAMPA BAY 183 14 12 /em) i) microamperes I ( 6 A yo I (microamp) tot. Blank ) on =— te | 1.4 = r= 0.8 0.9 E (applied voltage/cm) Fic. 3. Example of anodic stripping current as a function of applied voltage for the selenium sample (2) and for a blank (9). contributions due to handling problems for trace analysis quantities and by the interference of overlapping peaks in the determination of the charge used to strip the selenium. Some of these results may be at best semi-quantitative since several of the samples were found to have levels of selenium above the linear range of calibration even after dilution. Since the sensitivity of the response of the charge, Q, against selenium concentration above 130 ppb becomes small, the actual values are much higher and warrant further study. The selenium levels found in these samples range from typical to unexpect- edly high values for the waters evaluated. The northern area of McKay Bay has some of the highest levels of inorganic pollutants in the Tampa Bay area (Long, 184 FLORIDA SCIENTIST [VOL 57 1993). Therefore, the high selenium levels found in this study and the Nelson study are consistent with the levels of other inorganic pollutants in the McKay Bay area. Even higher levels of selenium were found in the lower stretches of the Palm River near the mouth and extending to the north bend (sites 2,5,6, and 7 in this study; N2, and N1 from the Nelson study). Samples 2, 5, and 6 were outside of the linear calibration range even after dilutions but limited resources prevented further efforts. At sites up river (3 and NO) selenium levels were relatively low but still above levels in open water areas. Even though Sites 1 and 2 were near a solid waste landfill, the selenium level was relatively low in the interior part of the estuary and high near the estuary opening at the mouth of the Palm River. In this study, site 8 which is located in the channel for the tidal exchange flow into the McKay Bay and a ship yard contained a high level of selenium at 132 ppb relative to the open water site samples (N3 and N4 in the Nelson study did not have detectable levels of selenium) The difference in the selenium values for sites 8, N3, and N4 are quite significant either in terms of the change due to site location or due to possible degradation of the open water area of McKay Bay. Site 10 was located near a coal fired power plant where all of the McKay Bay tidal exchange flowed past and was found to have selenium values lower than site 8. Sample 10 was also evaluated at three dilutions to yield the same concentration within experimental error. Coal fly ash has been reported to have appreciable levels of selenium (Essington et al., 1987; Niss et al., 1993). Selenium levels in the Palm River sites 2,5, 6, and 7 tend to correlate with the values Nelson (1976) obtained for site N1 . In particular, there was an increase in the selenium concentration in the waters just past the landfill corresponding to sites 2 and N2. MacKay Bay (site 8) was found to have a level of 132 ppb as compared with <0.2 ppb reported by Nelson at sites N3 and N4. A sample of open seawater from the Pacific Ocean was analyzed for selenium content by the methods used in this work for comparison with these results. Selenium was not detected and no anodic peak current was observed at the applied potential of 1.05 volts. The McKay Bay region of Tampa Bay represents an extreme case of environ- mental modification by anthropogenic activities. A large fraction of the shore line has been claimed from mangroves by dredging the bay area and filling the wetlands. Overall, the apparent increase in selenium levels found between the Nelson study in 1976 and the present could be due to a continuing redistribution of the selenium sediments disturbed by the dredging activity and/or due to the land use around the bay. There are several coal-fired electrical power generation plants, a railroad yard and a ship yard located around McKay Bay which could also contribute to selenium distribution. Further study of this region and a larger sample population maybe needed in order to determine the magnitude of this problem. Some of the selenium values in McKay Bay are comparable to a 1978 study which rated selenium levels at 203 ppb in the San Luis Drain Canal in California as among the highest levels in the U. S. (Cutter, 1989). No. 4 1994] BENSON ET.AL.