The Ecology of Hydric Hammocks - USGS National Wetlands ...

Copies of this publication may be obtained from the Publications Unit, U.S. Fish and Wildlife Service, 18thand C Streets, N.W., Mail Stop 1111, Arlington Square Building, Washington, DC 20240, or may bepurchased from the National Technical Information Service (NTIS), 5285 Port Royal Road, Springfield, VA

T,I"?ynryNa! :::8azm/etland3 Eeaeareh CenterU. S. Fish 2nd wildlife Service101 5 Gawe R~alcvardelidell, LA 70458Biological Report 85(7.26)September 1989THE ECOLOGY OF HYDRIC HAMMOCKS: A COMMUNITY PROFILESusan W. VinceStephen R. HumphreyRobert W. bimonsFlorida Muscum of Natural tli storyUniversity of FloridaGaincsvillc, FL 32611Project OfficerJames A. AllenNational Wetlands Research CentcrU.S. Fish and Wildlife Service1010 Cause BoulevardSlide1 1. LA 70458Performed forU.S. Department. of the InteriorFish and Wild1 ife ServiceResearch and DevelopmentWashington, DC 20240

DISCLAIMERThe opinions and recommendations expressed in this report are those ofthe authors and do not necessarily reflect the views of the U.S. Fishand Wildlife Service, nor does the mention of trade names constituteendorsement or recommendation for use by the Federal Government.Library of Congress Cataloging-in-Publication DataVince, Susan W.The ecology of hydric hammocks : a community profile / by Susan W. Vince,Stephen R. Humphrey, Robert W. Simons; project officer, James Allen ; performed forU.S. Department of the Interior, Fish and Wildlife Service, Research andDevelopment.p. cm. -- (Biological report ; 85 (7.26) (Sept. 1989))Includes bibliographical references.Supt. of Docs. no.: 149.89/2:85(7.26)1. Wetland ecology--Florida. 2. Forest ecology--Florida. I. Humphrey, StephenR. 11. Simons, Robert W. 111. Allen, James A. IV. Title. V. Title: Hydrichammocks. VI. Series: Biological report (Washington, D.C.) ; 85-7.26.Suggested citation:Vince, S.W., S.R. Humphrey, and R.W. Simons. 1989. The ecologyof hydric hammocks: a community profile. U.S. Fish W i ldl. Serv.Biol. Rep, 85(7.26). 81 pp.

PREFACEThis community profile is one in a series of Fish and WildlifeService publ ications compiled to provide a state-of-knowledge synthesisof scientific information and 1 i terature on various coastal habitats.The subject of this profile is the hydric hammock, a distinctive type offorested wetland occurring at low elevations along the gulf coast ofFlorida from Aripeka to St. Marks and at various inland sites inFlorida.Re1 atively 1 i ttle research has been conducted in hydric hammocks,and no thorough effort has been made previously to define thiscommunity. Consequently, no consensus has existed about the extent andnature of this community; some publ i shed works and active researchershave differed in their judgments about it; and the entity sometimes isignored and often is lumped with other types of mixed hardwood forests.The purpose of this profile is to establish or clarify an identificationand understanding of the hydric-hammock community. Information for theprof i 1 e was gathered from publ i shed and unpubl i shed 1 i terature, frompersonal communication with many technical experts, and from our ownfield experience. The profile includes some new data gathered in thefield for the purpose of defining this community.It is hoped that the content and format of the profile will beuseful to a broad spectrum of users, including other scientists,students, resource managers and planners, teachers, and interestedcitizens. The profi 1 e includes structural and functional aspects of thecommunity: its physical setting, plant and animal composition anddynamics, interactions of its flora and fauna, and its relationshipswith other communities.

CONVERSION TABLEMetric to U.S. Customarymillimeters (mm) 0.03937centimeters (crn) 0.3937meters (tn) 3.281meters (m) 0.5468kilometers (hn) 0.6214kilometers (hn) 0.5396square meters (m2) 10.76square kilometers (km2) 0.3861hectares (ha) 2.47 1liters (1) 0.2642cubic meters (m3) 35.31cubic meters (m3) 0.0008110milligrams (mg) 0.00003527grams (g) 0.03527kilograms (kg) 2.205metric tons (t) 2205.0metric tons (1) 1.102kilocalories (kcal) 3.968Celslus degrees (-C')i .8(-C)+32inchesinchesfeet (ft)fathomsstatute miles (mi)nautical miles (nrni)square feet (ft2) 2square miles (mi )acresgallons (gal)cubic feet (ft3)acre-feetounces (oz)ounces (oz)pounds (Ib)pounds (Ib)short tons (ton)British thermal units (Btu)Fahrenheit degrees ( O F )Y.S. Customarv to Metricinchcsinchcsfeelfathomsstalute milesnautical milessquare feetsquare milesacresgallonscubic feetacre-feetouncesouncespoundspoundsshort tonsBritish thermal unitsFahrenhe~t degreesmillimeterscentimetersmetersmeterskilometerskilometerssquare meterssquare kilometershectaresliterscubic meterscubic metersmilligramsgramskilogramsmetreic tonsmetric tonskilocaloriesCelsius degrees

CONTENTSPREFACE ..........................................................CONVERSION FACTORS ...............................................FIGURES ..........................................................TABLES ...........................................................ACKNOWLEDGMENTS ..................................................CHAPTER 1 . INTRODUCTION .........................................CHAPTER 2 . PHYSICAL SETTING .....................................2.1 CLIMATE .............................................2.2 PHYSIOGRAPHY AND GEOLOGY ............................2.3 SOILS ...............................................2.4 HYDROLOGY ...........................................2.5 PHYSICAL CONTROL OF DISTRIBUTION OF HYDRIC HAMMOCKS2.6 SUMMARY .............................................CHAPTER 3 . VEGETATION OF HYDRIC HAMMOCKS ........................3.1 INTRODUCTION ........................................3.2 VEGETATION PATTERNS .................................3.3 SOURCES OF VARIATION ................................3.3.1 Geography ....................................3.3.2 Hydrology ....................................3.3.3 Edaphic Conditions ...........................3.3.4 Fire .........................................3.3.5 Other Disturbances ...........................3.3.6 Summary ......................................3.4 PRIMARY PRODUCTION AND ITS FATE .....................CHAPTER 4 . ANIMALS ..............................................4.1 INTRODUCTION ........................................4.2 REPTILES AND AMPHIBIANS .............................4.3 BIRDS ...............................................4.3.1 Community Structure ..........................4.3.2 Selected Species .............................4.4 MAMMALS .............................................4.4.1 Community Structure ..........................4.4.2 Selected Species .............................CHAPTER 5 . PLANT-ANIMAL INTERACTIONS ............................CHAPTER 6 . LINKAGES WITH OTHER ECOSYSTEMS .......................6.1 WITH ESTUARIES ......................................6.2 WITH ADJACENT AND DISTANT HABITATS ..................REFERENCES ....................................................Paqeiiii vv ivi iii x155691116172020253535364041454848

FIGURESNumber1AL...............Distribut,ion of hyiiric ltanlrnocks in Florida.Sc&53!::;.a- , ;~i-.f:.{!~:.+ - , ,,.,. _ -, . _.:,,..: ci % :.- ,C a r.l;r.If31!: Ci : :::.-,t th?-pt? Inrntinn'; in=Florida, 1941.70 average. .................................Physiograpt~ic 1-eg ions of Florida. ...................--.-..Geologic cross section of peninst11 ar northern ~lorida. - . - .Geologic map of Florida showing don~inant rocks at or nearthe sul-face. ..............................................Marine tert-acps of Florida. ...............................Limestone outcrop in a gulf coastal hammock. ..............Sequence of plant corrlrntrnities From the lower St. JohnsRiver to upland scrubby flat.woods. ......................-.General ized profile of a spring-fed strearn and adjoininghydric k:ck ............................................Rainfall and ground-water level in 1986 at Tiger Creekhydric hammock, central Florida. ..........................Water- table profiles across three plant colnmuni t iesadjacent to l igcr Creek. ...............................-..Flooding of hydric hammock along the Myakka River, SarasotaCounty, Florida. ..........................................Flooding and tfrydown of Sanchez Prairie. ..................Two vegetation tt'ansects fro111 river swamp to crpl and forestin the Okl awaha Rivcr basin, central Florida. .............Topographic map of the upper St.. Johns River floodpl ai nbctweftn Pt!r.z!~ !?4n and l a k ~ Pninsrft. ....................Aerial phot,ograph of t~ydric hammock on the west bank ofthe St. Johns River .......................................A large swt?c?tguni in dense hydric hammock a'long Tiger Creek,Florida ...................................................Cabbage palm predoainates in hydric hammock along UpperMyakka Lakc., Myakka River State Park. .....................Hydric harnrnock at the in1 and edge of Gulf Hammock ........Hydric hammock along the upper St. Johns River. ..........*Coastal hydric hammock where it adjoins salt marsh, Gu7 fHammock, Levy County. .....................................Loblolly pine hydric hammock, Silver Springs, MarionCounty. ...................................................vertical structure of hydric hammock dominated by ? ob1 ol 1pi ne. .....................................................A hydric hammock strongly affected by seepage, WekivaSprings, Semi no1 e County. .......................... .......Species composition of trees, shrubs, and saplings in twohydric hammocks. ................................... .......Edge of hydric hammock and salt marsh along the GulfMexico near the Wi thl acoochee River, Citrus County , . - - . - -.Natural ranges in Florida of four species of trees commonhydric hammock. ................................ ........of cabbage palm in Florida ...................~istributionof

Stitti-~~~~dr 06t.s 6-f "Lr'ceb ;,I hydt ;i ii&tlllttii~A. ..............Distribution of tree species a1 ong a presumed floodinggradient in Sanchez Prairie, San Felasco Hammock StatePreserve ..................................................Expansion of hydric hammock into freshwater marsh, MyakkaRiver State Park. .........................................Expanding hydric hammock, Myakka River State Park. ........Change in the tree-species composition of a hydric hammockwith increasing distance from its salt-marsh boundary .....Fire in the cabbage palm edge of hydric hammock, SeminoleRanch .....................................................?re?-species composition of two hydric haanocks in VyakkaRiver State Park.. .....................................Species composition of four nearby portions of a hydrichammockstand in the northern part of Tosohatchee StatePreserve, Orange County. ..................................A grazing exclosure in Gulf Hammock demonstrates theeffects of cattle and deer on the ground cover of hydrichammock...................................................Tree bl owdowns due to hurricanes in coastal hydri c hammock.Phenology of bird diversity and abundance and the number ofspecies of plants producing fruits edible to birds in amesic hammock .............................................Breeding and winter ranges of selected migrant birds thatlive in hydric hammock, ...................................

NumberTABLESPaqeClassifications of freshwater, forested wet1 ands in FS orida. 2Characteristics of some soil series associated with hydrichammockvegetation. ....................................... 10Comparison of surface soil characteristics among threehardwood wetland types in Florida ......................... 11Plants occurring in hydric hammocks. ...................... 2 1Composition of hydric-hammock stands. ..................... 2 5Fire sensitivity in five tree species common to hydrichammocks .................................................. 42Occurrence of reptiles and amphibians in three variants ofhydric hammock. ........................................... 5 2Occurrence of reptiles and amphibians in a loblolly pinehydric hammock. ........................................... 54Occurrence of reptiles and amphibians in two drift fencearrays at a hydric-hammock site. .......................... 5 5Occurrence of reptiles and amphibians in two drift fencearrays in coastal hydric hammock, loblolly pine variant. .. 5 5Bird popul ati ons and community characteristics a1 ong theroute of the proposed Cross-Florida Barge Canal. .......... 57Occurrence of mammals in three variants of hydric hammockalong the route of the proposed Cross-Florida Barge Canal. 61viii

ACKNOWLEDGMENTSWe are grateful to G. Ronnie Best, Andre Clewell, Robert Dye,Steve Gatewood, J.D. Slabaugh, Robert Tighe, and Charles Wharton foroffering information based on their studies and experience. We thar?kKarl a Brandt , W i 11 i am Buckner, Nancy Szot, Andrew Vi nce, and Gordon Wardfor their assistance in the field and G. Kenneth Scudder for showing usnumerous field sites and providing photographs. We appreciate theadvice of David L. Auth, Katherine C. Ewel, John A. Fluno, and Paul E.Moler on ways to improve the manuscript.

CHAPTER 1. INTRODUCTION"Hammock, hommock, or hummock?"queried R. M. t-iarper in 1905; his artswerwas one of many attempts bybotanists to reduce confusion over theclassification of southeastern plantcommunities. Harper (1905) concludedthat a hummock is a geographic feature,a rounded knoll, whereas"hammock" and "hommock" are synonymsfor dense, hardwood forests that occurin limited areas amid the wetprairies, marshes, and pine forests ofthe coastal plain. The etymology of"hammock" and "hommock" is obscure(Oxford Engl i sh Dictionary 1933).Columbus' sailors learned to use net"hammocks, " hanging beds, fromCaribbean Indians; perhaps they misappliedthe Arawak term, hamacas, intendedfor the woods in which thesleeping nets were hung. Bartram(1791) and other early expositors ofFlorida's natural history employed the"hommock" spef 1 ing, perhaps in keepingwith native pronunciation, but"hammock" became the prevalent form inthe writings of the twentieth century.Hammocks are a small but distinctpart of the natural landscape of theFlorida peninsula north of Lake Okeechobee.In this region, the uplandsare dominated by three fire-adaptedcommunities: (1) a savanna of wiregrass(Aristida stricta) and otherherbs with an overstory of longleafpine (Pinus palustris) and scattereddeciduous oaks on we1 1 -drained deepsands; (2) a scrub community consistingof a sand pine (Pinus clausa)overstory and dense evergreen understoryon deep sands; and (3) pineforests (pine flatwoods) with an understoryof saw palmetto (Serenoarepensf, gall berry (Tlex glabra) , andother shrubs and herbs on flat, poorlydrained sites. Forests of bay trees~ y ~ r ~ r s ~ ~ ~ 2 :d : ksyswamps or bayheads, occupy depressionson the sides or at the bottom ofslopes where soils are kept moist orwet by seepage, Along drainages, indepressions in the uplands, and inother low areas are swamps andmarshes. Of lesser extent, and oftenprotected from fire by nearby bodiesof water, are the scattered patches ofmixed hardwood forest called hammocks(or "routhcrn mixed hardwood fcrest, "Monk 1965).Some hammocks reside on poorlydrained soils or on soils with highwater tables, and occasional floodingsaturates the soils for a time sufficientto modify plant species composition.These are hydric hammocks.Other names for this wetland forestinclude "1 ow hammock" (Harper 1915),"wetland hardwood hammock" (Soil Conservatian Service 1981), and " low1 andoak hammock" (Dunn 1982).Hydric ham-mocks often have a broad-leaved evergreenappearance and typically contain1 ive and swamp 1 aurel oaks (Quercusvirginiana and Quercus laurifol ia),cabbage palm (Sabal palmetto), southernred-cedar (Juniperus silicicola),sweetgum (Liquidambar styracif 7ua) ,and hornbeam (Carpinus carol iniana) .Hydric hammocks are distinguished fromthe drier mesic hammocks by the absenceor scarcity of a number of treespecies including pignut hickory(Carya glabra) , hop hornbeam (Ostryavirginiana) , winged elm (Ulmus alata) ,and southern magnol ia (Magnolia grandiflora).While some hydrophytic treespecies common to mixed hardwoodswamps, such as red maple (Acerrubrum) and swamp tupelo (Nyssa sylvaticavar. biflora), are shared byhydric hammocks, others, notably ashes

\fr.a2;1i1:~s spp. ) and tldlil cvprcss!T7% .' -1 i?l~~tt~h~m). are rare, Eiyc:r rc ~dtr~acks fr-ctjuently are situ~ledon gentle slopes betwecr~ rneslc h,nfn.ltockor plne flatwoods and r~vcr swamp, wetpra~rie, or marsh. The prcctbc houndaries between the rnnmlun~ tit.:, are eliisive, yet a dist~nct cctmbrr~atlnn ofplants, animals, and pl-iyslcal conditronscharacterize Iiydrlc har:?mr\rks andmake then recogr>izable in ilreat vluss,the three classificationsagree that hydri c hammocks di ffer frommost oth~r wetlands in their vegetatlor1 and tiydroi ogy. Swamps of mixedhardwoods or cypress tend to bef iooded more frequently and regularly,for a longer period, and to a greaterdepth than hydric hammocks. Bayheads(a1 so call ed bay swamps or baygalls)are distinguished by their constantseepage, peat substrate, and abundanceof bay trees. River flow influencesfloodplain and bottom1 and forests,Table 1. Classifications of freshwater, forested wetlands in Florida. Communities on the same line areapproximately equivalent. In some cases, a class in one list Is Subdivided into several classes inanother list.So i 1 Conservat ionService (1981)f lorida Natural Areas Wharton et a?. (1977)Inventory (1984)Bottomland hardwoodsSwamp hardwoodsCypress swampScrub cypressShrub bog-bay swampPitcher plant bogWet1 and hardwood hammockCabbage palm hammockFloodplain forest;Floodplain swampFreshwater tidal swampStrand swampDomeBasin swampBayy a1 1 ; BogSeepage slopeHydri c hammockWet fl atwoodsBottom1 and forestAlluvial river swampBlackwater river swampBackswampSpringrun swampTidewater swampCypress strandCypress pond (dome)Gum pond and swampLake border swampDwarf cypressBay swamp; Shrub bogHerb bogBog swampSeepage swampSouth Florida hammockHydric hammock

whereas hydric hammock is a still-waterwetland (Wharton et a7. 1977).Typicz??;~ on alluvial fIoodp:afns,bottomland hardwood forests contain anumber of species that are absent fromhydric hammocks, notably overcup oak(Quercus lyrata), water hickory (Caryaaquat ica) , cottonwood (Popu7us heterophy7la), and sycamore (Platanus occidentalis).The most extensive stands of hydrichammock are found in Florida along theGulf of Mexico from Aripeka in thesouth to St. Marks in the panhandle(Figure 1). Almost a continuous belt,the hammock forest lies just landwardof salt marsh. Another large standinhabits part of the west bank of theupper St. Johns River and, inland ofcoastal dunes, a narrowr strip of hydrichammock parallels the Atlanticshore of northeastern Florida. Manysmall er stands are scattered about,primarily in the northern and centralregions of peninsular Florida. Hydrichammocks are scarce in the panhandlewest of St. Marks, and their southernlimit is subject to debate. Hydrichammocks of cabbage palm, live oak,swamp laurel oak, and red maple to thewest and north of Lake Okeechobee weredescribed by R. Harper (1927); probablythis is the southern extent oftheir range (Figure 1). Farther tothe west, hydric hammock dominated byORANGE LAKESILVER SPRINGSINOLE COUNTYEKlVA SPRINGSHILLSBOROUGHRIVERMYAKKA RIVERIFigure 1. Distribution of hydric hammocks in Florida. Locations of the largest stands and thosesampled in the present study are shown. Numerous stands too small to delineate at this scale occurthroughout the northern and central sections of the peninsula. Dashed line is the contour of 15-melevation.

cabbage palm and live oak adjoins themarshes of the Myakka River. We followthe examples of R. Harper (1927)and Davis (1943) in differentiatingthe temperate-subtropical hydric hammocksfrom the tropical hammocks ortree islands common to southernFlorida. The tropical hammocks featurea diverse assemblage of West Indian,flood-intolerant plants (Olmstedand Loope 1984). They also may inhabitslight mounds within marsh,prairie, and mangrove vegetation,physiographic settings different fromthose of most hydric hammocks. TheFakahatchee strand in southwesternFlorida is difficult to classify becauseit contains stands of cabbagepalm and swamp laurel oak that resemblehydric hammock. However, theseforests most likely are transientstages resulting from the harvestingof cypress.Hydric hammocks seem to be rare outsideof Florida. Wharton (1977) notedonly small patches of comparable forestin Georgia, mainly in the coastal1 owl ands near Savannah. Reviews ofthe vegetation of South Carolina(Barry 1980) and the lowland forestsof its northern coa'stal pl ain (Jones1981) included no descriptions offorests similar in vegetation and hydrologyto hydric hammocks. A1 thoughHarper (1905) reported that hammocksranged from North Carolina to Floridaand Mississippi, the vegetation surveysof Georgia and South Carolinasuggest that, beyond Florida, the hammocksare mainly of the drier mesicand xeric types.A1 1 wet1 ands--forested and nonforested,saltwater and freshwater--covered more than 8.3 n~i 71 ion acres ofFlorida, about 22% of its land area,1472-73 i;' ---a- '""A'la1t1paVat A J u Y j . Fur eskdwetlands accounted for about 2.1 millionha; of that amount, hydric hammockprobably compri sed 120,000 to140,000 ha. Hydric hammock is estimatedto cover 80,000 to 100,000 ha atpresent (Simons et a7. 1988). A precisedetermination is not availablebecause the boundaries of hydric hammocksare difficult to delineate onaeri a1 photographs. Neither can dataon its extent be exlrd,td from theperiodic inventories of southeasternforests (e.g., Bechtold and Knight1982), because hydric hammock does notcorrespond to a U.S. Forest Serviceforest type or physiographic class.However, it is evident that the areaof hydric hammock has significantlydeclined in modern times, due primarilyto clearing for agriculture, realestate development, and, especi a1 ly inthe past 20 years, for pine plantations.The management guide that accompaniesthis profile details thevarious uses of hydric hammocks, bothdel i berate and inadvertent a1 terati onsof this community, and the impacts onforest structure, wild1 ife habitat,and hydrology (Simons et al. 1988),This community profile summarizes whatis known of the ecology of hydric hammocksand includes new data on thevariation of plant composition amongstands. Since hydric hammocks havebeen 1 i ttle studied, many inferencesare made from similar communities.The scarce scientific attention affordedhydric hammocks in no way reflectstheir value. The beauty of hydric-hammockforests and their provi -sion of wildlife habitat, timber, andflood control argue for their preservationand sound management.

CHAPTER 2. PHYSICAL SElTiNG2.3 CLIMATEFlorida boasts a subtropical climate,but that is technically correctonly for the southernmost and coastalportions of the state (Dohrenwend andHarris 1975). Winter temperatures innorthern Florida are much lower thanin the southern part of the State:mean minimum values are 5 and 16 OC,respectively (Jordan 1984), and extremesin the north range well belowfreezing. Low-temperature extremes inwinter require many plants and animalsto overwinter in typical temperatezonemanner. Summer temperatures, influencedby warm, moist air from theGulf of Mexico and the Atlantic Ocean,are re1 atively uniform throughout thestate. In July the mean daily maximumis about 31 to 33 OC and the meandaily minimum ranges from 22 to 24 OC.The average growing season is 270 daysin the north and more than 320 days inthe south (Tanner and Smith 1981).Rai nfall is plentiful everywhere inFlorida, averaging between 122 and 152cm annually (Jordan 1984), but itsdistribution is highly uneven over theyear (Figure 2). Rainfall is at amaximum in summer and a minimum inspring (Apri 1 -May) and fa1 1 (October-November), when condi ti ans becomedroughty in some years. SouthernFlorida receives most of its rain inthe period from June through September,and its hydric hammocks may floodthen. Northern Florida receives anadditional pulse of rain in late winter.Because of 1 ewer evapotranspi rationrates, winterprobablycontributes as much or more thansummer rain to surface- and groundwatersupplies. Hydric hammocks innorthern Florida often are flooded atFigure 2. Seasonal distribution of rainfall at threelocations in Florida, 1941-70 average (from Jordan1984).the beginning of the growing seasondue to the late winter rains. Especiallyalong the gulf coast, hydrichammocks also may flood from hurricanesin late summer and early fall.Thunderstorms prevail on 70 or moredays each year in Florida (Jordan1984) and bring 1 ightning as well asrain. Bolts of 1 ightning can strikeand kill individual trees and ignitefires. Fires are re1 atively uncommonin hydric hammocks, but their frequentoccurrence in some adjacent communitiesaffects the extent and compositionof hydric hammocks. Before humansactively Suppressed fires andbuilt roads and other barriers totheir spread, fires must have burned

far more frequently and extensivelyacross the Florida 1 andscape.2.2 PHYSIOGRAPHY AND GEOLOGYFlorida appears flat and featurelesscompared with the dramatic 1 andformsof the western United States. Allsites are close to the sea horizontallyand vertically: no place ismore than 120 km from the At1 anticOcean or the Gulf of Mexico, andFlorida's highest point is only 105 mabove sea level. Much of the state iselevated 15 m or less (Figure 1). Minorchanges in topography--di fferenceson the order of centimeters ratherthan meters--greatly affect the nature,extent, and distribution ofFlorida's plant communities.Florida lies within the Gulf and Atlanticcoastal-plain province and containsfive major physiographic regions(Figure 3). Finer geomorphic divisionsare described by Puri and Vernon(1964) and White (1970). Maximum localre1 ief occurs within the westernMARIANNALOWLANDSFigure 3. Physiographic regions of Florida. TheFloridan Plateau includes Florida and its submergedcontinental shelves. The edge of theplateau is indicated by the go m depth contour(from Cooke 1939).highl ands of the panhandle, whererolling uplands are cut by narrow,steep-wai i eu biftjd~u vdi 7 ey~, and 1 snext greatest in the central highl andsthat run north and south through muchof the peninsula. Fringing the entirepeninsula and extending to the extremewest of the panhandle, the coastallow1 ands have 1 i ttl e re1 i ef and generallyrise slowly via step-like terracesto the central ridges. The landsurfaces of the Okeechobee plain ofsouthern Florida and the Big Bend area(northeastern corner of .ehe Gulf ofMexico) are virtually flat.Florida's topography and physiographicdivisions have resulted fromthe interplay of sediment deposition(mainly marine), erosion, and solution.Since at least the early Cretaceousperiod, the Florida peninsulahas existed as part of the muchbroader Floridan Pl ateau, a stablecarbonate platform (Puri and Vernon1964). The relative amounts of landand water on the plateau have variedwith fluctuating sea level; at present,the emergent peninsula and thesubmerged continental she1 f that ringsthe state are roughly equal in area(Figure 3). The Floridan Plateau wasformed in warm shallow seas by the depositionof thousands of feet of Cretaceousand Tertiary limestone anddo1 omi te (Figure 4). Some evapori tes(e.g., anhydri te) a1 so were deposited,but only small amounts of clastic(sand, silts, clay) sediments accumulated.The deposits amassed atop thecoastal-plain foundation of sedimentaryand igneous rocks and theirthickness and slope were influenced bythe Peninsular Arch (Figure 4). Extendingdown the peninsula from southernGeorgia to the southeast, the Archis Florida's dominant subsurfacestructure. This deformation probablyresulted from crustal stresses thatwarped the coastal -plain floor duringthe 1 ate Paleocene or early Mesozoic(Applin 1951). In comparison, theoverlying carbonate deposits have experienced1 ittle deformation. Outcroppings of Eocene and 01 igocene1 imestones indicate the Ocal a Up1 i ftin west-central Florida (Figures 4 ands

Recent and PIe~sIoLene sands clay marl peatM~~cene-PI~~cene sands clay marl PDOSP~~IIC rockEocene IhmesloneQm Recent and Pleistocenemarine and freshwatermarls, peats, sands, muds .01 PleistoceneQs Pleistocene shell hash.clay and sandPC Pliocene clasticsMC Miocene clasticsMIMiocene limestone01 Oligocene limestoneEl Eocene limestone ciPaleocene l8mestoneTJCretaceo~S Sedtmonlary rocksFigure 4. Geologic cross section of peninsularnorthern Florida. Vertical scale is exaggeratedabout 160 times (adapted from Faulkner 1973).Figure 5. Geologic map of Florida showingdominant rocks at or near the surface. Areas withsurface limestone are shaded (adapted from Vernonand Puri 1964; Tanner and Smith 1981).5), the State's major surficial structure(Vernon 1951).The beginning of the Miocene periodin Florida was marked by a dramaticchange in depositional patterns (Puriand Vernon 1964; Winker and Howard1977; Riggs 1984). A flood of terrigenoussediments began to enter thepeninsula from the north; the sandsand clays were transported mainly tothe east and southeast. Deposition ofcarbonate sediments, predominantly insouthern and eastern Florida, continuedintermittently. Phosphoritesformed .on shal low marine platforms,especially around the Ocal a Up1 ift(Riggs 1984). By the end of theMiocene, a blanket of sand, relativelyimpermeabl e cl ay, and marl coveredmost of the carbonate plateau. Thecrest of the Ocala Up1 ift probably remainedabove sea level during much ofthe Miocene and all of the Pl iocene,and received no sediment deposits(Vernon 1951). Sand depositionreached its maximum extent during thePl iocene, forming a 1 arge beach-ridgeplain in the panhandle and parallelshoreline ridges that flanked thesides of the north-south axis of thepeninsula (Winker and Howard 1977).Thus, Florida's high1 ands were born;even today their surface rocks showthe dominance of Mio-Pliocene depositsof clastic sediments (Figure 5). Thelongshore transport of sand in Mioceneand more recent times was far greateralong the Atlantic coast of the peninsulathan on the gulf shore, shown bythe significantly thicker strata ofMiocene and post-Miocene sediments beneaththe eastern half of the peninsula(Figure 4). The surface of theAt1 antic coastal low1 ands i s coveredmainly by Pleistocene and Recent sanddeposits (Figure 5). In contrast, thecoastal lowlands along the Gulf ofMexico, from Pasco County in the southto just past St. Marks in the easternpanhandle, retain Tertiary 1 imestoneat or near the surface. Solution ofthe carbonates has resulted in the developmentof a karst plain becauseboth the water table and the limestoneare at or near the land surface. This

is precisely the region where hydrichammocks are most extensive (Figure1).Sea-1 evel fluctuations have a1 sogreatly influenced Florida's topography.The Mia-Pl iocene shore1 ineridges are the oldest and highest of aseries formed by the regressing ocean.Overall , sea 1 evel has dropped duringthe past 10 to 15 million years, butthe process has not been continuous.Reversals of the 1 ong-term decl ine,most evident during the Pleistoceneperiod, occurred when sea level wasa1 ternately 1 owered and rai sed by thedevelopment and subsequent me1 t i ng ofglacial ice in high latitudes. At theheight of the Wisconsin glaciationabout 17,000 years ago, sea level mayhave dropped to 130 m below the presentmean (Mil 1 iman and Emery 1968) ;during the preceding intergl aci a1 period,the ocean rose to a level 8-11 mabove the present value. When the seawas stationary for a long period,waves and wind cut seaward-facing escarpmentsand built dunes and shore-1 ine ridges. Adjacent, shallow oceanfloors were eroded, forming nearlylevel marine terraces that persistedwhen sea level fell again (Figure 6).The Pam1 ico marine plain, constructedduring the Pleistocene, includesFlorida's land surface lying between 3and 8 m in elevation. Much of itsarea in the gulf coastal counties ofDixie, Levy, and Citrus is occupied byhydric hammock (Vernon 1951; Puri eta?. 1967). On the Atlantic side ofthe peninsula, strips of hydric hammockare found, instead, on the morerecent Silver Bluff terrace (U.S. SoilConservation Service 1980).The advances and retreats of the sealeft their imprints on other featuresaf Florida's environment as we1 1 .When the sea was low, especially duringPliocene and Pleistocene times,surface drainage rapidly developed,eroding soft sediments that coveredthe 1 imestone bedrock (Stri ngf iel d andLe Grand 1966; Faulkner 1973). Inmany places, the 1 imestone became exposed,infiltrated with water, andriddled with solution cavities. Circulationof water in the Tertiary03 Wlcornlco, Penholoway,Talbot@j Pamllco0 Silver BluffFigure 6. Marine terraces of Florida. The elevationsof the terraces are Silver Bluff, cl-3 m;Pamlico, 2-8 m; Wicomico, 8-31 m; Coharie etc.,31-66 m; Pliocene, 66-98 m (adapted from Healy1975).1 imestone accelerated, replacingaboveground fl ows, promoting even moresolution, and becoming the Floridanaquifer, one of the most elaborateground-water storage and drainage systemsin the world (Stringfield and LeGrand 1966).Low stages of sea level in thePl ei stocene were accompanied by declinesin the ground-water table(Watts 1971). Because solution oflimestone proceeds mainly at and justbelow the water table (Stringfield andLe Grand 1966), lowered water tablesaided growth of the subsurfacedrainage system. Depressed water tablesa1 so meant drier surface terrain.Remains of the mammal assemblage livingin the early post-glacial periodof the Holocene (about 8,000 BP) suggestthat xeric, open woodland andgrassland were prevalent (Martin andWebb 1974). Pollen in sediments ofFlorida and Georgia 1 akes corroboratethe predominance of oak and prairievegetation at that time of lowered sealevel. As the sea, and presumably thewater table, rose through theHolocene, pines rep1 aced oaks and mod-

ern vegetation associations formed--pine fl atwoods, hammocks, cypressswamps, and bayheads (Watts 1980:Watts and Stuiver 1980).Hydric-hammock soils formed mainlyin sandy and loamy marine sedimentsover soft and hard limestone. A thinveneer of recent and residual weatheredsands covers much of the Tertiarylimestone in the gulf coastal hammocks.Thompson (1980) measured thedepth of the limestone bedrock in sixhydric-hammock stands at the northernend of the region; the extent of shallowlimestone (less than 90 cm beneaththe surface) ranged from 8% to 66% ofthe area. Outcrops of limestone arecommon in the gulf coastal hammocks(Figure 7). In contrast, extensivehydric hammocks grow on deep sandalong the upper St. Johns River (U.S.Soil Conservation Service 1960).There a1 kal ine materials, such asmarl, lie at least 107 cm below thesurface. However, the flow of thisstretch of the St. Johns River isgreatly influenced by the discharge ofhigh1 y mineral ized and sal ine water(residual sea water trapped in a deepaquifer) from upstream artesiansprings (Lichtler et a1. 1968). Highconcentrations of calci um and othersalts derived from the solution of1 imestone may become avail able to thehydric hammocks during high-water periods.Other hydric-hammock soils(e.g., the Tuscawilla and Parkwood seriesin the Atlantic coastal lowlands)contain shell and limestone fragmentsrather than solid limestone substratum(U.S. Soil Conservation Service 1974;1980) .The soils associated with hydrichammockvegetation are nearly 1 eve1and somewhat poorly to poorly drained(U.S. Soil Conservation Service 1981).Figure 7. Limestone outcrop in a gulf coastal hammock.

Sandy surface layers and sandy loamsubsoils arc com~on (Table 2). Theweakly cemented Bh horizon (hardpan)typical of pine flatwoods is absent.Clayey soi 1 s support hydric-hammockvegetation in some areas. Near theSilver Springs run in north-centralFlorida, hydric hammocks grow on phosphaticclayey soils (loamy fine sands)that contain 1 imestone nodules andshells. Known locally as "gumbo," thesoils belong to the Eureka and Paisleyseries and are poorly drained (U.S.Soil Conservation Service 1979).Shallow limestone bedrock in the gulfcoastal hammocks of Levy County, centralFlorida, is capped by severalinches to several feet of sandy loamor sandy clay loam (J. D. Slabaugh,U.S. Department of Agriculture SoilConservation Service, Bronson, Fl a. ;pers. comm.). In Waccasassa finesandy loam (Table 2), permeabil ity ismoderately slow t.o slow. The loamycap is, in turn, covered by a mantleof sand that gradually thickens withincreasing distance from the tidalmarshes. Due to the irregular configurationof the bedrock, the sandy mantlealso varies in thickness overshort distances. Slash pine, Pinuse]]iottii, grows On soils with thesandy mantle; conversely, hydric-hammockveyeial iurr w;~:I ve1-y ;CW VT ftOslash pines occupies the soils lackingthe mantle.In contrast to the highly acid soilsof pine flatwoods and bayheads, hydric-hammocksoils generally areslightly acid to mildly alkaline(Tables 2 and 3). Five hydric hammockstands at the northern reach of thegulf coastal hammocks had soil pH valuesbetween 5.7 dtid 5.3, end 5 ixthhad a value of 5.1 (Thompson 1980).The clayey soi 1 s that support hydrichammocks near Silver Springs are moderatelyacid (U.S. Soil ConservationService 1979). Organic matter contentof hydric-hammock soils is variablebut often low (Table 2), particularlyin relation to the amounts in bayheadsoils. More aptly termed "peat," bayheadsoil in northwestern and centralFlorida contains l5%-98% organic matter(Wharton et al. 1977; Clewell eta 1 . 1982). Thompson (1980) reportedorganic matter concentrations of 18%-54% in surficial samples of hydrichammocksoils, but the results probablywere biased by the inclusion ofTable 2. Characteristics of some sol1 series associated with hydric-hammock vegetation (from U.S. SoilConservation Service 1974; 1977; 1980; data for the Waccasassa series are from J. D. Slabaugh, pers.comrn.).Depth Texture pH Organic Cations (Meq./100 g)(cm) matter (%)Soil series % sand % silt % clay Ca Mg K NaAripeka 0-8 91 7 2 6.2 1.6 3.1 1.6 0.1 0.033-38 70 6 24 7.7 1 .O 7.7 1.5 0.2 0.8Pa rkwood 0-18 87 7 6 7.6 5.0 15.8 2.0 0.4 0.7fine sand 28-43 69 7 24 8.3 1.3Tuscawi 11 a 0-8 88 11 1 6.0 3.1 9.2 1.0 0.2 0.525-33 82 2 14 6.9 0.9 12.2 0.6 0.1 0.2

Table 3. Comparison of surface soil characteristics among three hardwood wetland types in Florida(data lor mixed hardwood swamps and bayheads from Monk 1966; data for hydric hammocksfrorn thisstudv. Table 2. and Thompson 1980).Pl ant community PH Cations (Meq./100 g)Mi xed hardwood swamp 3.8-6.8 1.6-134.9Bayhead 3.6-4.3 0.5-12.7Hydric hammock 5.1-7.6 6.4-18.9the surface litter layer. The Aucillahydric hammock occurs on 15 cm of organic-richsand (Wharton et a7. 1977).This wetland adjoins a spring run andshares some features with bayheads:near constant soil saturation and thepresence of trees such as sweetbays,Magnolia virginiana, that thrive onorganic soils. Such hydric hammocks,more constantly wetted than a1 ternatelyflooded and dried, probablyhave the highest content of soil organicmatter. Nutrient (cation) concentrationsare moderate in hydrichammocksoils (Table 2), falling atthe low end of the range reported forsoils in mixed hardwood swamps (Table3). Bayhead soils tend to have stilllower concentrations, especially ofcalcium and magnesium (Monk 1966).2.4 HYDROLOGYWater is the driving force in wet-1 ands, governing their unique characteristics and functions. Unfortunately,the hydrology of hydric hammocksis poorly known. Insufficientinformation is available on the natureand rates of hydrological inputs andoutputs, and the frequency and durationof flooding, to construct a waterbudget for a hydric hammock. The followingdiscussion is 1 argely conjectural,based on limited data and personalobservations.Sources of water to hydric hammocksare rainfall, river inundation, over-land flow and seepage from adjacentuplands, and discharge from deepaquifers. The re1 ative contributionsof these inputs vary among individualhammocks and with season. Wharton eta1. (1977) distinguished hydric hammocksfrom floodplain forests (a1 socall ed bottom1 and hardwood forests andriver swamps) by the source of water:the former are inundated by local water,while the latter are flooded byrivers. However, the demarcation isnot as clear-cut as the definitionsimply. For example, the major hydrologicalinputs to the hydric hammocksin the upper St. Johns River basinprobably are seepage from upland areasand rainfall. Yet these forests alsoare flooded occasionally by the riverduring periods of high water. In thelower basin, narrow strips of hydrichammock 1 ie between river swamp andscrubby pine fl atwoods communi ties(Figure 8). Though Laessle (1942) assertedthat the hammock soils werekept saturated by seepage from ups1 opesandy areas and were not subject toflooding, this description probably ismore appl i cab1 e to nearby bayheads(Figure 8); the hydric hammocks arecovered periodically by the St. JohnsRiver. Hydric hammocks also are foundon slightly elevated areas adjacent tomany spring runs, such as the Homosassa,Wekiva, and Aucill a Rivers,in northern and central Florida.Springs feed the rivers and maintain are1 atively constant, high water tablethrough a steady discharge of waterfrom the Floridan aquifer (Figure 9).