—SELENIUM LEVELS IN REGIONS OF TAMPA BAY 185 LITERATURE CITED AnpreEws, R. W. AND D. C. Jounson. 1975. Voltammetric deposition and stripping of selenium(IV) at a rotating gold disk electrode in 0.1M perchloric acid. Anal. Chem. 47: 294-299. CiayMaN, C. B. (ed.) 1993. The American Medical Association Home Medical Encyclopedia. Random House, New York, 890. Cutter, G. A. 1989. Freshwater Systems. Pp. 248 In: Occurence and Distribution of Selenium. M. Ihnat, (ed.) CRC Press. Boca Raton, FL. EssincTon, M. E., L. K. SPACKMAN, J. D. HarBour, AND K. D. Hartman. 1987. Physical and chemical characteristics of retorted and combusted western reference oil shale. DOE Report DOE/FE/ 60177-23433. Laramie, WY. Haycartu, P. M., A. F. Harrison, AND K. C. Jones. 1993. Geographical and seasonal variation in deposition of selenium to vegetation. Environ. Sci. Technol. 27, 2878-2884. Jackson, M. L., D. A. GILLETTE, E. F. DANIELSON, I. H. BLirrorp, R. A. BRYSON, AND J. K. Syers. 1973. Global dustfall during the Quaternary as related to environments. Soil Sci. 116:135-145. ,C. S. Li, AND D. F. Martin. 1986. Doland characteristics affect heart and gastrointestinal cancer death rates among Florida counties? Florida. Scient. 49:82-97. Kar son, U. AND W. T. FRANKENBERGER, JR. 1986. Single-column ion chromatography of selenite in soil extracts. Anal. Chem. 1986, 58, 2704-2708. KirscumaN, J. (dir.) 1978. Nutritional almanac, rev. ed. Nutritional Search, Minneapolis, Minn. Pp. 80- 81. Lone, E. R. 1993. Sediment toxicity in Tampa Bay: spatial extent and magnitude. 57th Meeting Florida Academy of Sciences, St. Petersburg, FL. Mertz, W. 1981. The essential trace elements. Science 213:1332-1338. Muth, O. H., J. E. OLDFIELD, J. R. SCHUBERT, AND L. F. REMMERT. 1959. White muscle diesease (myopathy) in lambs and calves. VI. Effects of selenium and vitamin E on lambs. Am. J. Vet. Res. 2231-234. NELsoNn, G. R. 1976. Analytical method for aqueous selenite ion in the environment. Masters Thesis. Univ. South Florida, Tampa, FL. Niss, N. D., J. F. ScHaBRON, AND T. H. Brown. 1993. Determination of selenium species in coal fly ash extracts. Environ. Sci. Technol. 27:827-829. OLDFIELD, J. E. 1991. Some safety considerations involving selenium. Bull. Selenium-Tellurium Assoc. Inc. Grimbergen, Belgium. Posey, R.S. AND R. W. ANDREWS, 1981. Determination of selenium (IV) by anodic stripping voltammetry with an in situ gold-plated rotating glassy carbon electrode. Anal. Chim. Acta 124:107-112. ROBBERECHT, H. AND R. VAN GREIKEN. 1982. Selenium in environmental waters: determination, speciation, and concentration levels. Talanta 29:823-844. SayrE, W. G., 1980: Selenium: A water pollutant from flue gas desulfurization. Am. Pollut. Contr. Ass. J., 30, (10). Uria, O., J. M. EsTe.a, V. Cerpa, J. L. BERNAL, M. J. Nozat, L. DEBAN, AND F. J. Gomez. 1990. A comparative study of a number of methods for sensitive selenium determinations in waters and fodder correctors. J. Environ. Sci. Health, A25(4):391-404. YANG, G. O., G. Y. Wanc, T. A. YIN, S. Z. Sun, R. H. ZHou, R. Y. Man, F. Y. Zual, S. H. Gou, H. Z. Wanc, AND D. Q. You. 1982. Relationship between the distribution of Keshan diesease and selenium status. Acta Nutrimenta Sinica 4:191-199 (in Chinese). Florida Scient. 57(4):177-185. 1994. Accepted: July 22, 1994. 186 FLORIDA SCIENTIST [VOL 57 REVIEW Garfield, Eugene. Of Nobel Class, Women in Science, Citation Classics ® and Other Essays, ISI Press, Philadelphia, xviii + 589 pp. $35.00 (free shipping on pre- paid orders). Cloth only THIS is volume 15 in Essays of an Information Scientist and covers 1992-1993. Like previous volumes, this is a compilation of well-written, thoroughly documented essays on a range of topics. One group deals with some specific examples of citation analysis, which was pioneered by the author. “The role of undergraduate colleges in research” was a particularly revealing analysis of 74 liberal arts institutions, including 50 “science active” colleges and the significant contribution that they make to United States research. Use of citation analysis to predict future Nobel laureates is a use that astute administrators would be wise to consider. (Future nobelists are cheaper to relocate than existing ones.) A second group are person-essays that deal with backgrounds and contributions of prize winners (the Nobel Prize, the Bernal Prize, the Priestly Medal, the NAS Award in Scientific Reviewing), Current Contents® contributors, and other interesting persons. My favorite essays are in the third group that demonstrate the remarkable range of interests of the author. Favorites here include an essay on why the poet Anne Sexton died, Huichol art and culture, and an essay on AIDS research. Some essays are joint contributions with background provided by the author and followed by a reprinted essay from another contributor. All are interesting. This is a companion book that should be available when one is travelling or when one wishes to be stimulated by interesting topics, interestingly written. — Dean F. Martin, University of South Florida No. 4 1994] 187 REVIEW Everglades Agricultural Area (EAA): Water, Soil, Crop, and Environmental Management. A.B. Bottcher and F.T. Izuno, eds. University Presses of Florida, 1994. xvi + 319 pp. $49.95 (cloth) THIS bookis the result of a five-year project funded by the South Florida Water Management District, the objective of which was to identify and evaluate potential on-site water quality practices for controlling losses of nutrients to environmentally sensitive areas. Known information about the EAA was widely scattered and needed to be summarized in a usable format for water managers, scientists, environmental- ists, and growers to develop appropriate policies and practices. The review also pointed out gaps in the available information. Chapters include: an introduction and overview of the area and the problems, and activities toward solutions, and a description of the history of water management in South Florida, both written by the editors; several chapters describing the soils of the EAA, nitrogen and phosphorus in the soils, water management and water quality research, monitoring and abate- ment programs and the economic importance of the EAA; and production of sugarcane, vegetables, sod, and rice and other aquatic crops. Early vegetable production did not come easy, because of lack of certain micronutrients (boron, copper, manganese and zinc) in the muck soils. In 1927, not a single peanut was harvested from a 1000-acre plantation, and the foreman of the Everglades Research Station made an offer of $5.00 per bean for every bean to be grown in “sawegrass land”. Each chapter contains an extensive list of references. The book contains 6 photo- graphs, 29 illustrations including maps, graphs and diagrams, and 56 tables. There are author and subject indices and a list of contributors and their affiliations. This book is a valuable resource of information for water quality and use managers, planners, farmers, and other agriculturists, and anyone looking for information on agriculture in the Everglades area. — Barbara B. Martin, University of South Florida, Tampa 188 FLORIDA SCIENTIST [VOL 57 ACKNOWLEDGMENT OF REVIEWERS It is a pleasure to acknowledge the dedication, cooperation, and service of the following persons, who generously donated time and expertise in reviewing manu- scripts for Volume 57. Some reviewers kindly served more than once. Elaine Anderson Daniel Austin Gregory R. Baker Carl Barfield Sue Boinski Dennis Borton Robert S. Braman Richard Cailteux Joseph T. Collins Bruce C. Cowell Jeff C. Davis Frederick B. Essig Arnold B. Grobman Dilna Victor Hale Walter C. Jaap Wayne King Robert