RIVER HYDRICSWAMP HAMMOCK SCRUBBY FLATWOODS BAYWEAD SCRUBBY FLATWOODSwater taMI_ surface of soilFigure 8. Sequence of plant communities from the lower St. Johns River to upland scrubbyflatwoods. In the higher flatwoods, water percolates to the hardpan and moves laterally until itemerges in the bayhead at the base of the slope. Ground water also seeps into the hydric hammock(from Laessle 1942).STREAM HYDRIC HAMMOCK MESlC HAMM--- APPROX MEAN WATER TABLEFigure 9. Generalized profile of a spring-fed stream and adjoining hydric hammock. The wavy linesindicate the annual fluctuation in river level. A high water table beneath the hydric hammock ismaintained by the spring run (adapted from Wharton et al. 1977).A1 ong the spring runs, hammock vegetationoccurs where the water table usuallyis below the surface and wherespring seepage is supplemented byrainfall and runoff from the uplands.These examples suggest that, whileconsiderable variation exists amonghydric hammocks, they differ from bothbottomland hardwood forests and bay-heads in their mixture of watersources. Bottom1 and hardwood forestsreceive most of their water from riverflow, and bayheads are characterizedby ground-water seepage..Within a hydric hammock, hydrologicalinputs vary with season.Rainfall is distributed very unevenly

throughout the year in Florida, andground-water 1 eve1 in hydri c hamnlocks. 1 T- ai.,I IULL.UUL-L-) ULLVI Y 11ly1Ji. A l l ~ 8 , ~Tjscj*Creek hammock in central Florida, thewater tabl e dropped, probably be1 OWthe root zone, during the period oflow rainfall in spring 1986 (Figure10). Ground-water 1 eve] reactedquickly to increased rainfall in summer,rising towards the land surface.Evapotranspiration must have removedmuch of the added water, yet the upwarddisplacement of the water tabl e($5 in Juncj :;as f;r 3723t~r th;nthe local input of rain (22 cm). Thehydric-hammock soi 1 s collected runoffand seepage from the uplands, and perhapsriver overflow, during the rainyseason. In droughty periods, up1 andseepage probably was the main sourceof water to the hammock.A bayhead situated in the TigerCreek watershed experienced far lessckrin~c -in water table depth with seasonthan the hydric hammock (Figure11). Preslrmably its major source ofwater, seepage, was more constant thanrainfall. Also, the peat soils typicalof bayheads hold more water than1 MlXtO HARDWOOD SWAMP4.1 1. ->-.\ ;\4(> i~pAn~---kA-2 ---- L1 4i)-L8041°C)AI €SO 200 240 780OISTANCF ALONG TRANSECT ( m)I I IF M A M J J A SMONTHFigure 10. Rainfall and ground-water level in1986 at Tiger Creek hydric hammock, centralFlorida. (Ground-water data are from RobertTighe, Center for Wetlands, University of Florida,Gainesville.)IIFigure 11. Water-table profiles across three plantcommunities adjacent to Tiger Creek, showingground-surface elevation (solid line) and maximum(closed circles) and minimum (open circles)heights of the water table during the period June1985 to September 1986 (February 1986 to September 1986 for hydric hammock). (Data are fromRobert Tighe, Center for Wetlands, University ofFlorida, Gainesvilie.)

soils with less organic matter, andthey maintain higher water levels dueto capillary action.Water-table depths in a mixed hardwoodswamp adjacent to Tiger Creekfluctuated slightly less than in thenearby hammock (Figure 11). The watertable sat at or above the surface ofmost of the swamp at the time of highestground-water level. In the hydrichammock, the water table remained belowgroundover the course of study.Although data are lacking, it is general1 y acknowl edged that hydri c hammocksflood less frequently, to ashallower depth, and for a shorter periodthan mixed hardwood swamps (e.g.,Brown and Starnes 1983).Most or all hydric. hammocks flood onoccasion (Figure 12), but depth, fre-quency, and duration vary from onehammock to another. Some oaks on thefringe uT Lilt: A v lildl~i!iiu~i(bear water lines about 1 m above theground. Extrapolation of high-waterlines into hydric hammock in the St.Johns River basin indicates that partof the forest floods annually to adepth of 20-30 cm (G. R. Best and P.Walt ace, Center for Wetlands, Universityof Florida, Gainesville; pers.comm.) .,ti; hYdr ic hallirllocks ffat~i-c a I-, igiiwater table seasonally or year-round.For two to six months during mostyears, the water table is very near(within 30 cm) or above the soil surfaceof the gulf coastal hammocks inLevy County (J. D. Slabaugh, pers.comm.) . Subsurface cl ay 1 ayers probablyhold the water table close to thesurface of the Aucilla hammock (C. H.Wharton, Institute of Ecology, University of Georg ia.; pers. cornm. ) . As describedpreviously (Figure 9), springruns maintain a high water table yearroundbeneath adjacent hydric hammock.Because the water table is high andthe land is flat, hydric-hammock soilsquickly become saturated during heavyrains, and excess water collects onthe surface. The frequency of inundationmay vary from once per decade,especi a1 ly where hurricanes induceflooding, to once or twice per year.Figure 12. Flooding of hydric hammock along theMyakka River, Sarasota County, Florida. Thisflood arrived as backwater from the river, threedays after rain.The rate of water loss from hydrichammock, and so the hydroperiod(duration of flooding) , is variable.The gulf coastal hammocks occasionallyare flooded by hurricanes and probablydrain within several weeks via smallcreeks. Due to their poorly drainedc1 ayey soi 1 s, hydric hammocks a1 ongthe Silver Springs run may remain inundatedfor two or three months followingrains in summer and winter.Sanchez Prairie, a small hydric hammockin north-central Florida, wascovered with up to a meter of waterfor several months in winter andspring 1986 (Figure 13); it also hadflooded the previous summer, but for ashorter time. The 1 onger hydroperiodin winter was due both to prolongedrains and to slow losses of water by

Figure 13. Flooding and drydown of Sanchez Prairie, a small hydric hammock in San FelascoHammock State Preserve, Alachua County, Florida. Common trees include live and swamp laureloaks, sweetgum, red maple, and loblolly pine.

evapotranspiration and sink-hol edrainage.Probably the major outputs of waterfrom hydri c hammocks are evapotranspirationand surface runoff.Ground-water recharge is unlikely tobe significant in most hydric hammocks;recharge is minimal or nonexistentwhere Tertiary limestone and,therefore, the Floridan aquifer, lieclose to or at the surface (Conover eta7. 1984). Surface water detained inhydric hammocks exits by sheet flowand small streams. Shall ow-lyingpermeable 1 imestone tends to restrictthe development of surface drainage,1 imi ting rivers in 1 imestone terrain,such as the Suwannee of northernFlorida, to few tributaries(Stringfield and Le Grand 1966). Thegulf coastal hammocks are transectedby small, poorly-defined streams anddrainageways. Discharge of hammockdrainage to receiving areas (e-g., theestuaries of the gulf coast) is probablyspread over time as well as area.The dense vegetation and flat topographyof hydric hammocks probably slowsthe movement of water and attenuatespeak storm flows.2.5 PHYSICAL CONTROL OF DISTRIBUTION OFHYDRIC HAMMOCKSA unique combination of physicalconditions is required for the existenceof hydric hammocks. It includesflat terrain, the influence of limestone,sandy or sandy clay soil, ahigh water table, and occasional-toseasonalinundation by a mixture ofrainfall, seepage, and, sometimes,river overflow. Southeastern statesbesides Florida 1 ack the configurationof physical features needed for thedevelopment of hydric hammocks. Theflat topography of much of peninsularFlorida is rep1 aced northward by morevaried and hilly terrain. Except inparts of South Carolina's coastalplain, limestone bedrock is buried farbeneath the surface (Stringfield andLe Grand 1966). Alluvial rivers originatein the mountains and Piedmontand carry high loads of sediment tothe coastal plain. The sediments aredeposited in extensive floodplains,providing substrate for bottomlandhardwood forests, the predominant typeof foresled w~t:~iLd {;;;&J ~ V I I el ai.1982).Hydric hammocks are scarce inFlorida's panhandle west of St. Marksfor the same reasons that they arerare outside of the State. Most ofFlorida's large river systems andfloodplains, and all of the alluvialrivers, are concentrated in that region(Wharton et al. 1977). Tributarycreeks often O~taFy deep, i3drrCWravines that dissect the hilly landscapeof the highlands (Clewell 1981).In the flatter coastal lowlands, limestonedips below the surface just westof St. Marks (Figure 5).The extent and distribution of hydrichammocks within peninsularFlorida are greatly influenced by topographyand physiography. The1 argest stands, the gulf coastal hammocks,inhabit ocean-smoothed terraceswith limestone and the water tableclose to the surface. Narrower bandsof hydric hammock commonly occur betweenmixed hardwood swamp and uplandforest in northern and central Florida(Figure 14a). Where the slope between1 owl and and up1 and increases, hydri chammock thins or disappears (Figure14b). On the west bank of the upperSt. Johns River, a swath of hydrichammock about 50 km long and averaging1-2 km wide abuts freshwater marsh(Figures 15, 16). In contrast, on theeast side cf the river, hydric hammockoccurs only in small patches amongmarsh and pine fl atwoods communities.The soil types are similar, if not thesame (U.S. Soil Conservation Service1960, 1974), and the elevations of thehammocks with respect to the St. JohnsRiver suggest that the frequency ofriver flooding is alike. One differencethat may account for the more extensivehydric-hammock forest on thewest bank of the river is the largenumber of tributary streams on thatside (Figures 15, 16). We speculatethat the streams, and their riverswamps, promote the establ i shment ofhydric hammocks via their effects on1 ocal hydro1 ogy and the containment of

+southari:magnolla+-------cabbage palmFlortda elmtransect length (m)-plw hickorysouthern rnagnolla+-----'Ilaurel oak+------Atransect length (m)Figure 14. Two vegetation transects from riverswamp to upland forest in the Oklawaha Riverbasin, central Florida. Solid bars indicate theapproximate width of hydric hammock on a gentlerise (a) and a steeper slope (b). The distributionsof various tree species are shown as bracketedlines; asterisks represent single individuals (forscientific names, refer to Table 4; adapled fromFlorida Game and Fresh Water Fish Commission1976).fire. Close study of the local distributionof hydric hammocks wouldcontribute greatly to our understandingof the re1 ationship between physi -ca1 factors and the development ofthis community.Figure 15. Topographic map of the upper St.Johns River floodplain between Puzzle Lake andLake Poinsett. The distribution of hydric hammockon the west side of the river is shown.Scattered hydric hammock stands are situated atabout 5-m elevation on the east side of the river,but they are too small to delineate on this map.2.6 SUMMARYAbiotic characteristics differ fromone hammock to another, and some areshared by other plant communities, buttheir configuration distinguishes hydrichammocks. All hydric hammocksreside on flat terrain, the most extensivestands inhabiting recently exposed,ocean-smoothed terraces. Hydric-hammocksoils generally aresandy, low to moderate in organic mattercontent, and slightly alkaline tomildly acidic; they lack the alluvialsediment on which bottomland hardwoodsoccur. The influence of limestone isa cbiqui tous feature. Limestone

Figure 16. Aerial photograph of hydric hammock on the west bank of the St. Johns River. View isfrom Puzzle Lake to the west, towards the mouth of the Econlockhatchee River (photo by G. KennethScudder).bedrock 1 ies close to or at the surfaceof many hammocks; in others, calciumis provided by flooding or seepingwater or by shells and limestonefragments in the soil.Another characteristic of hydrichammocks, the high water table, isproduced in a variety of ways. Exposuresof Tertiary limestone in thegulf coastal hammocks bring the Fl ori -dan aquifer close to the surface. Hydrichammocks adjacent to spring runshave high water tables due to dischargesfrom deep aquifers. In somehammocks, subsurface clay layers confineseepage and percolating rainfallto shallower strata.Hydric hammocks receive water fromrainfall, river overflow, and seepageand runoff from uplands. The mixtureof water sources varies among hammocks,but none is dominated by seep-age or riverine floodwater characteristicsof bayheads and bottomlandhardwood forests, respectively. Atone end of the spectrum, some hydrichammocks are strongly influenced byseepage. The soils are often saturatedand relatively high in organicmatter. However, ground-water levelsin these hammocks are not as constantas in bayheads, and may drop below theroot zone for brief periods. In contrast,some hydric hammocks probablyexperience extremes of flooding anddrydown, the former occasioned by periodicriver overflow. Most hydrichammocks fa1 1 between these two types.-Rainfall is the main source of water,but seepage and riverine floodwatera1 so contribute.Rainfall (acting directly or viaover1 and or river flows) occasionallyraises the water table above the surfaceof hydric hammocks. Floodingw

frequency may be once per year in flooding . Hydroperi od a1 so variessouthern Florida, timed with the rainyaniong hammocks, depending on the rateper iuci irl >ullililer . Kydr i~ ~I~IIIIIIUL~~ iit v-f tlkapott diisl) i t'at ;oil, the extent ofnorthern Florida may flood twice per drai nageway s, and soi 1 type. However,year, in slimmer and in winter. Some both hydroperiod and flooding frehydrichammocks are inundated less of- quency in hydric hammocks are lessten, perhaps as infrequently as oncethan in mixed hardwood swamps.per decade, when hurricanes induce

CHAPTER 3. VEGETATION OF HYDRIC HAMMOCKSSet amid the marshes, prairies, andpine flatwoods of Florida, hammocksare conspicuous in their height anddensity of vegetation. A closedcanopy and, often, the dominance ofbroad-1 eaved evergreen species (Harper1915; Laessle 1942; Monk 1965; Greller1980) create heavy shade within astand. The lush, jungle-1 i ke appearanceof some hydric hammocks (Figure17) t;e::c; 5 !1~152?y tc~pcra:eflora. Many species, including sweetgum,live and swamp laurel oaks, redmap1 e, and sugarberry (Celtis laevigata),range far to the north as wellas throughout Florida.Virtually all of the plant speciescommon to hydric hammocks (Table 4)are included in other communities.Only pink-root (Spigel ia loganioides),an herb, is known to bc en3ccicFigure 17. A large sweetgum in dense hydric hammock along Tiger Creek, Florida. Rows of holesOn the trunk are from repeated feeding by overwintering yellow-bellied sapsuckers.20

Table 4. Plants occurring in hydric hammocksa. Abundance classes are abundant (A), common (C),occasional (O), and rare (R).Scientific name Common name AbundanceWoody Pl ants:Acer barbatumAcer negundoAcer rubrumAescu7us paviaAmpel opsis arboreaBaccharis halimifoliaBerchemia scandensBignonia capreol ataBumelia reclinataCa7 1 icarpa americanaCamps is radicansCarpinus carol inianaCarya aquat icaCarya glabraCe7tis laevigataCercis canadensisCornus foeminaCrataegus viridisCrataegus marshalliiDecumaria barbaraDiospyros virginianaEugenia axil larisForest iera 7 igustrinaFraxinus carol inianaFraxinus paucif loraFraxinus pennsylvanicaGeisemium se~npervl rensGleditsia aquaticaGordonia lasianthusHypericum hypericoidesIlex cassine17ex coriacea17ex deciduaIlex glabra17ex opacaIlex vomitoriaI1 7 icium parviflorumItea virginicaJuniperus si 7 icico7aLiquidambar styracifluaLiriodendron tulipiferaLyonia lucidaMagnolia virginianaMorus rubraMyrica ceriferaNyssa sylvatica var. bif lordParthenocissus quinquefo7iaPersea palustrisFlorida map1 ebox-elderred maplered buckeyepepper vinegroundselrattan vinecross vinebuckthornbeautyberrytrumpet creeperhornbeamwater hickorypignut hickorysugarberryredbudswamp dogwoodgreen hawparsley hawcl imbing hydrangeapersimmonwhite stopperprivetpop ashswamp ashgreen ashyellow jessaminewater-1 ocustloblolly-baySt. Andrew's-crossdahoonbig gal 1 berrypossum-hawgal 1 berryAmerican hol lyyauponyellow aniseVirginia-willowsouthern red-cedarsweetgumtulip treefetterbushsweetbayred mu1 berrywax-myrtl eswamp tupeloVirginia creeperswampbay(Continued)

Table 4. (Continued).Scientific name Common name AbundancePinus elliottiiPinus serot inaPinus taedaQuercus laurifol ia bQuercus michauxiiQuercus nigraQuercus shumard i iQuercus virginianaRhaphidophyllum hystrixRubus argutusSaba 7 palmettoSabal minorSageretia minutif loraSambucus canadensisSebast iana fruticosaSerenoa repensSmilax spp.Tilia carolinianaToxicodendron radicansUlmus alataUlmus americana var. f loridanaUlmus crassif07 iaVaccinium elliottiiVaccinium fuscatumViburnum dentatum var. scabrellumViburnum obovatumVitis aestivalisVitis rotundifol iaslash pinepond pineloblolly pineswamp 1 aurel oakswamp chestnut oakwater oakshumard oaklive oakneed1 e palmhighbush blackberrycabbage palmbluestem palmettoclimbing buckthornelderberrysebasti an-bushsaw palmettogreenbriarbasswoodpoison ivywinged elmFlorida elmcedar elmmay berryswamp blueberrysouthern arrow-woodwal ter viburnumsummer grapebull ace grapeHerbaceous Plants:Arisaema triphyllumArnoglossum divers if07 iumArundinaria gigantea'Aster spp.Azo77a caro7inianaBoehmeria cylindricaBotrychium spp.Caca 1 ia suaveo 1 ensCarex spp.Chasmanthium spp.Cirsium spp.Cladium jamaicenseClematis crispaConyza canadensisCyperus sp.Desmodium spp.Dicondra caroliniensisDryopteris ludovicianaElephant opus nudatusjack-in-the-pulpiti ndian-pl antainswitch caneastermosquito fernbog hempgrape ferni ndian-pl antainsedgesspi kegrassesthistlessawgrass1 eather-flowerhorseweedf 1 at sedgebeggarweedpony-footFlorida shield fernpurple elephant' s-foot(Con t i nued)

Table 4. (Continued).Scientific nameElytraria carol inensisEpidendrum conopseumErechtites hieracifoliaEupatorium capi7lifo7iumEupatorium jacundumGalactia spp.Ga7ium spp.Hydrocotyle spp.Hypoxis leptocarpaHyptis alataImperata sp.Juncus spp.Leersia hexandraLemna spp.Lorinseria areolataMelothria pendulaMikania scandensMitchella repensMuhl enbergia schreberiOplismenus setariusOsmunda cinnamomeaPanicum commutatumPanicum rigidulumPanicum spp.Paspalum f loridanumPaspalum spp.Ph 1 ebod ium aureumPhyllanthus liebmannianusPo lygonum hydropiperoidesPolypodium po7ypodioidesPonthieva racemosaPsychotria undataRhynchospora spp.Ruellia caroliniensisSalvia lyrataSalvinia rotundif07 iaSanicula canadensisScleria triglomerataSenecio glabellusSpigelia loganioidesSpiranthes longilabrisSpirodela spp.Stenotaphrum secundatumSisyrinchium atlanticumThe7ypteris spp.Tillandsia bartramiiTillandsia recurvataTillandsia setaceaTillandsia usneoidesTrichostema dichotomum(Continued)Common namescal e-stemgreen-fly orchidf i reweeddog-fennelagerati namilk peabedstrawpenny-wortswamp (ye1 1 ow) star-grassmusky mintcogon grassrushsouthern cut grassduckweedchain ferncreeping-cucumbercl imbing hempweedpartridge berrynimbl eweedwoods grasscinnamon fernvari abl e pani cumred-top panicumpanic grassFlorida paspal umpaspal umgo1 dfoot fernpine-wood daintiesmild water-pepperresurrection fernshadow-wi tchwild coffeebeak rushwild petunialyre-leaf sagewater spanglessnakeroottall nut-grassbutterweedpink-rootlong-1 ip ladies'-tressesduckweedSt. Augustine grassbl ue-eyed-grasswood fernneedl e-leaf ai rpl antball mossneedle-leaf ai rpl antSpanish mossblue curlsAbundance

Table 4. (Concluded).Scientific name Common name AbundanceUrena lobataVerbesina virginicaVernonia spp.Viola affinisVittaria lineataWoodwardia virginicacaesar weedfrostweedironweedFlorida violetshoestring fernchain ferna The species list was derived by R.W. Simons from numerous field trips andconsultations with D.W. Hall (Herbarium, University of Florida), D.B. Ward(Department of Botany, University of Florida) , W. S. Judd (Department ofBotany, University of F1 orida) , R.K. Godfrey (Department of Bi 01 ogi cal Sci -ences, Florida State University), D.K. Younker (Florida Department of NaturalResources), and others; from a review of site surveys done for theFlorida Natural Areas Inventory; and from literature (Nash 1895; Harper1914, 1915; Laessle 1942; Florida Game and Fresh Water Commission 1976; Simonsand Hintermister 1984; Simons et al. 1984).Historically, most authors treated laurel oaks as a single species but differedon the appropriate scientific name; the current consensus is to treatthe wetland form as swamp laurel (or diamond-leaf) oak, Q. laurifolia, andthe upland form as laurel oak, Q. hemisphaerica. Wool fenden (1967) identifiedthis oak as Quercus hemisphaerica, laurel oak. Little (1979) consideredlaurel oak and swamp laurel oak to be the same species, Quercus 7aurifolia.Includes Arundinaria tecta, sometimes called river cane or bamboo.(Godfrey 1979). About one-ha1 f of thewoody plant species listed in Table 4a1 so are characteristic of southeasternbottomland hardwood forests(Wharton et a7. 1982). Harper (1915)surveyed the vegetation of a part ofcentral Florida and observed 19 abundanttree species in low (hydric) hammocks;of these, seven also were commonin high (mesic) hammock, six occurredfrequently in swamps, two wereabundant in both high hammock andswamp, and the rest had uncertainidentifications. The species 1 ists ofhydric hammock, mesic hammock, andmi xed hardwood swamp communitiesgreatly over1 ap, but the relativeabundances of the species differ(Florida Game and Fresh Water FishCommission 1976). Some pl ant species(e.g., Florida elm) are considerablymore common in hydric hammocks thanelsewhere. While the plant species ofhydric hammocks are not distinctive,their assemblage is.The canopy of hydric hammocks isabout 17-21 m high and is dominated byone or more oak species, cabbage palm,or a combination of these. Also commonare sweetgum, loblolly pine,Florida elm, and red maple. Canopyclosure is 75%-90%. The trees may be1 aced with vines--typically trumpetcreeper, pepper vine, poison ivy, andwild grape--and epiphytes may be abun-dant on the tree trunks and limbs.Young canopy trees, especi a1 ly cabbagepalm and sweetgum, may be frequent inthe understory, but in some hydrichammocks, hornbeam dominates. Bluestempalmetto, Smilax species, or amixture of shrubs and sap1 ings may occupythe shrub layer. However, cover

of the subcanopy and shrub layers isextremely variable; this vegetationoften is so sparse that vi si bi l i ty isgood at eye level. Likewise, theground layer may be absent or consistof a dense growth of ferns, sedges,grasses, and 5mi7ax species.Vertical structure and species compositionof the vegetation vary considerablyfrom one hydric hammock toanother. The objectives of this chapterare to document some of the differencesamong hydric hammock standsand to suggest factors responsible forthis variation. Because hydric hammockvegetation has been 1 i ttle studied,only hypotheses, based upon ourobservations and re1 evant data fromother communities, are presented. Wehope these conjectures will spur furtherresearch. Many piant species ofhydric hammocks are significant to animalsas well as to humans (Simons etal. 1988). To preserve and managethese resources, it is necessary tounderstand the factors responsible forstand composition.3.2 VEGETATION PATTERNSVery few quantitative descriptionsof hydric-hammock vegetation have beenpublished; these are supplemented byour surveys (Table 5) of stands locatedthroughout the range of hydrichammock in Florida (Figure 1). TheTable 5. Composition of hydric hammock stands (importance valuesa of trees 2 10 cm dbh).- - - - --si tesb--Species ESJ TOS MYR TC SEM WEK NGH OPH HR' SS SFH IGHCabbage palmLive oakSwamp laurel oakSweetqumFlorida elmLob1 01 1 y pineRed mapleHornbeamWater oakSweetbaySouthern red-cedarSugarberrySwamp tupeloGreen ashWinged elmSwamp bayBasswoodShumard oakWax-myrtl ePersimmonPignut hickoryWater-1 ocustRed mu1 berryBald cypressWater hickorySwamp ash80 75 72 56 52 50 43 27 20 267 13 23 6 33 13 7 30 32 2 7 6 14 10 8 11 1 22 223 18 11 tl 7 4 10 17 23 17t 1 6 12 5 2 6 2 1 52 8 1 38 5 46 14 4 10 52 4 tl 4 29 16tl tl 2 7 7 2 1< 1 10 5 1 1010 4 2 14 2 4 34 2 7 tl1 6 (1 66 1 1< 1 2 2 tl2 3< 1 1 33 1 < 11 (1 1t 1 1 22 tl 2< 1 < 1 26 t 132(Continued)

Table 5. (Concluded).~i tesbSpecies ESJ TOS MYR TC SEM WEK NGH OPH HR' SS SFH IGHLoblol ly-bayDahoonWhite ashSouthern magnol i aPrunus sp.Dev i 1 woodPop ashSlash pineSpruce pineGreen hawNumberofspecies 4 11 5 12 11 10 22 13 16 10 12 17Basal area(m2/ha) 54 71 67 57 29 61 46 42 97 42 49 43Trees in sample 263 473 224 234 113 69 1,979 113 425 273 197 140aThe importance value of a species is one-half the sum of the relative density(percentage of total tree density in the sample) and the relative basalarea (percentage total basal area in the sample). Sites were sampled duringthe present study except where otherwise noted. Ten-meter-wide belt transectswere used to sample the trees. The number and length of transectsvaried with the size and heterogeneity of the hydric hammock.Site locations are shown on Figure 1. Code for hydric hammocks:ESJ = east side of upper St. Johns River (Seminole Ranch Recreation Area);TOS = Tosohatchee State Reserve, west side of upper St. Johns River;MYR = Myakka River State Park (Shep's Island);TC = Tiger Creek Nature Preserve;SEM = Seminole County, northwest corner adjacent to the St. Johns River(data from G. R. Best and P. Wallace, Center For Wetlands, University ofFlorida, Gainesvil le) ;WEK = Wekiva Springs run, Wekiwa Springs State Park (the spelling of thepark differs from that of the spring and river);NGH = northern gulf coastal hammocks, St. Marks National Wildlife Refuge(Thompson 1980) ;OPH = Orange Lake palm hammock, northeastern shore of Orange Lake, AlachuaCounty;HR = Hi llsborough River State Park (Wool fenden 1967) ;SS = Silver Springs run (north side), Marion County;SFH = San Felasco Hammock State Preserve (Sanchez Prairie);IGH = inland reach of Gulf Hammock (near Otter Creek, Levy County).All trees, no matter what diameter, were recorded along a 1-m wide, 300-mlongtransect. Besides the species for which data on density and diameterwere recorded, six additional tree species were noted in the plot(Wool fenden 1967).

sites were not chosen randomly;rather, they were selected to encompassthe hydrological and soil conditionsassociated with this communityand to demonstrate the degree of variationin plant composition.Hydric hammocks range from nearlymonospecific stands of cabbage palm(Figure 18) to diverse hardwoodforests 1 acking this species (Figure19; Table 5). Tree species (woodyplants 210 cm in diameter at breastheight) numbered from 5 to 22 perstand in the present study. Differencesin sample size may have accountedfor some, but not all, of thevariation in species richness amongthe forests. Hydric hammocks lyingbetween the pine flatwoods and freshwatermarshes of the St. Johns (Figure20) and Myakka (Figure 18) Riverbasins were the least diverse. Notonly were few species present--mainlycabbage palm, 1 ive oak, and southernred-cedar--but they were very unevenlyrepresented. Hydric hammocks of theGulf Hammock region (Figure 191, exceptingstands immediately adjacent tosalt marsh, contained many morespecies, and the trees were apportionedmore evenly among species.Monk and McGinnis (1966) compared diversityof tree species of ten forestcommunity types in north centralFlorida, including five categories ofsouthern mi xed hardwood forest(hammock) del ineated by content ofsoil moisture and soil calcium. The"wet calcareous cl imax southern mixedhardwoods" type was judged to be themost diverse. Locations of the standsused in the analysis were not given;however, the wet calcareous hardwoodstands probably were hydri c hammockssimilar to the diverse forests we sampledin the inland Gulf Hammock regionand in San Felasco Hammock State Preserve(Table 5).Figure 18. Cabbage palm predominates in hydric hammock along Upper Myakka Lake, Myakka RiverState Park.

Figure 19. Hydric hammocksat the inland edge of Gulf Hammock feature a diverse mixture of swamplaurel oak, sweetgum, Florida elm, hornbeam, loblolly pine, sweetbay, and red maple. Cabbage palmis absent.Figure 20. Hydric hammock forest of cabbage palm, five oak, and southern red-cedar adjoinsfreshwater prairie along the upper St. Johns River.

When present in a hydric hammock,cabbage palm was nearly always the?redorninant tree (Tab1 e 5). Ouercusspecies, especi a1 1 y 1 i ve oak and swamplaurel oak, were always important, andsweetgum and Florida elm often wereabundant. Tree species that frequentlyappeared in moderate numbersincluded red map1 e, water oak, sweetbay,sugarberry, swamp tupelo, greenash, and swampbay. Some species were1 ocal ly abundant, but absent or scarceel sewhere. For exampl e, southern redcedarshared the canopy with cabbagepalm and live oak in hydric hammocksalong the upper St. Johns River and inthe coastal reaches of the gulfcoastal hammocks (Figure 21); mostother hydric hammocks lacked thisspecies (Table 5). Loblolly pine wasfar more common in some hydric hammocksthan in others and was the dominantspecies in hydric forest nearSilver Springs run (Table 5; Figure22). There, mature pines averaged 25-35 m in height and towered over a subcanoDvof typical hvdri c-hammockspecies: cabbage palm, live oak,sweetgum, and water oak (Figure 23).This association of tree species wasdistinguished from hydric hammock andtermed "short-1 eaf pine and cabbagepalmetto bottoms" by Harper (1915) and"loblolly pine hammock" by the FloridaGame and Fresh Water Fish Commission(1976). Instead, we consider the SilverSprings forest to be a variant ofhydric hammock that contains an unusuallyhigh abundance of loblol ly pine.Loblol ly pine-dominated hammock isvery extensive in the Silver Springsregion, and it also occurs in smallpatches within many large hydric hammocks.Some tree species are characteristic, although infrequent, members ofFigure 21. Coastal hydric hammock where it adjoins salt marsh, Gulf Hammock, Levy County. Thecanopy consists entirely of cabbage palm, southern red-cedar, and live oak.

Figure 22. Loblolly pine hydric hammock, Silver Springs, Marion County.Figure 23. Vertical structure of hydric hammock dominated by loblolly pine.30

hydric harriniock forests. These speciesi ncl i~de persimmon and red mill berry,-..A +"-:- --,-2~.t :br:;:Cc c~~,,, Cl\,.,nU,," >>,b 1 ,3 . ",,, d">

(a)swamplaureloakSweet-Florodared horn -maple beam bald shrubsWFigi~re 25. Species composition of common trees (>I0 cm dbh; hatched bars) and shrubs andsaplings (2.5-10 dbh; open bars) in two hydric hammocks: (a) Seminole County (data from G. R. Bestand P. Wallace, Center for Wetlands, University of Florida); (b) northern gulf coastal hammock(Thompson 1980). Importance values (defined in Table 5) were calculated for each size-class withina hammock. Species with importance values of 5 or less in both tree and sapling classes wereomitted.Tree saplings dominated the 2.5-10cm dbh :izc class in hydric hammocksof Seminole County and the northerngulf coast (Figure 25); shrubs accountedfor less than 20% of the basalarea. With the exception of cabbagepalm, a1 1 tree species were representedin the smaller-sized category.The converse was not true: red maplesapl ings were abundant in the SeminoleCounty hammock, but mature trees ofthat species were absent. In the sameforest, swamp 1 accrel oak, sweetgum,and loblolly pine had higher importancevalues as sapl ings than astrees. Apparently the canopy containedgaps that promoted colonizationof these shade-intolerant species(Putnam et a7. 1960).Hornbeam and swampbay were more importantas "sapl ings" than as "trees,"because they attain maturity at asmall size (Figure 25). Hornbeam wasan abundant member of the subcanopy ofthe Seminole County hammock, the OrangeLakc palm hammock, and the inlandreach of Gulf Hammock, and it musthave dominated this stratum in theHi1 1 sborough River hydric hammock(Table 5). However, this species wasscarce in the northern gulf coastalhammock (Thompson 1980) and non-existentin many of the hydric hammocksthat we visited, including SilverSprings, Wekiva Springs, Myakka River,and Sanchez Prairie (in San FelascoHammock State Preserve). Hornbeam'sabsence from the last forest is puzzling,since the species is common insurrounding mesic hammock.Yaupon, wax-myrtle, and dahoon werethe most common shrubs on the northerngulf coast (Thompson 1980), whereasswamp dogwood and wax-myrtle were theonly shrubs sampled in the SeminoleCounty hydric hammock (G. R. Best andP. Wall ace, pers. comm. ). Wax-myrtle

was the most frequent shrub in the hydrichanimccks that we surveyed. Fewsnruo species uLLur red r eyui dr ;y, dr~dseveral were restr-~cted to one or afew hammocks. Blue-stem palmetto appearedin several stands within theSi 1 ver Springs hydric hammock. Needlepalm was present only at WekivaSprings and Tiger Creek. The drierparts of the Sanchez Prairie hydrichammock featured blueberry bushes(Vaccinium spp. ) , but the wetter areashad no shrubs. Among the hydric hamiiid~khwe ;l is i Led, biil u?. veyetatf \u':~densest at Wekiva Springs.tiydric hammocks with a high abundanceof cabbage palm, such as Tosohatcheeand Myakka River (Table 5),generally had few plants in the shrublayer, except for very young palms,and no ground vegetation (Figure 18).The extent of the ground layer was extremelyvariable both among and withinthe reinainiiig fcr-ests. In the SilverSprings hydric hammock, dense patchesof spi kegrasses (Chasmanthium spp.)inhabited clearings. Vegetation,mainly grasses and ferns, covered from74% to 96% of the ground in six standsat the northern end of the Gulf Hammockregion (Thompson 1980). Thepl ants were present in sl ightl y openareas and absent from low, wet places,suggesting that they were 1 imi ted bylow 1 ight and extended f1 ooding.Herbaceous vegetation was most lushand diverse in the Tiger Creek, SilverSprings, and in1 and Gulf Hammockforests. In those hydric hammocks, aprofusion of ferns, grasses, sedges,and herbs covered the ground (Figure19).The largest contiguous stand of hydrichammock, known locally as GulfHammock, comprises the section of gulfcoastal hammock between the Suwanneeand With1 acoochee Rivers. Originallycovering more than 40,000 ha, GulfHammock has been reduced by clearingfor pine plantations and agriculture(Simons et dl. 1988). Gulf Hammock ishighly diverse, encompassing most ofthe variation in structure and composi tion found among hydric hammocks.Few hammocks have such a complete setaf the more common plant species, norof the rare ones, such as cedar elm,American plum, pink-root, pine-wooddaintiss, hiid the two indian plantainsCaca7 i a suaveolens and Arnog7ossum diversifolium.The interface of salt marsh and GulfHammock forest (Figure 26) is very irregular.Many islands of hammock arefound in the marsh, and the marsh extendsinto the forest along tidalcreeks, sometimes for considerabledistances. In most places along theboundary, the forest begins abruptlywith a dense, 12- 15-meter-tall standof cabbage palm, live oak, southernred-cedar, and occasionally water locust.Loblolly pine sometimes occursat the forest edge, and in a few areasof coastal Gulf Hammock, it dominatesthe forest, often in association witha dense ground cover of St. AugustineFigure 26. Edge of hydric hammock and saltmarsh in Gulf Hammock, along the Gulf of Mexiconear the Withlacoochee River, Citrus County,Florida.

grass. Sal t-tolerant shrubs such as1r.a frotescens (marsh elder), Lycrrtmcarol inianum (Chri stmasberry), andBaccharis angust if07 ia, B.g70merulifolia, and B. ha7imifolia(saltbushes) are abundant at the forestedge in some locations. Vines arerare at the edge, and both vines andshrubs are scarce in the forest interioruntil at least 1 km inland. Theground cover is very sparse. Thiscoastal part of Gulf Hammock is quitesimil ar to the hydric hammocks borderingthe marshes along the St. Johnsand Myakka Rivers.With increasing distance from thesalt marsh, cabbage palm, live oak,and southern red-cedar decl ine andhardwoods increase in dominance. Beginningat about 2 km inland, GulfHammock is divisible into three mainvegetation types. Swamps of bald cypress,red maple, swamp tupelo, andgreen ash occur in low areas and alongthe poorly defined creek drainages.Forests that might be considered eithermesic or hydric hammock occupysl ightly elevated ridges. The overstorycontains swamp chestnut, Shumard,live, laurel, and water oaks,plus sweetgum, southern magnol i a, sugarberry,winged elm, Florida elm,Florida maple, lob101 ly pine, southernred-cedar, pignut hickory, persimmon,red mulberry, and basswood. The understoryand ground cover also are diverse.The third type of vegetation,the majority of Gulf Hammock, is betweenthese two "extremes" in speciescomposition and is clearly hydric hammock.This major hydric hammock (Figure19) consists mainly of swamp laureloak and sweetgum in combination withlive oak, water oak, loblolly pine,Florida elm, basswood, persimmon, redmaple, sweetbay, sugarberry, and cabbagepalm. The average height of thedominant canopy trees is about 30 m,with scattered lob1 01 ly pines emerging3-6 m above the canopy. The scatteredlive oaks are by far the largesttrees, averaging about twice the trunkdiameter and crown spread of the othercanopy tree species. The ground atthe bases of these big live oaks oftenis raised by the root system to form amound. This microtoaoara~hv is particularlywell suited for the estab-1 ishment of magnolias, and, in conjunctionwith the spreading crown ofthe live oak, for the growth of severalspecies of vines (bullace andsummer grape, pepper vine, rattanvine, and climbing buckthorn). A subcanopydominated by hornbeam, wax-myrtle,swamp dogwood, and various treesap1 ings often is we1 1 developed. Theground layer commonly is a dense mixtureof grasses, sedges, and ferns,but in wet areas the cover is mostlyleaf 1 itter with only a scattering ofherbaceous pl ants.Patches of hydric hammock dominatedby loblolly pine are found within GulfHammock. Some of these are naturaland others are the result of human activity.Some of the higher ridgeswere cleared for farmjng 1 ong ago, apparentlyby German immigrants, andthen abandoned. Now stands of largeloblolly pines, locally known as"German Islands", cover the oldfields. A modern activity with similarresults is the clearcutting of extensiveareas of Gulf Hammock foll owedby the planting of loblolly pine.Gulf Hammock and its surroundingsare very flat, with one exception.Old sand dunes .covered with sand pinescrub vegetation are found on thenorth side, just inland from CedarKey. A mixture of swamp and hydrichammock occurs adjacent to the dunes,but both of these types are somewhatdifferent in composition than elsewherein Gulf Hammock, presumably dueto the continuous supply of waterseeping out of the dunes and to thethicker layer of organic muck. Swamptupelo, the dominant tree or codominantwith green ash, is much moreabundant than in other swamps of theregion. The hydric hammock also isdifferent in its abundance of needlepalm, which is quite scarce elsewherein Gulf Hammock. This part of GulfHammock is most similar to the hydrichammocks at Wekiva Springs, MormonBranch Botanical Area in the Ocala NationalForest, and Tiger Creek.