Kluson Steve Kostewicz John M. Lawrence James N. Layne Robert Livingston Edward Long Ralph E. Moon Henry R. Mushinsky Robert Powell Paul Shafland Janice Snook Peter Stiling Walter K. Taylor Diane TeStrake Lovett Williams, Jr. Florida Scientist QUARTERLY JOURNAL of the FLORIDA ACADEMY OF SCIENCES VOLUME 57 DEAN F. MarTIN Editor BARBARA B. MARTIN Co-Editor Published by the FLORIDA ACADEMY OF SCIENCES, INC. Indialantic, Florida 1994 The Florida Scientist continues the series formerly issued as the Quarterly Journal of the Florida Academy of Sciences. The Annual Program Issue is published independently of the journal and is issued as a separately paged Supplement. Copyright © by the FLoripa ACADEMY OF SCIENCES, Inc. 1994 No. 4 1994] CONTENTS OF FLORIDA SCIENTIST VOLUME 57 NUMBER ONE, Two Predation of Hatchling Diamondback Terrapin, Malaclemys terrapin (Schoepff), by the Ghost Crab, Ocypode quadrata Fabricius). Il.......... Rudolf G. Arndt Reproductive Cycle of the Indo-Pacific Gecko, Hemidactylus garnotii, in SoM TALC TA IE. avons Same ee etn een, ame A CEO. is Pr he Walter E. Meshaka, Jr. Effect of Treated Kraft Black Liquor on Hydrilla verticillata (Royle)........... Andrew L. Hassell, Barbara B. Martin, Dean F. Martin, and Jess M. Van Dyke Macroinvertebrates of the Northern Everglades: Species Composition and Peon iner SRM LUNG: scr... .ack mea Ae) chi beet h 3. dose a hs caters coal Russell B. Rader Silver Accumulation in Three Species of Fish (Family: Centrarchidae) in Shotemmoake te dbneaeMMmemt PROMOS ..chcsechanes sia sesaes ssndeeo eI oo eRe sesdaeaeh Kym Rouse Campbell rend INVA reso cecncihcdoveo sna cusecesotsoctsreasesoneren ot Joseph A. Stanko The Use of Computer-Assisted Molecular Modeling in College General CLLENITS UPY gee qatosoveesbobnooa sees acura canara a eee eno James R. Yount A Mass Stranding of Leach’s Storm-Petrel in Georgia and Florida ............... Carol A. Ruckdeschel, C. Robert Shoop, and George W. Stiple Yeast Interactions Inferred From Natural Distribution Patterns ................... NUMBER THREE Fog Temporarily Increases Water Potential in Florida Scrub Oaks. ............. Eric Menges Amanageatec bibliography of Lyng byalicctsntasssaesen- see. dseeskuncessnaviccneeneessennuceoens Dean F. Martin, Barbara B. Martin, Elsie D. Gross and Karen Brown Morphometric Quality of Wild Turkeys Harvested from Public vs. Private WN Fadteh GB LAL OV Tal Cl Ve oie se ne AO soo ek svn noise clnon cane eaennceetane David T. Cobb Observations on Reproduction of an Endangered Cactus, Cereus Robinii Brum tetee) Pll mye WS Oi ee tects ee tee ia otrst oie cacti ao aah da casein acidsbbanebenuasounaeeecie Michael K. Hennessey and Dale H. Habeck 19] 1 6 10 22 34 42 43 48 50 62 63 64 65 88 93 192 FLORIDA SCIENTIST [VOL 57 Prevanence of Whipworm (Trichuris) Ova in Two Free-Ranging Popula- tions of Rhesus Macaques (Macaca Mulatta) in the Florida Keys ......... Linda L. Taylor, Robert G. Lessnau and Shawn M. Lehman 102 Patterns of Coral Abundance Defining Nearshore Hard-Bottom Communi- ties of the Florida Keys o...0...c.002 staid. citdn eee rr M. Chiappone and K.M. Sullivan 108 AWATGS .05o.00de08eieccteceecsbbeselanbalilbeQdnn LIA DORA ORI eee 126 BOOK ROME Who scscclecces Reese ee Frederick B. Essig 128 Corrigem™uin isjd. eos. ehh sesssaecescsensseendouieessasuaeee dene 128 NUMBER FOUR Recognition Characters and Juxtaposition of Florida and Mississippi Slimy Salamanders (Plethodon Glutinosus Complex) .......0c.cecssesscssessenesettenetes James Lazell 129 Benthic Invertebrates and Allied Macrofauna in the Suwannee River and Estuary Ecosystem, Florida ..0......:.csessscn ent stieeiibecscndieince er William T. Mason, Jr., Robert A. Mattson and John H. Epler 141 Florida's Tomato Producers: Are They Moving Toward Sustainability? ........ M. E. Swisher and E. P. Bastidas 161 Identification of Sailfin Catfishes Introduced to Florida ..........ccccceeeeeeeeeeee Lawrence M. Page 171 Additions to the Flora of South Florida: Four New Species of Naturalized Tropical Trees :si..cibjcs.d diareeha Miki ee a os John B. Pascarella 173 Selenium Levels in the Palm River and MacKay Bay Region of Tampa Bay Robert F. Benson, Joseph A. Saraceno, David R. Lowell and David W. Brett 177 BOOK REVICW ch cses cccasicdesscdianisieiacscebocdban csdaccs Stanncsence chen ee 186 OOK REWACW. aikcs acs ceeeekeen eee Barbara B. Martin 187 Acknowledgment of Reviewers «.........0.....d0..esssnesnosesessnsnsretnesens3oe eee 188 Index, VOl ST sckecee he a I ee eee 191 INSTRUCTIONS TO AUTHORS Individuals who publish in the Florida Scientist must be active members in the Florida Academy of Sciences. Submit a typewritten original and two copies of the text, illustrations, and tables. All typewritten material—including the abstract, literature citations, footnotes, tables, and figure legends—shall be double-spaced. Use one side of 8'/2 x 11 inch (21'/2 cm x 28 cm) good quality bond paper for the original; the copy may be xeroxed. Margins should be at least 3 cm all around. Number the pages through the Literature Cited section. Avoid footnotes and do not use mimeo, slick, erasable, or ruled paper. Use metric units for all measurements. Assistance with production costs will be negotiated directly with authors of pages which exceed 10 printed pages. Excess over 10 pages by an author in a single volume will result in charges for additional pages. ApprEss follows the author’s address. | ApsTRACT—All manuscripts shall have a short, concise, single-paragraphed abstract. The abstract follows immediately the author’s address. ACKNOWLEDGMENTS are given in the body of the text preceding immediately the Literature Cited section. LITERATURE CITED section follows the text. Double-space every line and follow the format in the current issue. Manuscripts of 5 or less, double-spaced typewritten pages should conform to the short article protocol. See recent issue for proper format of both long and short articles. TABLES shall be typed on separate sheets of paper, double-spaced throughout. Each table shall contain a short heading. Do not use vertical rulings. Maximum character width, including spaces, of tables is 98.0. This includes a minimum space of 5 characters between columns. Tables are charged to authors at $13 per quarter page, $29 per half page, and $39 per full page. ILLUsTRATIONS—All drawings shall be done in good quality India ink, on good board or drafting paper. Letter by using a lettering guide or equivalent. Typewritten letters on illustrations are unaccept- able. Drawings and photographs should be large enough to allow 1/3 to 1/2 reduction in size. Photographs shall be glossy prints of good contrast. Whenever possible, mount photographs in lots size. The author’s name and figure number should be penciled lightly on the back of each figure. Figure legends must be listed on a separate page and not on the drawing or photograph. The legend must be double spaced. Illustrations are charged to authors at $6.00 per page or fraction. We ask authors to prepare accepted manuscripts in personal computer formats. Please do not submit a manuscript on disk. Once reviewers have recommended a manuscript be accepted, you will be given directions for submitting the final manuscript on disk. Proor must be returned promptly. Notification of address change and proofreading are the author's responsibility. Alterations after the type has been set will be charged to the author. REPRINTS may be ordered through the FAS office from the printer on forms provided at the time that proofs are sent to authors for checking. FLORIDA ACADEMY OF SCIENCES INSTITUTIONAL MEMBERS FOR 1994 Archbold Biological Station Florida Institute of Technology Brevard Community College, Florida Advanced St. Petersburg (Florida) Public Library Technology Center Library, Palm Bay St. Thomas University, Miami, FL The British Library, Boston Spa, Weatherby, University of Miami (?) UK University of South Florida CoRPORATE MEMBERS FOR 1994 Mote Marine Laboratory, Sarasota Membership applications, subscriptions, renewals, changes of