Gulf Hammock is rep1 aced abruptly atits inland edqe by pine flatwoods.However, strips of hydric hammock extendinland along streams such as~ocky Creek, Otter Creek (inland GulfHammock site, Table 5), and the WaccasassaRiver, and these forests graduallybecome less diverse. Cedar elmand American plum do not occur withinthese inland strips. Other trees thatdisappear, roughly in order of disappearance,are red buckeye, water locust,Florida maple, Shumard oak,swamp chestnut oak, cabbage palm,wi nged elm, sugarberry, basswood, and,finally, loblolly pine. The hydrichammocks farthest i nl and a1 ong thestreams consist mainly of swamp 1 aureloak, sweetgum, red map1 e, sweetbay,and Florida elm; occasional speciesi ncl ude 1 i ve oak, persimmon, swampbay,dahoon, and slash pine. Wax-myrtleis an abundant shrub. This forestis most similar to other hydric hammocksfound adjacent to f 1 atwoods andto sandhill streams such as TigerCreek.3.3 SOURCES OF VARIATIONComplex interacting factors probablyi nfl uence the species composition ofhydric hammock and distinguish it fromother communi ties. We propose thatthe geographical location of a stand,the hydrological regime, edaphic conditions, and fire frequency and intensityare the major determinants of thestructure of hydric hammock. Disturbancesother than fire and flooding,both natural and human, also maygreatly a1 ter the forest's diversityand biomass.conditions, soil type, and fire. However,a number of species have restrictedgeographic distributions thatpreclude their presence in some hydrichammocks. Absence of such characteristicbut not ubiquitous species maycause doubt about the classificationof a stand of forest unless a complexmodel is kept in mind.Temperate species dominate the floraof hydric hammocks but have varioussouthward limits to their distributions(Figure 27). The ranges oftwo abundant species, sweetgum andloblolly pine, do not extend south ofcentral Florida. A number of lesscommon trees, including green ash,Shumard oak, swamp chestnut oak, cedarelm, and winged elm, are restricted tonorthern Florida. So too are some occasional shrubs and herbs: buckthorn,green haw, red buckeye, and switchcane. The depauperate flora of hydrichammock in Myakka River State Park(Table 5) may result, at least partly,from its location south of the rangesof many species common to hydric hammocks.A few species are found only in asmall number of hydric hammocks be-green ash ] \ Shumard oakI \Plant species whose geographicaldistributions completely include thatof hydric hammock may be characteristiccomponents of this communi tythroughout its range. Live oak, swamplaurel oak, red maple, and many otherspecies that occur in hydric hammocks(Tab1 e 4) range throughout Florida(Little 1978); the absence of any ofthese species from a particular standis due to factors such as hydrologicalFigure 27. Natural ranges in Florida of fourspecies of trees common to hydric hammock(from Little 1978). The range of green ash acfuallyextends southeastward at least to WekivaSprings, Orange County.

cause they are tropical plants confinedmostly to south Flnrida- Noneoccur in abundance and only one, whitestopper, is a tree. Isolated populationsof this species occur north ofLake Okeechobee (Little 1978); one ispresent in hydric hammock in TosohatcheeState Reserve. The geographicaldistributions of several herbaceousplants common to hydric hammocks,including goldfoot fern andwild coffee, extend north from theCaribbean to central Florida.A striking example of the influenceof a species' range on the compositionof hydric hammock is cabbage palm, thepredominant tree species in many hydric hammocks. From North Carol i na tonorthern Florida and west across thepanhandle, this species is essentiallyrestricted to the coast (Figure 28).South of a line from about Cedar Keyon the gulf coast to St. Augustine onthe Atlantic, cabbage palm rangesFigure 28. Distribution of cabbage palm inFlorida (from Little 1978). Asterisks on the insetmap indicate isolated populations of cabbagepalm outside of its continuous range (from Brown1973).across Florida' s peni nsul a. North ofthis line, a few interior populationsare scattered a1 ong rivers. Hydrichammocks in the northern region, forexample in San Felasco Hammock StatePreserve (Tab1 e 5), 1 ack cabbage palm.The inland Gulf Hammock site that wesampled (Table 5) is situated at theedge of cabbage palm's coastal range.Cabbage palm was absent from much ofthe forest, but it was common, especiallyin the understory, in someparts.Numerous studies have demonstratedor inferred that patterns of vegetationin southern forested wetlands arestrongly influenced by flooding(Bedinger 1978; Huffman and Forsythe1981; Wharton et a7. 1982; Leitman eta7. 1983). Frequency, duration,depth, and timing of flooding are factorsthat affect the species compositionof a forested wetland. The effectsof flooding on these communitiesare mediated by physical and chemicalchanges in the soils and varying responsesto the alterations by plantspecies.The nature and extent of the physicaland chemical processes that followsoil inundation largely depend on theduration of submergence and on soilproperties (Ponnamperuma 1984). Whena soil is flooded or saturated, gasexchange between the soil and air isgreatly restricted. The slowing ofgas diffusion is particularly great insoils with high clay content. Oxygensupply to the soil is drastically cut,and roots and microorganisms depletethe oxygen in the soil water veryrapidly. Gases produced by soi 1metabol ism (e.g., carbon dioxide) accumulate.Anaerobic conditions inducea number of chemical changes in thesoil, many of which are detrimental toplant growth. Nitrate is replaced byammonium, often less preferred for uptakeand assimilation by plantspecies. Oxidized forms of iron, manganese,and sulfur are reduced to potentiallytoxic forms, including sulfides. Ethanol, another potent i a1toxin, is a by-product of anaerobicrespiration in most plant roots.

A sequence of changes in plantmat a hnl i grn and ~hvsi 01 ogi ca9 orocessesfollow the onset of anaerobiosis inwater1 ogged soi 1 s (Kozl owski 1984a,b) . Decreased water adsorption andclosure of the stomata result in a4 slowed rate of photosynthesis. Rootpermeabi 1 i ty is reduced, affecting theImovement of ions including nutrients.Leaf chlorosis and abscission, and retardedplant growth (especiallyheight-growth) foll ow. Morphologicalchanqes, such as the formation ofaerenchyma tissue and the growth ofadventitious roots, may take place.If flooding is prolonged, the plantdies. Flood-to1 erant plants, characteristicof wetlands, possess a varietyof morphological and physiologicaladaptations (reviewed in Hook andCrawford 1978) that avoid or mitigateflooding stresses. These adaptationscommonly facil ite oxygen flux to theroots and enhance the ability of plantroots to respire anaerobical ly withoutharmful effects.Plant responses to flooding arestrongly infl uenced by the duration,timing, and depth of inundation, butthe condition of the floodwater alsois significant. Flowing water is betterto1 erated than standing water,presumably because of the higher concentrationsof dissolved oxygen in theformer. The effects of flood durationare clear from the preceding paragraph;changes in plant physiology becomeprogressively more severe withincreased period of anaerobiosi s. Theadverse effects of fl ooding are exacerbatedduring the growing season whenoxygen demands are greatest. On theother hand, flooding during the dormantseason has re1 atively 1 i ttl e impacton the physiology and survival ofmost tree species ,(Gill 1970). Depthof inundation is critical to plantsurvival because oxygen diffusion tothe roots is greatly slowed by passagethrough water. Water depth with respectto plant height also affectsplant response to flooding. Most Shumardoak and sweetgum seedlings survived60 days when flooded only to theroot collar (Hosner and Boyce 1962),but all died after 20 days of completesubmersion (Hosner 1960). Water deepenough to cover most of the plant de-creases 1 ight intensity and interfereswith stomata1 function.The ability to withstand floodingvaries among plant species, sometimesamong populations, and with plant ageand plant size (Whitlow and Harris1979). Most assessments of the relativetolerance of woody plants toflooding are based on the results ofexperimental inundation of seed1 ingsand reservoir flooding of establishedforests. Results of studies examiningthe responses of wetland hardwoodtrees were summarized by Gill (1970),Teskey and Hinckley (1977), Whitlowand Harris (1979), and McKnight et al.(1981). Some tree species common tohydri c hammocks were included. Caremust be taken in applying the resultsof these tests to interpretations ofnatural communities. Reservoir floodingis considerably less erratic thannatural flooding and, in nature, numerousfactors interact with inundationto influence vegetation patterns.Nevertheless, some conclusions fromstudies of flood tolerance help in explainingplant distributions withinand among hydric hammocks.McKnight et a1. (1981) assignedsouthern bottomland trees to four toleranceclasses that varied in two regards:length of time during thegrowing season that the species canwithstand flooded or saturated soils,and the extent of anatomical and physiologi cal adaptations. Only one commonhydri c-hammock species, swamp tupelo, was considered to be tolerant.This species can survive long periodsof inundation, and its seeds remainviable when submerged in water formonths. Most hydric-hammock species(red maple, sweetgum, Florida elm,loblolly pine, sweetbay, persimmon,green ash, swamp l aurel oak) are moderatelytolerant of flooding; somemorphol ogical or physiol ogical adaptationsto flooding may develop (Figure29), but they do not enable the treeto survive flooding for an entiregrowing season. Mature trees of sixhydric-hammock species remainedhealthy when flooded for 17%-37% ofthe growing season; red maple withstoodthe longest flooding period, and

were saturated (Teskey and tiinckleyI???).Figure 29. Buttressed roots of trees in hydrichammock, including Florida elm (In front) andswamp laurel oak (behind), provide stability inwet soils and perhaps help to aerate the rootsystem.water oak and loblolly pine the shortest(Teskey and Hinckley 1977).Weakly tolerant species include hornbeam,red mulberry, and several oaks--water, live, Shumard, and swamp chestnut.According to McKnight et al.(1981), these species are able to surviveonly short hydroperiods (a fewdays to a few weeks), and they possessno apparent adaptations to inundation.Seedlings of hydric hammock trees generallyare more sensitive to floodingthan mature trees, although, as notedabove, mortality rates depend on thedegree of submersion. In short-termexperiments, death usually occurredwithin 20-30 days when the seedlingswere compl etel y submersed and within0.5-3 months when they were partiallyflooded; no seed1 ings died when soilsThe damming of the Oklawaha River innorth-central Florida provided addi -tional evidence of the differences inflood tolerance among hydric-hammockspecies (Harms et a1 . 1980). Probablymost of the flooded forest was riverswamp dominated by deciduous hardwoods,but part was hydric hammock(Florida Game and Fresh Water FishLcirnmr ssion 1476 j . Tiir-ee yedt :, dfteidam construction, tree mortal ity wasstrongly correlated with water depth,with essentially 100% mortality at averagedepths of 1.2 m or more, decreasingas prevailing water level declined. However, mortal i ty rates significantlyvaried among tree species.Bald cypress, swamp tupelo, and cabbagepalm were the most tolerant offlooding. Cabbage palms were scatteredthroughout the forest and fewdied, irrespective of water depth.They accounted for most of the survivorsin the most deeply flooded sectionof the forest. Two species ofoaks, water and swamp laurel, wereleast tolerant of flooding, survivingonly at the shallowest depths (aboutthe same and 20 cm deeper than controlsites outside of the reservoir). Nodoubt water levels in the reservoirfluctuated less thdn in tile undi5-turbed floodplain and long periods ofcontinuous flooding occurred. However,data on the water elevation ofthe lake indicate that prolonged drydownsalso took place (Harms et al.1980).The placement of tree species withinhydric hammocks probably ref1 ects inlarge part their relative tolerancesto flooding. Species abundances a1 onga presumed flooding gradient inSanchez Prairie, San Fel asco HammockState Preserve, are shown in Figure30. At the time of sampling, shallowwater stood at the beginning of thetransect, in a stand dominated by baldcypress and red maple. The transectclimbed a very gentle slope into hydrichammock, ending where live andwater oaks dominated and the groundwas quite dry. Bald cypress was restrictedto the wetter (swamp) end of

I ~ U mapleswamp laurel oak 3:1100bald cypress a- swamp tupelo -,100wgreen ash m jo $sweetgum m rn 3:6IO~IOIIY pine 3: !i100live oak- dowater oakI0 40 80 120 160 200DISTANCE (m)Figure 30. Distribution of tree species along apresumed flooding gradient in Sanchez Prairie,San Felasco Hammock State Preserve. The transectbegins in swamp conditions (left) and endsin drier hydric hammock (right).the gradient, and live oak, water oak,and loblolly pine to the drier(hammock) part. Red maple and swamp1 aurel oak ranged almost throughoutthe swamp/hammock transect. Swamplaurel oak was replaced by 1 ive andwater oaks at the dry end of the gradient,their relative positions in accordancewith flooding to1 erances inferredfrom the studies cited above.Another study transected 200 m of aSemi no1 e County hydri c hammock (Table5), descending 80 cm to river swampdominated by bald cypress, red maple,and sweetgum (G. R. Best and P. Wallace,pers. comm.). The first half ofthe transect probably was slightlyabove average high water, whereas thesecond half was slightly below. Sugarberry'and hornbeam were found onlyin the first 50 m, while swamp laureloak extended to 150 m. Sweetgum andFlorida elm ranged throughout the twohammock sections and into the swamp.Cabbage palm, the dominant tree inthis hydric hammock (Table 5), alsowas ubiquitous, but its abundance declined greatly in the swamp. Redmaple occurred in swamp and hydrichammock, though only as saplings inthe latter. NO 1 ive oaks were sampled,although this species was ob-served in other, generally drier,parts of the hydric hammock.Some of the differences in compositionof plant species among hydrichammocks also result from variablepatterns of flooding and drydown. TheSeminole County hammock probablyfloods more often and for a longer periodthan many others, which was reflectedin the vegetation: Floridaelm was an important species, redmaple and swamp laurel oak sap1 ingswere abundant, no 1 ive oaks were sampled,and bald cypresses were present.Live oak was absent also from theWekiva Springs and Tiger Creek hydrichammocks, both of which receive seepageand are almost always moist. Inthe Wekiva Springs hammock, a high watertable is maintained by dischargefrom a deep aquifer (see Figure 9).The hydrology of the Tiger Creek hammockwas discussed in Chapter 2.4;thi s wet1 and receives up1 and seepagedepending on the season. The profusionof shrubs and ground vegetationin these hammocks probably resultsfrom the nearly constant saturation ofthe soils. The influence of seepageon plant species composition is especial1 y pronounced at Wekiva Springs.Numerous bay trees, needle palm, andcinnamon fern make this forest intermediatein composition between"typical" hydric hammock and bayhead,a community characterized by constantseepage. The re1 atively constantmoisture regime of the Wekiva Springsand Tiger Creek hammocks contrastswith the extremes of drought andflooding experienced by other hydrichammocks. We suggest that hydric hammockslow in species diversity anddominated by cabbage palm and live oakexi st where long, dry periods are interruptedby occasional episodes offlooding. Examples are found in theupper St. Johns and Myakka Riverbasins, where river overflow only sporadicallyfloods the hydric hammocks.In Myakka River State Park, hydrichammock has expanded into freshwatermarsh during the past 30 years (Figure31), at the same time as river levelshave declined due to upstream diversions(Robert Dye, Florida Departmentof Natural Resources; pers. comm.).Whereas the old part of the hydric

Upper Myakka Lake/-HAMMOCKINVASION0.5 kmFigure 31. Expansion of hydric hammock(Shep's Island) into freshwater marsh, MyakkaRiver State Park, 1957-1972.hammock is composed of cabbage palmand live oak, the invading vegetationconsists of swamp laurel oak, Floridaelm, and, closer to the marsh, waterlocust and pop ash (Figure 32). Severeflooding during the summer of1982 killed many swamp laurel oaks,but not 1 ive oaks. Apparently thelatter species can withstand occasionaldeep inundation but not prolongedsoi 1 saturation.3.3.3 Eda~hic ConditionsParticular plant species may be favoredin some hydric hammocks becauseof variations in soil type. Changesin soil texture, organic content, andpH can markedly affect drainage andnutrient avail abil i ty, and therefore,the growth and survival of variousspecies. For example, sweetbay generallyis associated with organic soils(Wharton et a7. 1976). This speciesis most abundant in hydric hammockssuch as the Wekiva Springs forest(Table 5) that are characterized byseepage of ground water and organicrichsoils. The abundance of lob101 lypine in the northern gulf coastal hammocksi s negatively correl ated wi thsoil pH (Thompson 1980). Among thehydric hammocks sampled in Table 5,loblolly pine was most important inthe Silver Springs forest; the clayeysoils in that region are more acidthan the fine sands common to most hydrichammocks. The abundance patternsof water oak, both within the northernGulf Hammock and among hydric hammocks,parallel those of loblollypine, suggesting that growth of bothspecies is favored by similar conditions.CABBAGE PALMLIVE OAKSWAMP LAURELOAKFLORIDA ELMWATER LOCUSTPOP ASHFigure 32. Profile of expanding hydric hammock, Myakka River State Park.

The salt concentration of hydrichammocksoi 1 s strongly influences communI ty SxruLLur e. i~drill!-mediately adjacent to salt marshes ofthe central gulf coast are composedalmost entirely of Cabbage palm,southern red-cedar, and live oak(Figure 21).termed coastalThese forests werehammocks by Swindell(1949), the Florida Game and Fresh WaterFish Commission (1976), andClewell (1981), but we consider themto be an extreme type of hydric hammockLi,ai, ;> > iiiipl i fie6 Ly o~ed;i >prayand floodwater. The dominance of cabbagepalm in association with 1 ive oakand red-cedar continues for about 2 kminland (Figure 33). As distance fromthe sal tmarsh boundary increases, theforest is enriched with additionalspecies. Loblolly pine is the firstaddition, sometimes occurring at theforest edge. Cedar elm occurs asscattered trees beginning perhaps 200-300 in froia the coast. At dbout 1 kn~,Florida elm, sugarberry, and Floridamaple become common in some places.Sweetgum, swamp laurel oak, and hornbeamare abundant at 2 km. Averagei~ ~I~I,~I~IUL~>maximum height of trees increases from12-14 m at the marsh edge to 18-20 mitridnri i.5 km. f hompson (193'3) notedthat the structure of hydric-hammockstands in the northern Gulf Hammockalso varied with distance from tidalmarsh. Cabbage palm and southern redcedardecreased i n re1 ative importanceand sweetgum increased, as stands werelocated farther from the marsh andsoil concentrati ons of solubl e saltsdecreased. Cabbage palm, 1 ive oak,and southern red-cedar are very salttoleititit{Kurr and Wagner 1957). Thelast species has been termed a calciphile(calcium-1 oving) because it a1 -most always is associated with 1 imestoneoutcrops or coastal (salty) regions.However, southern red-cedardoes not require salt; it may be restrictedto sa1 t-influenced hydrichammocks because it is excluded fromothers by less sal t-tolerant species.Absent from many hydric hammocks,southern red-cedar is most abundant atthe seaward edge of the gulf coastalhammocks and in hydric hammocks of theupper St. Johns River basin (Table 5).The species composition of hydric hammocksin these two areas is very similar:almost exclusively cabbage palm,live oak, and southern red-cedar. Thehydric-hammock soi 1 s adjacent to themarshes of the St. Johns River arebathed by occasional river overflowcontaining salts originating as upstreamartesian flow of fossil sea water(see Chapter 2.3).3h0 6i0 960 12b0 15b0 1800 2 100 2400DISTANCE FROM SALT MARSH (m)Figure 33. Change in tree-species compositionof a hydric hammock with increasing distancefrom its salt-marsh boundary. The point-quartermethod was used, at 50-rn intervals, to Sample thetrees along a transect in Gulf Hammock.Hammocks commonly are thought to befree of fire and, indeed, to owe theirexistence to protection from this disturbance(e.g. , We1 1 s 1942). Harper(1915) noted of hydric hammocks: "Thehumus probably seldom or never getsdry enough to burn, so that fire doesnot need to be reckoned with." He emphasizedtwo features of this communitythat tend to reduce fire frequencyand intensity: wet soils and(usually) the lack of a substantial1 i tter 1 ayer. However, we observedfire scars in virtually every hydrichammock we visited. Almost all of thecharred trunks were of cabbage palms.Fire mark5 are preserved on these

trees for many years because they donot slough off the trunks (iaessle1942). P!G 63~bt hydric ha:ccck: Szrnmuch less often than fire-adapted communiti es such as pine fl atwoods, butwe propose that fires are sufficientlyfrequent and intense in some stands toinfl uence plant composi tion (Figure34).The effects of fire on hydric-hammockvegetation are mediated by differencesin suscepti bi 1 i ty among theplants. Seedl ings and sap1 ing-sizedtrees are more 1 i kely to be killed byfire than larger trees, because thebark tends to be less thick and thecrowns can be reached by the flames.However, even large trees of some hydric-hammockspecies, especially hardwoods,can be injured by fire, makingthem subject to attacks by fungi andinsects (Fowell s 1965). Bark thick-ness varies among species, accountingfor some, but not all, of the differnncszin firp C I I C C P ibil ~ ~ i t,y (Tab1 P6). When bark thickness was held constant,fourteen species of southerntrees still varied in fire resistance,presumably due to differences in thestructure, composition, density, andmoisture content of the bark (Hare1965). Generally, conifers were moreresistant than hardwoods, and fieldobservations support these findings.Pine fl atwoods (and planted pine plantations) are maintained by regularfires that kill hardwood reproduction.Two severely burned cypress swamps experienced18% and 23% declines in theabundance of the dominant tree, pondcypress (Taxodium distichum var.nutans, a conifer), but 98% and 83%decreases in the hardwoods (mainlyswamp tupelo, sweetgum, and sweetbay)(Ewe1 and Mitsch 1978). Loblolly pineis among the most fire-tolerantspecies in hydric hammocks, and almostall the hardwoods are susceptible tofire. However, one conifer, southernred-cedar, is quite fire-suscepti ble(Putnam et a7. 1960).Table 6. Fire sensitivity in five tree species commonto hydric hammocks. A standardized flamewas applied to living bark and the mean time inseconds for cambium to reach a lethal temperatureof 60 "C was recorded (from Hare 1965).Bark thickness (cm)Species 0.5 0.8 1.0Sweetgum 25.6 48.6 101.8Red map1 e 29.0 56.8 117.6Water oak 30.2 61.0 136.0Figure 34. Fire in the cabbage palm edge Ofhydric hammock, Seminole Ranch, BrevardCounty; top, June 1977, bottom, May 1978 (photoat top by G. Kenneth Scudder).Sweetbay 30.8 67.0 152.0Loblolly pine 35.6 84.2 179.2

Cabbage palm is the most fire-toleranttree in hydric hammocks, survivingeven htlvrr r r; 1 ii. Aii f lii~fi~lbui-ii iiia cabbage pal m/l i ve oak/southern redcedarstand in Tosohatchee State Reservekilled all trees except cabbagepalms (Randall E. Hester, Florida Departmentof Natural Resources; pers.comm. ) . Harlow (1959) ascri bed nearlymonospecific stands of cabbage palm tofrequent fires (every 2-3 years).Fires favor cabbage palm, and theirproduction of flammable 1 i tter in turn~)-c;,;;i?tes f i I-e. Less frequent and intensefires probably favor 1 ive oak inaddition to cabbage palm. Though manyauthors (e.g., Putnam et a1. 1960;Fowell s 1965) have claimed that 1 iveoak is fire-suscepti ble, Laessle andMonk (1961) observed signs of fire ina1 1 eight of the 1 ive-oak-dominatedforests they examined in northeasternFlorida. Because coastal and in1 andstands featured similar vegetation andfire scars, Laessle and Mank (1961)concluded that salt spray was less importantthan occasional fire, coupledwith the tenacity and longevity oflive oak, in maintaining live oak forest.More frequent and intense firesresulted, in coastal areas, in a lowthicket of vegetation containing sawpalmetto and dwarfed forms of severaloak species (Laessle and Monk 1961).Some of the variation in speciescomposi ti on among hydri c hammocks(Table 5) also may result from difFerencesin fire frequency and intensity.The domination of loblolly pine in hydrichammocks a1 ong Si 1 ver Springs runprobably is favored by occasionallight fires as well as soil conditions.If fire (and cattle grazing)were prevented, the forest would convergetoward more "typical" hydri chammock with an abundance of cabbagepalm, oaks, and sweetgum (Florida Gameand Fresh Water Fish Commission 1976).Hydric hammocks highly dominated bycabbage palm, and some live oak, arefound in the Myakka and St. JohnsRiver basins (Table 5). In these areas,hammocks often are bounded bothup- and down-slope by communitiescharacterized by frequent fire--pineflaiwoods and freshwater marsh. Geographicallocation plays a role in thelow diversity of hydric hammock in theMyakka region (see section 3.31), butother factors also may be important.Figure 35 contrasts the tree speciescomposition of two hydric hammockswithin Myakka River State Park. The0 Shep's IslandDeer Prairie SloughswampFloridaFigure 35. Tree-species composition of two hydric hammocks in Myakka River State Park: Shep'sisland, adjacent to Upper Myakka Lake and Deer Prairie Slough, about 20 km to the east.43

first, Shep's Island (Table 51, adjacentto Upper Myakka Lake, consists ofcabbage palm and live oak and has I lr;-tle shrub and ground Vegetation. Thesecond hammock, located about 20 km tothe east in Deer Prairie Slough, containsa number of hardwoods absentfrom Shep's Island, as well as copiousshrubs and herbs. The hydric hanlmockin Deer Prairie Slough occupies aslightly elevated area that is completelysurrounded b~ mi xed hardwoodswamp. We suggest that this contrastin hammock vegetation i s due to d ii i F-ference in fire frequency and intensity.Presumably fires that sweep thevast marshes of the Myakka region alsoburn Shep's Island but are damped bythe swamp in Deer Prairie Slough andonly rarely reach the hammock there.It also is possible that the hydrologicalconditions of the two hammocksdiffer enough to account for some ofthe differences in species composi -tion. The swamp may modulate theflooding regime of the Deer PrairieSiough ild111111u~k i u ~ h t h ~ th: dcr:t'c::of flooding is less than in the Shep'sIsland hammock and drydowns are not assevere.The plant composition of the Tosohatcheehydric hammock (upper St.Johns River) varies greatly on a localscale (Figure 36). Portions of thehammock jmmedi ately adjaceqt to freshwatermarsh (Figure 36c, d) are theleast diverse. Fire scars are pervasive,signal 1 i ng more frequent burnsin these sites than in hammock portionsnext to a mixed hardwood swamp(Figure 36b) and a creek (Figure 36a).The low basal area of forest (d) andits short stature (the average heightof the canopy is 12 m, in contrast to17 m in site (b)) probably are due to80 - cabbage'760-40I-Iswdmpoak 01QnUt Floridasouthernred- sweet--cedar gum redw r-7 r 1 ?kor~ 25 I I 1 ] mulberrycabbagered -ZE!%sugar- I/ / 1 livewaxmyrtleFigure 36. Species comf30sition Of nearby Portions of a hydric-hammock stand in the northernpart of Tosohatchee State preserves Orange Total basal area is listed for each site.

frequent fires, a1 though timber harvesting(evident from stumps of southernred-cedar) also may be signifi -cant. It is likely that sites (a) and(b) are more diverse not only becauseof less frequent fire, but because ofless extreme hydrological regimes--their soils probably are kept moreconstantly saturated. The low diversityat sites (c) and (d) may also result,in part, from higher salt concentrationsthat exclude sal t-intolerantspecies.The boundaries of hydri c hammocks,as well as their species composition,may be affected by fire. When fire issuppressed in an adjoining communi ty,often by human actions, hydric hammockmay expand. In the Alafia River systemof west-central Florida, Clewel 1et al. (1982) documented the movementof wetland hardwoods (live oak, swamplaurel oak, water oak, and sweetgum)into pine flatwoods. They attributedthe extension to a reduction in firefrequency brought about by agri culturalactivities in the uplands.Though the expansion of hydric hammockinto freshwater marsh in Myakka RiverState Park is related partly tochanges in the hydro1 ogi cal regime(see section 3.32), fire suppressionalso may be responsible. Fire suppressionwithin the park has accompaniedincreased recreational use; frequencyalso has decreased in the watersheddue to land-use changes.3.3.5 Other DisturbancesHydric hammocks have a long historyof human use and disturbance that ischronicled in the accompanying managementhandbook (Simons et a7. 1989).Probably few areas have remained untouchedby human activities. Timberharvesting and 1 ivestock grazing probablyhave had the greatest influenceon the plant composition of contemporaryhydric hammocks.Logging of hydric-hammock forestshas greatly varied in intensity andscale, from selective cuts of scatteredindividual s to clearcutting ofhundreds of acres. Until the twentiethcentury, the cuts generally wereselective; the chosen species changedwith m2rket dz~:c?. I, :.,-highly valued in the era of woodenships, especially in the early andmid-1800's when the United Statesbuilt its navy. Pencil production inthe late 1800's was dominated bysouthern red-cedar, resul ti nq in theover-exploitation of this species inGulf Hammock. The large stumps thathave endured in the hydric hammocksshow that regenerated southern redcedarshave no+ yet reached their f:rmer size. In the 1800's and 1900's,sweetgum and some other hardwoods ofhydric hammocks were cut selectivelyfor manufacture of furniture andcrates, and these harvests too aremanifested in present size distri butions.Sweetgum trees between 50 and75 cm dbh are common in the TigerCreek hydri c hammock, 1 ong protectedfrom harvesting, and in remote areasof the gulf coastal hammocks, butelsewhere the trees are rarely greaterthan 50 cm dbh. Where past harvestingis particularly evident, for examplein the Orange Lake palm/oak hammock(Tab1 e 5), sweetgirms are less than 25cm dbh. Highgrading, the practice ofharvesting all merchantable treeswhile leaving species and individualtrees of low value, became popular inthe twentieth century. This type of1 oggi ng in hydri c hammocks general 1 yretains all cabbage palm, live oak,and an assortment of crooked and decayedtrees. The dominance of cabbagepalm and live oak in some hydric hammocks,for example the Orange Lakeforest, must in part be due to highgradingand selective harvesting practicesthat bypassed these species.Clearcutting became the predominantmode of harvest in hydric hammock inthe past 20 years. Sometimes thecl earcuts are a1 1 owed to regeneratenaturally, but more often, and particularly in the gulf coastal hammocks,they are planted with loblolly pine.uc; ;ak ~;l;Spanish explorers brought cattle andhogs to Florida in the sixteenth century,and wild hogs have roamed hydrichammocks and many other habitats eversince. Cattle still are grazed insome hydric hammocks where ground vegetationis relatively lush, such asthe inland reaches of the gulf coastal

hammocks and in loblolly pine-domi -natd hammnrkq. Cr27i179 hy cattle nrferal hogs can greatly influence theplant species composition of hydrichammock. Hogs avidly consume mast,infl uenci ng oak regenerati on, andtheir rooting for tubers, roots, andsmall invertebrates severely disturbsthe soil. Cattle compact the soil andreduce browse and groundcover vegetation,including hardwood regeneration(Figure 37). Where cattle density ishigh. hydric hammock may resemble amanicured lawn with an occasional tree(often live oak). The dominance ofcabbage palm and live oak in some hydrichammocks, for example the OrangeLake forest, probably is perpetuatedby preferential consumption of hardwoodseedlings and saplings by grazers.The major natural disturbances inhydric hammocks are floodins. fire,grazing, and wind. Browsing by nativewild3 ife. main1 v deer, influences theabundance and composition of groundcover,tree seedl ings, and shrub vegetation.Gaps from fallen trees arefound in all hydric hammocks, but theyare especial ly common in coastal areasvulnerable to hurricanes (Figure 38).Succession, whether on naturallycleared or on harvested sites, has notbeen documented in hydri c hammocks.However. some trends can be predictedbased upon species differences inshade to1 erance, re1 ati ve growth rate,longevity, requirements for seed1 ingestabl ishment, and ability to sproutfollowing disturbance (Putnam et a?.1960; Fowells 1965; McKnight et a7.1981). The composition of the regeneratingforest is affected by a numberof factors including the type, severity,and extent of disturbance, thecomposition of the pre-disturbanceFigure 37. A grazing exclosure in Gulf Hammock demonstrates the effects of cattle and deer on theground cover of hydric hammock. Wood fern is a predominant species both inside and outside theexclosure, but its height is greatly reduced by grazers.

Figure 38. Tree blowdowns due to hurricanes are a common sight in gulf coastal hammocks. Thissouthern red-cedar was located in a patch of hammock surrounded by salt marsh.stand (especially ages and species oftrees), and seed sources. If the disturbance(fire, hurricane, 1 oggi ng,etc. ) is re1 atively moderate, removingmainly canopy trees, then the new forestwill be dominated by the growth ofremai ninq seed1 i nqs and sap1 ings andsprouts from vigorous sprouters suchas sweetgum and persimmon. More severedisturbances favor sprouting ornew regeneration; the outcome probablydepends on the original composition ofthe stand. Virtually all hardwoodscommon to hydric hammock are capableof sprouting, but in many species thisability is inversely related to thesize of the tree. Therefore, if younghardwoods are present in the pre-di s-turbance stand and the disturbance isnot so intense as to destroy theirroots, they may then sprout vigorouslyto produce a new canopy. The new forestis likely to contain sweetgum andhornbeam, but a1 so oaks, sugarberry,and other hardwoods. On the otherhand, if the canopy contained mostlyolder hardwoods and little understorywas present prior to the disturbance,then severe disturbance will initiatenew regeneration in which 1 oblol lypine and sweetgum are dominantspecies. Sweetgum is the major colonizerfollowing most disturbances inhydri c hammock because usually it isan abundant member of the forest, itis a pro1 ific producer of seeds, andit vigorously sprouts from roots andstumps. However, loblolly pine ismore favored by severe, 1 arge-scal edisturbances, such as clearcuts withthe slash removed. Loblolly pine requiresmineral soil and the absence ofcompeting vegetation for good seedgermination and growth, and it regeneratesand grows very rapidly whenthese conditions are present in hydrichammock. The development of "mature"hydric hammock, with its abundance of

phosphorus-rich water; less productivecypress domes obtained 1 i ttlephospnorus aecause tneir water wasstagnant and derived mainly from rainfa1l . Hydric hammocks probably aremore productive than cypress domes(956 gm dry wt/m2; Brown 19811, and1 ess productive than bottom1 and h rdwood forests (1,374 gm dry wt./m 5 /yr-in a Louisiana stand; Connor and Day1976). Hydric hammocks are consideredto be still-water wetlands in contrastto those dominated by river flow; however,surface waEer may move acrossthe gentle slopes of hammocks duringand following floods. Flooding frequencyvaries among hydric hammocks,but in many, inundation occurs at theheight of the rainy season(s) --ei theronce or twice per year. Unlike cypressdomes, hydric hammocks receivewater from a variety of sources, includingrainfall, river overflow, andup1 and seepage.In forested wetlands, the two componentsof above-ground production, stemgrowth and the production of leaves,fruit, and flowers (measured as litterfa1 1 ) , are about equally important(Connor and Day 1982). Li tter-fallproduction in wet1 ands is positivelyrelated to the movement of water(Brinson et a7. 1981), but the relationshipwith flooding frequency isless clear. In Florida's ApalachicolaRiver floodplain, annual 1 i tter-fallin swamp communities is less than inhigh-ground, lev e stands (760 versus874 gm dry wt ./m 2 /yr respectively; El -der and Cairns 1982). The 1 evee soilsare saturated only during the floodseason, and their forests are dominatedby sweetgum, sugarberry, andswamp laurel oak. The first twospecies are especial? y high producersof leaf 1 itter; grape vines also contributeheavily to leaf fall in the1 evee forests. Litter production di f-fered only slightly between an uplandterrace forest and its adjacent floodplainforest in the South Carolinacoast 1 plain (606 versus 667 gm drywt./m/yr ? respectively; Shure andGottschal k 1985). Situated just beyondthe flooding margin, the terraceforest contained water oak, swampchestnut oak, poplar, and pignut hick-ory. The floodplain swamps of bothrivers were composed largely of tupelosand ash; the Apalachicola swampalso had abundant bald cypress, andthe South Carol ina forest includedsweetgum and red maple. The drierforests of the two systems have treespecies in common with hydric hammockand their flooding frequencies probablycircumscribe the range among hammocks.Litter production in hydrichammock is likely to be on the sameorder as the levee and upland terraceforests, but the seasonal pattern maybe quite different. In theApal achicol a River floodplain, 1 i tterfall peaks sharply in autumn (Elderand Cairns 1982). Most of the treespecies, including sweetgum, Americanelm, and hornbeam, follow this phenol -ogy, but swamp laurel oak sustainshigh leaf-fa1 1 production from Octoberthrough March. Whereas floodplainswamps typically are dominated by deciduousspecies, hydric hammocks containan abundance of evergreens.Still , evergreen 1 eaves are not permanent.In some cases, leaf fall ispulsed; 1 ive oak and sweetbay droptheir leaves at the same time as newones emerge in spring. Other species,such as cabbage palm, lose leavesthroughout the year.Litter must decompose slowly withinthe hydric-hammock community. Therate of decomposition is likely to beless than in wetlands with flowing waterand with a higher frequency andgreater duration of flooding . Leavesimmersed in the Apalachicola River decomposedsignificantly faster thanthose set on an unflooded levee site(Elder and Cairns 1982). No doubt waterflow enhanced the fragmentationand leaching of the material, a compositeof five major floodplainspecies. On the levee, even the mostrapidly decomposing leaves 1 ost lessthan half their mass in six months.Rate of leaf decomposition that decreasedacross floodpl ain transectscorrel ated with reduced flooding; however,changes in the species compositionof the litter probably were theproximate cause (Elder and Cairns1982; Shure et al. 1985). In theSouth Carolina floodplain system describedin the preceding paragraph,

locally collected litter lost 85% ofits mass in one year at a streambanksite, but only 58% annually in theterrace forest dominated by oaks(Shure et a]. 1985). The leaf decompositionrates of most floodplainspecies, including tupelo, ash, blackgum, red maple, and sweetgum, werevery rapid, but those of bald cypressand swamp laurel oak were considerablyslower (Elder and Cairns 1982; Shureet a?. 1985). Leaves that were recalcitrant to decomposition tended tohave higher carbon:nitrogen ratios andgreater concentrations of 1 ignin andcell ulose. Decomposition rates of oak1 eaves besides swamp 1 aurel were notmeasured, but they also are 1 ikely tobe slow because of similar chemicalcharacter] st~cs. Ai rnougn i i r;r;er decompositionis likely to be slow inhydric hammocks because of pl antspecies composition, low flooding frequency,and the absence of strong waterflow, probably only a small amountof 1 itter and decomposed material iswashed out by occasional floods. Theamount and form (part icul ate, dissolved)of export from a hammock dependson current velocity, the timingof floods in reiation to the seasonalpattern of litter fall, and uptakerates of dissolved nutrients by hammockplants.