address, and orders for back numbers should be addressed to the Executive Secretary, Florida Academy of Sciences, P.O. Box 033012, Indialantic, FL 32903-0012 No. 4 1994] PUBLICATIONS FOR SALE by the Florida Academy of Sciences Complete sets. Broken sets. Individual numbers. Immediate delivery. A few num bers reprinted by photo-offset. All prices strictly net. Special bargain: $500 for near complete sets. Prices quoted include domestic postage. Some issues may not b available. All are $20 per volume or $5 per issue, except for symposium issues. PROCEEDINGS OF THE FLORIDA ACADEMY OF SCIENCES (1936-1944) Volumes 1-7 QUARTERLY JOURNAL OF THE FLORIDA ACADEMY OF SCIENCE (1945-1972) Volume 8-35 FLORIDA SCIENTIST (1973) Florida’s Estuaries—Management or Mismanagement?—Academy Symposium FLORIDA SCIENTIST 37(4)—$5.00 Land Spreading of Secondary Effluent—Academy Symposium FLORIDA SCIENTIST 38(4)—$5.00 Solar Energy—Academy Symposium FLoripa SCIENTIST 39(3)—$5.00 (includes do-it-yourself instructions) Anthropology—Academy Symposium FLORIDA SCIENTIST 43(3)—$7.50 Shark Biology—Academy Symposium FLORIDA SCIENTIST 45(1)—$8.00 Future of the Indian River System—Academy Symposium FLoripA SCIENTIST 46(3/4)—$15.00 Second Indian River Research Symposium—Academy Symposium FLORIDA SCIENTIST 53(3)—$15.00 Please send payment with order. If required, an invoice will be sent on purchas | over $20 from a recognized institution. Please include Florida sales tax if shipped tc} Florida address, unless a copy of a valid Florida Consumer’s Sales Tax Certificate’ | enclosed. NON-PROFIT | ORGANIZATION | The Florida Academy of Sciences, Inc. P.O. Box 033012 Indialantic, Florida 32903-0012 U.S. Postage PAID Permit #26 Melbourne, FL Pa hye eS Fe ONS SO ae OP ated de des oe ne T ; mit Me0OMLan Inetitutian Puree Bat ee Ba eae BN ise : : : Wibrary - Acquisitions Room 25 ANHE Siig oe Rommel om aoe te Saas Washington DC 2os4e¢ 0nd hy diel ees OL F AAV Ycl Kier pie ‘7 inen:~ ee ITION 4a TION igi TION 41 yet ITION ‘O07 ‘TES SMITHSONIAN ILSNI ASTID LS ria o> NVINOSHLIWS ae ted — = zt ie = Ey > 0 rs “7 AO as = ae o = wo NOILNLILSNI 7) rd = << : LU 3 0 gs Ka > GQ = =o o _ LIBRARIES SMITHSONIAN __ Zz if 2 ia age Ta Ay, is a = “ Gup = = “GY; = o v0 mem O = as z= NOILALILSNI NVINOSHLINS ; hig ae e : = = a = e ~ - 2 ae ie LIBRARIES SMITHSONIAN ee WM = ; = < : = = NA 2 g ws = NOILAL ” Ww LSNI NVINOSHLINS 2 a ql ac = ‘i < pa <= cn = a a g a SSIYVYUYSIT LIBRARIES , c z 4 o = Oo 30 = Pe > = > 0] i a = 2 | w = O INSTITUTION NOILALILSNI _ NVINOSHLIWS, |S; ‘ = =< = = . 5 = ue 2 go Zz eS 27% S3luvYyg ei BRARI ES SMITHSONIAN _! i = rm) yy aa = a Yi a f Cc ie: c X a a o oa Oo 2 - 25 INSTITUTION NOILALILSNI Sa : : sq — ow = ra rm oe = 2 > | jw a = (72) YO Zz os Zz] SSIYVYEIT LIBRARIES IN: Pd eis w” 2 a= a < WA = th = > = a oH on ge a INSTITUTION | NOILALILSNI_NVINOSHLNS Ti = a ul oc “ oc | ra ‘= ie me = os = sf INSTITUTION NOILALILSNI_ NVINOSHE _* 2 5 _ 0 2 2 = «gn >" = S3JiIu¥VvYyg iL BRARI ES SMITHSON wn = 7) us Pe rm) = wo nen ~ + me a) 2 -_ > ra > o m 2 m 4 m = ” om < Ww Zz w I1_LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S31) = ‘y ” a ® ~” a 2) = = < = < Sh, = x 5 a, oe x 3 z 8 NSE S aap Be = BS ae) r Be SANE OO bs 8 oO 4 = SS = = \ >" = > ON NOILMLILSNI_NVINOSHLINS S3INVYSIT LIBRARIES INST ud 2 uJ = # z = = a 4 ie +4 1 4 = i = 65 ) § : te S : € m0 5 oa wy 5 or = = z2 o z = = i] LIBRARIES INSTITUTION NOILNLILSNI NVINOSHLINS S3)]2 r z s z 00 er) ° = fe \ x cad = rag 5 Oo o z OD . = IN NOILNLILSNI SSIYVUGIT LIBRARIES SMITHSONIAN INST! wo a w a wee". w < . = re = = Ws. = = 5 z A = AWS =. = S 2 a = = YY = E > ; J = > = ‘ > = z ry) = 2) Sey eee ”) 117 LIBRARIES SMITHSONIAN NOILNLILSNI NVINOSHLIWNS S314 B “ 0 2 ge “ a “Ss z . XY = fa = 3 = = =i @ coe <= Oo \ x : © za : OD 2) a ~~. N . ow in 7) a fe) rT WAN Oo Sere 8 bY = Zz = WN Zz, = ZY = > = NY >’ = > ” z a) Rae = w” 2 N —NVINOSHLINS Saluvaaly LIBRARIES SMITHSONIAN INSTI) ” tie = ” f z= te el Oe 2 Oe gm es ” w c < c a - z o = = E = ” = z < z= fo] ” x= E = n 3 9088 01354 2014 "Mee