CHAPTER 4. ANIMALS4.1 INTRODUCTIONVirtually nothing has been publishedabout invertebrates inhabiting hydrichammock. This habitat is important tocertain butterfl ies (John A. Fluno,Department of Entomology, Ohio StateUniversity, retired; pers. comm.) ,Sugarberry is the sole host plant forthe snout butterfly (Libytheana bachmanii), the hackberry butterfly(Asterocampa celtis), and the tawnyemperor (Asterocampa clyton) . Sugarberryand water elm (P7anera aguatica)are primary hosts for the questionmarkbutterfly (Polygonia interrogat ionis),with false nettle as an occasionalhost. Caterpillars of the easterntiger swallowtail (Papilio glaucus)feed on Fraxinus, L iriodendron, andNagno7 i a virginiana.Very few animals are endemic to hydrichammock. The crayfish Procambarusgeodytes, a primary burrower,appears to have an endemic distributionin the hydric hammock along SilverRiver and in other forested wetlandsof the Oklawaha River watershed(Franz 1976). A subspecies of thesoutheastern shrew (Sorex longirostriseionis) was describ.ed by Davis (1957)from the hydric hammock around HomosassaSprings, Citrus County,Florida. However, the actual distri -bution of this form and its validityas a taxon are uncertain (Humphrey etal. 1986). Despite their commonnames, neither the Gulf Hammock ratsnake nor the Gulf Hammock dwarf sirenare restricted to hydric hammock,though the rat snake occurs there; thesiren occurs in ponds in and beyondhydric hammock.Though the biology of most vertebratesfound in hydric hammock isfairly well known, very few have beenstudied there. The vertebrate faunaof hydric hammock is not unique to thehabitat, resembl ing faunas of mostother forested habitats in peninsularFlorida. However, the hydric-hammockfauna is luxuriant. Compared withother types of forest in the region,hydric hammock has a highly diversevertebrate fauna and a high abundanceof selected species.4.2 REPTILES AND AMPHIBIANSThe herpetofauna has no speciesunique to hydric hammock and is representativeof the region (Table 7).Characteristic species include southernblack racer, rat snake, Floridabox turtle, green anole, ground skink,broad-headed skink, southern toad,green treefrog, squirrel treefrog, andeastern narrow-mouth toad. Thecoastal variant of hydric hammocklacks many species found in the inlandvariants of hydric hammock wherefloodwater is fresh rather than brackish.The blue-striped ribbon snakeand blue-striped garter snake arewidespread in several habitats of thegulf coastal region, but they werefound only in the coastal sample ofhydric hammock shown in Table 7. Inland hydric hammock has an associationof reptiles and amphibians that appearsto be ecotonal between mesichammock and swamp forest. Lob1 01 lypine-dominated hydric hammock includesa number of species characteristic ofpine flatwoods (pinewoods snake, scarletking snake, pinewoods treefrog).Some quantitative data are avail ableon the reptile and amphibian community

Table 7. Occurrence of reptiles and amphibians in three variants of hydric hammock along the routeof the proposed Cross-Florida Barge Canal, including associated aquatic habitats such as ponds andstreams (Florida Game and Fresh Water Fish Commission 1976).SpeciesLob1 01 1 yCoastal Inland pinehydric hydric hydrichammocka hammock hammockEastern mud snake (Farancia abacura abacura) P-0Southern ringneck snake (Diadophis punctatus punctatus) P-F P-F?ine~oadz sszke (Rhadiqaea f!avf 7ata)Rough green snake (Opheodrys aestivus) P-F P-FSouthern black racer (Coluber constrictor priapus) P-F P-FEastern coachwhip snake (Masticophis flage7lum flagellum) P-0 P-IEastern indigo snake (Drymarchon corais couperi) P-0 C-FRed rat snake (Elaphe guttata guttata) P-0 P-0Yellow rat snake (Elaphe obsoleta quadrivittata)P-FGulf Hammock rat snake (Elaphe o. 9. x E. o. spiloides) C-0 P-FScarlet kingsnake (Lamprope7tis trianguium elapsoides) F-0Eastern kingsnake (Lamprope7tis getulus getulus) P-0Florida scarlet snake (Cemophora coccinea coccinea) P-0 P-0Rough earth snake (Virginfa striatula)Florida water snake (Nerodia fasciata pictiventris)Peninsula ribbon snake (Thamnophis sauritus sackeni)P-0P-FBlue-striped ribbon snake (Thamnophis sauritus nitae)C-FEastern garter snake (Thamnophis sirtalis sirtalis) P-0Blue-striped garter snake (Thamnophis sirtalis similis) C - FStriped crayfish snake (Regina a1 leni)Northern Florida swamp snake (Seminatrix pygaea pygaea) P-0Florida brown snake (Storeria dekayi victa) P-0 P-0Florida red-bellied snake (Storeria occipitomaculata obscura)P-FEastern coral snake (Micrurus fulvius fulvius) P-0 P-0Florida cottonmouth (Agkistrodon piscivorous conanti)P-FDusky pigmy rattlesnake (Sistrurus miliarius barbouri) P-F P-0Eastern diamondback rattlesnake (Crotalus adamanteus) P-0 P-0Stinkpot turtle (Stet-notherus odoratus)P-CStriped mud turtle (Kinosternon baurii)P-FFlorida mud turtle (Kinosternon subrubrum steindachneri)Florida box turtle (Terrapene carolina bauri) P-0 P-FPeninsula cooter (Pseudemys f loridana peninsularis) P-0Green anole (Anolis carolinensis carolinensis) P-F P-FSouthern fence lizard (Scelcporus undulatus undulatus) F-0Eastern glass 1 izard (Ophisaurus ventral is)P-FIs1 and gl ass 1 izard. (Ophisaurus compressus)P- IGround skink (Scincel7a laterale) P-F P-FBroad-headed skink (Eumeces laticeps) P- I C-FSoutheastern five-lined skink (Eumeces inexpectatus)P-FTwo-toed amphi uma (Amphiuma means) P-0Striped newt (Notophtha7mus perstriatus) P-0Peninsula newt (Notophthalmus viridescens piaropicola)P-FMote salamander (Ambystoma talpojdeum) P-0Slimy salamander (Plethodon glutinosus glutinosus)C-FDwarf salamander (Eurycea quadridigitata) P-0Southern dusky salamander (Desmognathus auriculatus) P-0Greater siren (Siren lacertina) P-0Eastern spadefoot toad (Scaphiophus ho7brooki holbrooki) P-0Oak toad (Bufo quercicus)Southern toad (Bufo terrestris) P-F P- FFlorida cricket frog (Acris gryllus dorsalis) P-0 P-FC-FP- IC- IP-FP- FP-FP-0P-0P-0P-0P-FP-FP- IC-FP-0P-FP-0P-FP-FP-0P-FP-0P-FP-F(Continued)

--Table 7. (Concluded),Loblol lyCoastal Inland pineSpecieshydric hydrichammocka hammockhydrichammockLittle grass frog (Limnaoedus ccularis) P-F P-FFlorida chorus frog (Pseudacris nigrita verrucosa)Green treefrog (Hyla cinerea) P- F P-FP- FP-FSouthern spring peeper (Hyla crucifer bartramiana) P-F P-0Pinewoods treefrog (Hyla femoralis)P- FRarking treefrog (Hyla qratiosa) P-0 P-FSquirrel treefrog (Hyla squirella) P- F P-F P-FCope's gray treefrog (Hyla chrysoscel is)Greenhouse frog (Eleutherodactylus planirostris)P-FP-FBullfrog (Rana catesbeiana) P-0 P- ISouthern leopard frog (Rana sphenocephala) P- F P-F P-FBronze frog (Rana clamitans clamitans)River frog (Rana heckscheri)P-0P- IPig frog (Rana gryl io)P-FEastern narrow-mouth toad (Gastrophryne carolinensis) P-F P-F P-FTotal number of species 2 8 51 50aThe coastal hammock sampled here was topographically diverse and included some mesicand xeric hammock. The sample also included the eastern hognose snake (Heterodonp7atyrhinos), Florida crowned snake (Tantil la relicta rei icta), Peninsula mole skink(Eumeces egregius onocrepis) , and Florida worm l izard (Rhineura f loridana) , which donot occur in hydric hammock. Other species characteristic of mesic or xeric hammockmay also have been over-represented in this sample.C = characteristicP = presentF = frequent0 = occasionalI - itlfreqiier~tof hydric hammocks. In a permanentplot in 1 oblol ly pine hydric hammock(Table 8), more than 97 individuals of8 species of reptiles and amphibianswere found in quadrat sampling, andmore than 245 individuals of 17species were found in time-constrainedsampl i ng . Compared wi th seven otherhabitats so sampled in the region, hydrichammock ranked second in the numberof species present and had morethan twice as many individuals as inany other habitat. The two methodstogether produced more than 342 individualsof 18 species, more than inany other habitat. The most abundantspecies were the green anole, groundskink, Florida cricket frog, 1 ittlegrass frog, squirrel treefrog, andgreenhouse frog. The little grassfrog and squirrel treefrog reachedtheir maximum abundance in this habi -tat.Arrays of drift fences fitted withpi t-fall traps and funnel -traps captureda different set of reptiles andamphibians (Tab1 e 9). Compared withten other habitats sampled by thistechnique, hydric hammock ranked secondin number of species and third innumber of individual s captured (withcaptures of the greenhouse frogdeleted because of its selection oftraps for shelter). The most commonspecies in hydric hammock were thenarrow-mouth toad, spadefoot toad, andground skink. The spadefoot toad

Table 8. Occurrence of reptiles and amphibians in a loblolly pine hydric hammock on the route of theproposed Cross-Florida Barge Canal (Florida Game and Fresh Water Fish Commission f 976).SpeciesSpring S urnrne r Autumn-Area- 1 imited search of a 1000-2 quadrat without time limitsScar1 et ki ngsnakeGreen anoleGround skinkOak toadLittle grass frogPinewoods treefrogSquirrel treefrogGreenhouse frogTotal species/individual s 5/9 7/>54Time-limited search for 6 man-hours without marked boundariesPinewoods snakeBlack racerFlorida cottonmouthFlorida box turtleGreen anoleGround skinkSoutheastern five-1 ined skinkUnidentified skinkOak toadSouthern toadFlorida cricket frogLittle grass frogPinewoods treefrogSquirrel treefrogGreenhouse frogSouthern 1 eopard frogEastern narrow-mouth toadTotal species/individual s 7/>62 12/>99reached its maximum abundance in thishabitat.During intensive sampling of theloblolly pine variant of coastal hydrichammock in St. Marks NationalWild1 ife Refuge, Wakulla County,Florida, 12 species of reptiles andamphibians were found (U.S. Fish andWildlife Service 119801): pigmy rattlesnake,ribbon snake, black racer,box turtle, green anole, ground skink,broad-headed skink, southern toad,green treefrog, squirrel treefrog,southern 1 eopard frog, and narrowmouthtoad. In continuous sampling inthis habitat (Table lo), the communityof reptiles and amphibians ranked secondin number of individuals and ninthin number of species, compared withsimilar samples from 11 other habitatsin that region. The most common

Table 9. Occurrence of reptiles and amphibiansin two drift fence arrays with pit-fall traps andfunnel-traps from 31 March lo 18 October 1975, ina hydric hammock on the route of the proposedCross-Florida Barge Canal (Florida Game andFresh Water Fish Commission 1976).Table 10. Occurrence of reptiles and amphibiansin two drift fence arrays with pit-fall traps andfunnel-traps from 9 December 1978 to 25 July1979, in coastal hydric hammock, loblolly pinevariant, in St. Marks National Wildlife Refuge,Wakulla Co., Florida (U.S. Fish and Wildlife Service1980).SpeciesNumberSpeciesNumberSouthern ri ngneck snakeEastern coachwhip snakeEastern garter snakeFlorida red-be1 1 ied snakeDusky pigmy rattlesnakeStriped mud turtleGreen anoleGround skinkBroad-headed ski nkSl imy salamanderEastern spadefoot toadSouthern toadLittle grass frogSouthern spring peeperSouthern 1 eopard frogBronze frogEastern narrow-mouth toadTotal speciesTotal individualsSouthern ringneck snakeSouthern bl ac k racerScarlet kingsnakeEastern ki ngsnakeFlorida scar1 et snakeFlorida water snakeRibbon snakeGarter snakeFlorida red-bell ied snakeFlorida mud turtleGreen anoleGround skinkBroad-headed ski nkOne-toed amphiuma(Amphiuma pho7eter)Sl imy sal amanderDwarf sal amanderSouthern toadCope's gray treefrogGreen treefrogSquirrel treefrogBull frogSouthern 1 eopard frogEastern narrow-mouth toadspecies were garter snake, red-be1 1 i edsnake, ground ski nk, broad-headedskink, green treefrog, squirreltreefrog, southern leopard frog, andnarrow-mouth toad.Studies of particul ar species ofreptiles or amphibians specifically inhydric hammock are almost non-existent.The lone exception is the easternindigo snake, a threatenedspecies. This species spends most ofthe winter in dens in hollow rootchannels and rodent burrows at thebases of large live oaks; other densites are armadi 11 o burrows, hol lowlogs, solution holes in limestone outcrops,and windrows of debris from1 umbering operations (Moler El9851 ) .Total speciesTotal individualsIndigo snakes breed in late autumn andwinter (Speake et al. 19781, whenmales make brief trips away from densapparently in search of females (Moler[I9851) . During summer, indigo snakesare often found in association withponds. Animals with access to bothclearcuts and uncut hammock seem toprefer the former, frequently occurringnear the ecotone of clearcut andhammock. Home ranges in winter average6.5 ha in size; in summer, theyaverage 158 ha (Moler 119851).

4.3 BIRDS4.3.1 Community StructureMore is known about the community ofbirds of hydric hammocks than aboutother groups of animals 1 iving there,but very little is known about specificbird populations in this habitat.The most common year-round resi -dents are the red-shouldered hawk,barred owl , red- be1 1 ied woodpecker,pifeated uoc~~ecker, northern f! ic4er,American crow, fish crow, blue jay,Carol ina wren, tufted ti tmouse, Carolinachickadee, and northern cardinal.The most common summer residentsare the great crested flycatcher,northern parul a warbler, and summertanager. The most common winter residentsare the eastern phoebe, Americanrobin, house wren, ruby-crownedking1 et, ye1 1 ow-rumped warbler, Ameri -can galdf inch, and whi te-throatedsparrow.Three variants of this habitat wereincluded in the study of flora andfauna of the route of the proposedCross-Florida Barge Canal (FloridaGame and Fresh Water Fish Commission1976). Raw data from that study havebeen analyzed by Humphrey and Nesbitt[1989] and are summarized in Table 11.rnastal hydric hammock wat characterizedby the highest absolute numberof kingfishers, fish crows, hermitthrushes, ye1 low-rumped warblers,northern cardinals in winter, andwhite-throated sparrows. This was theonly habitat in which gray kingbirdswere recorded. In1 and hydric hammockwas characterized by the highest absolutenumber of red-shouldered hawks,mourning doves, pileated woodpeckers,Carol ina wrens, bl ue-gray gnatcatchers,and black-and-white warblers.The loblolly pine variant of hydrichammock was characterized by the highestabsolute number of black vultures,downy woodpeckers, hairy woodpeckers,American crows in winter, wood pewees,brown-headed nuthatches, yellowthroatedvireos, and summer tanagers.Large, rare species for which hydrichammock was important included nestingospreys and American swallow-tailedkites (Elanoides forficatus) and post-breeding wood storks (Rycteriaamericana). an endanqered species.Hydric hammock supported large numbersof overwintering passeri nes. Notrecorded in these samples but veryabundant, especi a1 ly foraging over theedge of hydric hammock with marsh andwith fl atwoods, were flocks of overwinteringtree swal 1 ows (Iridoprocnebicolor).Bird communities in the three typesof hydric hammock (Table 11) were consistentlyamong the most diverse ofthose of 14 habitats in the region.In the breeding season, the diversityindex H1(Shannon and Weaver 1949) washighest in loblolly pine hydric hammock,second in rank in coastal hydrichammock, and fifth in inland hydrichammock. As measured by number ofspecies, diversity in the breedingseason ranked fourth, fifth, and sixthin the three types of hydric hammock.In winter, diversity as measured bynumber of species was higher in thethree types of hydric hammock than inany other habitat, and diversity asmeasured by H' was highest in inlandhydric hammock, fourth in rank in hydricloblolly pine hammock, and eighthin coastal hydric hammock. Abundanceof birds, as measured by number of individuals,was intermediate in thethree types of hydric hammock.A1 though each of the three types ofhydric hammock can be differentiatedby a few species of birds occurringonly or in highest number there, thesimilarity of these bird communitiesis much more striking than their differences(Humphrey and Nesbi tt[1989]). This similarity is not limitedto hydric hammocks; it extends toall types of forests sampled in theregion, suggesting that bird communi -ties do not differentiate among typesof forests nearly as finely as doforesters and ecologists. Instead ofbeing habitat-specific entities, thesebird associations appear to form accordingto the adaptations of individualspecies in response to gradientsof various features of the environment.Humphrey and Nesbi tt suggestedtwo such gradients on the basis of theavai 1 able data--three-dimensional

Table dl. Bird populations and community characteristics along the route of the proposed Cross-Florida Barge Canal. The sampling unit was the sum of counts made on a day at a site in a habitat; theWU~Q;;~;;~ 3: brk:r~ :CLT!E r2nrjecf frem 305 to 330 minutes In one day. Tabular values are averages ofthe one-day counts expressed as birds per hour. For species significantly affected by season inanalysis of varianceacross 14 habitats, counts areshown separately forwinter and summer; otherwise,they are combined. Some species were not observed anywhere in some seasons.VariableLob1 011 yCoastal Inland pineSeasonhydrichammockhydrichammockhydrichammockPopu7ation countsGreen-backed heron (Butorides striatus)Ki 1 ldeer (Charadrius vociferus)Turkey vulture (Cathartes aura)Black vulture (Coragyps atratus)Red-shouldered hawk (Buteo lineatus)Osprey (Pandion haliaetus)Northern bobwhite (Colinus virginianus)Mourning dove (Zenaida macroura)Yei i OM-bil 1 rd cuckoo (Coccyzus amcricanus)Barred owl (Strix varia)Be1 ted kingfisher (Ceryle alcyon)Red-bellied woodpecker (Melanerpes carolinus)Northern fl i cker (Col aptes auratus)Ye1 1 ow-be1 1 i ed sapsucker (Sphyrapicus varius)Downy woodpecker (Picoides pubescens)Hairy woodpecker (Picoides villosus)Pileated woodpecker (Dryocopus pileatus)Gray kingbird (Tyrannus dominicensis)Great crested flycatcher (Myiarchus crinitus)Eastern wood-pewee [Contopus virens)Eastern phoebe (Sayomis phoebe)Acadian flycatcher (Empidonax virescens)Purple martin (Progne subis)Blue jay (Cyanocitta cristata)American crow (Corvus brachyrhynchos)American crow (Corvus brachyrhynchos)Fish crow (Corvus ossifragus)Tufted titmouse (Parus bicolor)Carolina chickadee (Parus carolinensis)Brown-headed nuthatch (Sitta pusilla)Carolina wren (Thryothorus ludovicianus)Carolina wren (Thryothorus ludovicianus)House.wren (Troglodytes aedon)Ruby-crowned kinglet (Regulus calendula)Blue-gray gnatcatcher (Polioptila caerulea)Eastern bluebird (Sialia sialis)Hermit thrush (Catharus guttatus)American robin (Turdus migratorius)Loggerhead shrike (Lanius ludovicianus)Gray catbird (Dumetel la carol inensis)Northern mockingbird (Mimus po7yg70ttos)Northern mockingbird (Mimus polyglottos)Brown thrasher (Toxostoma rufum)winterwinter*bothwinter*winter*breedingbreeding*breeding*both*breedingboth*breeding*both*winter*winter*breeding*both*breedingbreeding*breedingwinterbreedingbreedingbreeding*breeding*winter*breeding*bothboth*breeding*breeding*winter*winterwinter*winter*breeding*winter*winterwinter*winter*breeding*winter*both(Continued)

Table 11. (Concluded).- ----VariableSeasonCoastalhydrichammockIn1 andhydrichammockLob1 01 1 ypinehydrichammockWhi te-eyed vireo (Vireo griseus)Yellow-throated vireo (Vireo f lavifrons)Sol itary vireo (Vireo so1 i tarius)Red-eyed vireo (Vireo 01 ivaceus)Northern parul a warbler (Parula americana)Black-and-white warbler (Mnioti?ta varia)Ye1 1 ow-rumped warbler (Dendroica coronata)Yellow-throated warbler (Dendroica dominica)Prairie warbler (Dendroica discolor)Pine warbler (Dendroica pinus)Common ye1 lowthroat (Geothlypis trichas)Eastern meadowl ark (Sturnel la magna)Redwi nged blackbird (Agelaius phoeniceus)Boat- tailed grackle (Quiscalus major)Common grackle (Quiscalus quiscula)Summer tanager (Piranga rubra)Northern cardinal (Cardina 1 i s cardinal is)Northern cardinal (Cardinal is caxdiilal is)American goldfinch (Carduelis tristis)Rufous-sided towhee (Pipilo erythrophthalamus)Rufous-sided towhee (Pipi lo erythrophtha 1 amus)Savannah sparrow (Passerculus sandwichensis)Bachman's sparrow (Aimophila aestivalis)Whi te-throated sparrow (Zonotrichia albicoll is)Swamp sparrow (Melospiza georgiana)Community characteristicsbreeding*breeding*winter*breeding*breedi ngwinter*winterbothbreedingwinter*bothwinter*breeding*bothbreeding*breeding*breeding*winter*winterbreeding*winter*winterbreeding*winter*winter1.5000.71.16.00.40001 .o00001.513.54.800.41.6000.40.8Number of speciesNumber of speciesDiversity index (H')Diversity index (H')Number of individualsNumber of individualsbreeding 24.0 22.0 23.5winter* 23.0 29.0 23.3breeding* 2.74 2.55 2.77winter* 2.30 2.96 2.60breeding* 70.2 95.0 92.2winter 57.6 80.0 94.4* Significant habitat effect for the season indicated (P < 0.05).structure of the vegetation rangingfrom homogeneous to heterogeneous,with hydric hammocks being near themu1 ti stratal extreme, and treecomposition ranging from a1 1 hardwoodto including substantial numbers ofconifers, with the loblolly pine variantof hydric hammock forming one extremeof this gradient. Severalspecies most strongly associated withthe conifer end of this gradient(ye? 1 ow- throated warbler, ye1 lowthroatedvireo, brown-headed nuthatch,and summer tanager) occur only or inhighest numbers in forests with codominantconifers.The importance of hydric hammock tooverwintering passerines can be surmisedfrom the high number of certainspecies in Table 11 (American robin,ruby-crowned kinglet, black-and-whitewarbler, ye1 1 ow-rumped warbler) , butthis phenomenon has not been describedadequately. The best information is

from censuses of a 13.5-ha hydric hammockin Hillsborough River State Park,Hi 1 l sborough Co., Florida (Wool fenden1967; 1968). Inis site supported 366individuals of 16 species per 40.5 haduring breeding versus 411 individualsof 33 species during winter. Speciesof overwintering migrants identifiedon this site (in addition to thosejust 1 i sted) include eastern phoebe,hermit thrush, ye1 low-be1 1 ied sapsucker,solitary vireo, ovenbird, orange-crownedwarbl er, palm warbler,common ye1 1 owthroat, and Americangoldfinch. Much attention has beengiven recently (e.g., Pasquier 1982)to the role of Neotropical forests insupporting populations of migrantpasserines that breed in eastern NorthAmerica, but the same role of forestsof the southeastern United States, andhydric hammock in particular, is lesswidely recognized. This role needs tobe much better documented and pub1 i -ci zed.Long-term trends in populations ofbreeding birds in Florida, based onbreeding bird survey counts from 1969to 1983 in all habitats, revealed significantchanges in the number of severalspecies occurring in hydric hammock(Cox 1987). These included increasesin number of the mourning doveand osprey and decreases in number ofthe northern fl icker, brown-headednuthatch, northern mockingbird, easternbl uebi rd, 1 oggerhead shrike, ye1 -1 ow- throated warbler, prairie warbler,common ye1 lowthroat, red-winged blackbird,and eastern meadowlark. Threeof the declining species are cavitynesters.4.3.2 Selected SoeciesThe wood duck (Aix sponsa) usuallyforages elsewhere but nests in cavitiesof live trees, including those inhydric hammock. This habitat may beimportant to wood ducks because of itsabundance of den trees and its proximityto water. Acorns and other mastare important fall and winter foods ofwood ducks (Landers et al. 1976).Wood ducks prefer to forage for mastfallen into shallow water or on theforest floor adjacent to water, and wehave observed this behavior in floodedhydri c hammocks.Wild turkey (Meleagris gallopavo)were subject to market hunting inFlorida until that commerce was bannedin 1901. The population in Gulf Hammockin 1948-49 was considered"reasonably good" (estimated at 500 to600 birds, or one per 71 ha of suitablehabitat) but far below carryingcapacity (Swindell 1949). Major pressureson the population were consideredto be hunting, which increasedafter World War 11, and reduction ofhabitat qua1 i ty by succession andcanopy closure foll owing 1 umbering.The lowest densities of turkeys in thearea are in extensive hammock unbrokenby clearings, which would providepoults with a suitable abundance ofinsects. Turkeys shift to the hammocksin autumn when acorns are avail -able and to hammock edges, flatwoods,and clearings in late winter and earlyspring when new herb growth becomesavai 1 abl e.The ivory- bi 11 ed woodpecker(Campephi 7us principal is) once inhabitedhydric hammock, as well as itsmodal habitat of river swamp and cypressswamp. Records of originaldistribution of this bird in "marlhammocks" (Tanner 1942) were up theSt. Marks River (Wakulla County);Pumpkin and Cal i fornia Swamps (DixieCounty) ; Suwannee Hammock, Rosewood,Otter Creek, Gulf Hammock, and Sim'sRidge (Levy County) ; Crystal River(Citrus County) ; Tampa (Hi11 sboroughCounty); Manatee County; Enterpriseand Turnbull Hammock (Volusia County) ;Jim Creek (Orange County); TaylorCreek and Wolf Creek (Osceol a County) ;Highl ands Hammock (Highl ands County) ;and Caloosahatchee region (LeeCounty) . The avai 1 able density esti -mates are that one pair of ivory-billsneeded 15.5 to 44 square kilometers ofgood habitat; the same stands supportedroughly 72 pileated woodpeckersand 252 red-be1 1 ied woodpeckers. Virginhydric hammock is characterized byvery old and standing dead trees,which supported the woodpeckers' maindiet of larvae of borers fbuprestid

and cerambyci d beet1 es) . Unl i ke 1 arvaeof other types of borers, theseoccur only between the bark and sapwood;hence they are available onlyfor a few years after death of thetree, and their overall density isquite low, even in virgin forest. Thenear extinction of this woodpecker isattributed mainly to cutting of virginforest, accelerated by hunting by Indiansfor ornaments, by scientists andcollectors for specimens, and by localresidents for curiosity and for food.The ivory-bi 1 l disappeared from mostof Florida by 1900-15. By 1935 only afew remnant populations were left; the1 ast re1 i able report from Gulf Hammockwas in 1934 (Tanner 1942).The Carol i na parakeet (Conuropsiscaro 7 inensis) a1 so occupied hydri chammock (McKinley 1985), but therecorded observations are too vague toprovide a sense of the importance ofthis association. Apparently thisspecies occurred primari 1 y in cypressswamps and pine flatwoods, where itfed on conifer seeds. However, thereare several specific references toCarolina parakeets seen or collectedin hydric hammock: Gulf Hammock, LevyCounty (Laurent 1906; Gordon 1909) ;hammock woods and cypress swampsaround Or1 ando and Sanford (Nehrl ing1896) ; and numerous 1 ocal i ti es thatare near both hydric hammock and primaryfeeding habitats.Although it did not appear in thesamples reported in Table 11 (perhapsdue to the highly clumped dispersionand nomadic travels of its flocks),the cedar waxwing (Bombyci 1 la cedrorum)is quite abundant in hydric hammock,especially in autumn and winter.The species is named for its preferencefor cedar berries (Martin et a7.1951); cedar berries may be exceptionallynutritious because of their highlipid content. A flock of waxwingsusually feeds until satiated and thenrests in a group on bare limbs of atree near the food source. Most seedspass rapidly through the digestivesystem and are deposited with the fecesnear but not under the parenttree, after which the birds often feedagain. This association of seed-dis-perser and fruit-producer may be aprimary force in maintaining the redcedarcomponeni uf iile iiYU;r ;L ;I~I~IIIIUL;Icommunity .4.4 MAMMALS4.4.1 Communitv StructureThe mammal s occurring in hydri c hammockhave been identified (Table 12),but the status designation providespoor resolution uf ab~:id~~;~t. In pitfalltrapping directed at reptiles andamphi bi ans, the small mammals capturedin hydric hammock or bayhead (datapooled) per 1,000 pit-nights were 2.70short-tai led shrews, 0.74 southeasternshrews, 0.25 least shrews, and 0.25golden mice (Florida Game and FreshWater Fish Commission 1976). In pitfalltrapping directed at shrews incoastal hydric hammock, the small mammalscaptured in 772 pit-nights werethree southeastern shrews, elevenshort-tai led shrews, and one leastshrew (Humphrey et a7. 1986). Mammaltracks counted per mile in loblollypine hydric hammock (Florida Game andFresh Water Fish Commission 1976) included11.33 for white-tailed deer,2.00 for armadillo, and 0.67 for feralhogs. Individual mammal s observedwhile night-lighting along 91 milestravelled in loblolly pine hydric hdmmock(Florida Game and Fresh WaterFish Commission 1976) included 34whi te-tailed deer, 20 armadillos, 5rabbits, 4 raccoons, and 3 opossums.Mi scel 1 aneous observations of mammal smade per 100 hours in a habitat type(Florida Game and Fresh Water FishCommission 1976) included: coastalhydric hammock--4 whi te-tailed deerand 4 long-tailed weasels; inland hydrichammock--3 whi te-tailed deer; andloblolly pine hydric hammock--22 armadillos,8 white-tailed deer, 5 raccoons,4 opossums, and 3 gray squirrels. The number of squirrel s seen orheard per hour whi 1 e conduct i ng poi ntcountsof birds (Florida Game andFresh Water Fish Commission 1976) washighest among 18 habitat types in inlandhydric hammock (2.40) -,ld thirdhighest in coastal hydric hammock(1.67), but none were detected in

Table d 2. Occurrence of mammals in three variants of hydric hammock along the route of the proposedCross-Florida Barge Canal, compiled from all sources (Florida Gameand Fresh Water Fish Commission1976).SpeciesLob1 01 1 yCoastal Inland pinehydric hydric hydrichammock hammock hammockVirginia opossum (Didelphis virginiana)Southeastern shrew (Sorex longirostris)Short-tailed shrew (B7arina carolinensis)Least shrew (Cryptotis parva)Red bat (Lasiurus borealis)Seminole bat (Lasiurus seminolus)Ye1 1 ow bat (Lasiurus intermedius)Evening bat (Nycticeius humeralis)Southeastern big-eared bat (Plecotus townsendii)Nine-banded armadillo (Dasypus novemcinctus)Eastern cottontail (Sylvilagus floridanus)Marsh rabbit (Sy 1vi lagus palustris)Gray squirrel (Sciurus carol inensis)Sherman's fox squirrel (Sciurus niger shermani)Southern flying squirrel (Glaucomys volans)Eastern woodrat (Neotoma floridana)Cotton mouse (Peromyscus gossypinus)Golden mouse (Ochrotomys nut tall i)Eastern harvest mouse (Rei throdontomys humul is)Marsh rice rat (Oryzomys palustris)Hispid cotton rat (Sigmodon hispidus)Gray fox (Urocyon cinereoargenteus)Long-tail ed weasel (Mustela frenata)Striped skunk (Mephitis mephitis)Raccoon (Procyon lotor)Florida panther (Fei is concolor coryi)Feral hog (Sus scrofa)Whi te-tailed deer (Odocoileus virginianus)C-FP-0P- FP-FP-FP- IP- IP- iP- IP- IC-FC-FP- IP-FP- IP-FP-0P-0P- IP- IP-0P-FC = characteristicP = presentF = frequent0 = occasionalI = infrequentloblolly pine hydric hammock. Theonly small mammals trapped in coastalhydric hammock and inland hydric hammockin the St. Marks Wildlife Refuge,Wakulla County, Florida, were cottonmice, golden mice, and easternwoodrats (U.S. Fish and Wildlife Service[1980]).Considering a1 1 the available information,at least the following speciesof mammals are characteristic of hydric hammock: opossum, southeasternshrew, short-tailed shrew, armadillo,gray squirrel, flying squirrel, cottonmouse, raccoon, feral hog, and whitetaileddeer. The Florida panther has

no specific habitat preference; it hasbeen reported i n coastal hammock(Pearson 1951) and hydri c hammock(Layne 1970). The panther probablywas characteristic of hydric hammockbefore the elimination of most breedingpopulations from northern Floridaby man. Like the panther, the Floridabl ack bear (Ursus f loridanusf loridanus) has no specific habitatpreference. Instead, the preferredhabitat of bears is a mosaic of wetlandand upland forests (Harlow 1961),including hydric hammocks. Bears arestill relatively common in and nearthe Ocala and Osceola NationalForests, which incl ude some hydrichammock habitat. Specific bear sighting~were reported for the Ocala NationalForest by Florida Game andFresh Water Fish Commission (1976) andfor the Osceol a National Forest (U.S.Fish and Wild1 ife Service 1978). Thesoutheastern brown bat (Plyotis austroriparius)and bobcat (iynx rufus)also occur in hydric hammocks (Pearson1954).4.4.2 Sel ccted SoeciesArmadillos have been studied extensivelyin Florida, though not withreference to specific habitats. Thediet of armadillos in Florida (Nesbittet al. 1977) consists mostly of insects(78% by volume) and includessmall quantities of earthworms (5%),reptiles and amphibians (1% andbirds and mammals (~1%). Though a varietyof vertebrates are included inthe diet, these account for ~0.01% ofthe items eaten (Wirtz et a]. 1985).Presumably an abundance of macroinvertebratesin the leaf 1 itter of hydrichammock is responsible for the abundanceof armadillos there, Armadillosdig burrows and den underground; inregularly flooded areas, burrows areplaced in patches of high ground.Gray squirrels occur at densities ofroughly 5 per ha in inland hydric hammockand 2.5 per ha in coastal hydrichammock in autumn (Jennings 1951).Staple foods of the gray squirrel inhydric hammock and adjacent forestsare the seeds of loblolly pine and theSeeds, buds, and flowers of hickories,Oaks, elms, magno1 ias, and red maples.The abcndzRcp r _ r ~ st2yl~ fnod' chiftsseasonal 1 y among 1 ~cal pl ant communi -ties, affecting the local distributionof sauirrel s. From september to mid-~anua'r~, acbrns' 'and hickory nuts areavailable in all habitats, squirrelsare widely distributed, and they becomefat. By mid-January the supplysf hard mast is exhausted, but budsand seeds of elm and maple b ~ ~ ~ m eabundant in hydri c hammock and riverswanp, and sqg;~yclr {perhepr the entire population) become concentratedin these communities, with a densityof 13.8 per ha recorded by Jennings.At thi s time squirrel s abandon coastalhydric hammock, because cedar berriesare no longer avail able; 1 ittle foodis produced and no squirrels are presentthere until autumn. In springand summer, squirrel s are dispersedthroughout river swamp, hydric hammock,and nresic hammock. Green nutsof hickories, oaks, and loblolly pinesare eaten in July and August. Usuallya gray squirrel can find at least twoof these plant communities by movingonly a few hundred feet, because manyhydric hammocks are small and ecotonalor, if large, are interspersed withdendri tic swamps and mesic ridges.Squirrels scatterhoard acorns in thesoil, especi a1 ly at elevated sitesnear the bases of trees and stumps.Acorns buried during drought periodsmay be covered subsequently by standingwater before they are consumed.The supply of stored food helps supportthe squirrel population throughthe winter and may be an essential resourcefor the spring breeding season.While lack of an acorn crop in one ora few species of oak is normal, a completefailure of acorn mast in allspecies is uncommon. In such a mastfailure, the few available acorns werecompletely harvested by wild1 ife bymid-October; thereafter the squirrelssurvived by recovering stored acorns(Jennings 1951). Gray squirrels nestin tree cavities (usually in liveoaks) during winter, and in the springthey build nests made of leaves andtwigs or of cabbage palm fibers. Ordinarilya SPrlng and a fall breedingseason Occur, but spring breeding does

no specific habitat preference; it hasbeen reported in coastal hammock(Pearson 1951) and hyaric nammoc~(Layne 1970). The panther probablywas characteristic of hydric hammockbefore the elimination of most breedingpopulations from northern Floridaby man, Like the panther, the Floridabl ack bear (Ursus f 1 oridanusf 7oridanus) has no specific habitatpreference. Instead, the preferredhabitat of bears is a mosaic of wetlandand upland forests (Harlow 1961),including hydric hammocks. Bears arestill relatively common in and nearthe Ocala and Osceola NationalForests, which include some hydrichammock habitat. Specific bear sighting~were reported for the Ocala NationalForest by Florida Game andFresh Water Fish Commission (1976) andfor the Osceola National Forest (U.S.Fish and Wildlife Service 1978). Thesoutheastern brown bat (Myotis austroriparius)and bobcat (Lynx rufus)also occur in hydric hammocks (Pearson1954) .4.4.2 Selected SoeciesArmadillos have been studied extensivelyin Florida, though not withreference to specific habitats. Thediet of armadillos in Florida (Nesbittet al. 1977) consists mostly of insects(78% by volume) and includessmall quantities of earthworms (5%),reptiles and amphibians (1% andbirds and mammals (~1%). Though a varietyof vertebrates are included inthe diet, these account for ~0.01% ofthe items eaten (Wirtz et a7. 1985).Presumably an abundance of macroinvertebratesin the leaf 1 itter of hydrichammock is responsible for the abundanceof armadi 11 os there. Armadi 110sdig burrows and den underground; inregul arly fl ooded areas, burrows areplaced in patches of high ground.Gray squirrels occur at densities ofroughly 5 per ha in inland hydric hammockand 2.5 per ha in coastal hydrichammock in autumn (Jennings 1951).Staple foods of the gray squirrel inhydric hammock and adjacent forestsare the seeds of lobfolly pine and theseeds, buds, and flowers of hickories,oaks, elms, magnolias, and red maples.TL,,,, - abundanc~ 3f rtz:!e fnndc chiftsseasonally among 1 ocal plant communi -ties, affecting the local distributionof squirrels. From September to mid-January, acorns and hickory nuts areavailable in all habitats, squirrelsare widely distributed, and they becomefat. By mid-January the supplyof hard mast is exhausted, but budsand seeds of elm and maple becomeabundant in hydric hammock and riverswamp, and squirrels (perbap: the entirepopulation) become concentratedin these communities, with a densityof 13.8 per ha recorded by Jennings.At this time squirrels abandon coastalhydri c hammock, because cedar berriesare no longer available; little foodis produced and no squirrels are presentthere until autumn. In springand summer, squirrel s are dispersedthroughout river swamp, hydri c hammock,and mesic hammock. Green nutsof hickories, oaks, and loblolly pinesare eaten in July and August. Usuallya gray squirrel can find at least twoof these plant communities by movingonly a few hundred feet, because manyhydric hammocks are small and ecotonalor, if large, are interspersed withdendri tic swamps and mesic ridges.Squirrel s scatterhoard acorns in thesoil , especi a1 ly at elevated sitesnear the bases of trees and stumps.Acorns buried during drought periodsmay be covered subsequently by standingwater before they are consumed.The supply of stored food helps supportthe squirrel popul ation throughthe winter and may be an essential resourcefor the spring breeding season.While 1 ack of an acorn crop in one ora few species of oak is normal, a completefailure of acorn mast in allspecies is uncommon. In such a mastfailure, the few available acorns werecompl etely harvested by wi ldl ife bymid-October; thereafter the squirrelssurvived by recovering stored acorns(Jennings 1951). Gray squirrel s nestin tree cavities (usually in liveoaks) during winter, and in the springthey build nests made of leaves andtwigs or of cabbage palm fibers. Ordinarilya spring and a fall breedingseason occur, but spring breeding does

not take place in years of mast fail -ure. Deaths due to starvation or diseasehave not been recordea In tnlshabitat.The eastern woodrat is most abundantin the ecotone between mesic and hydric hammocks (Pearson 1952). Breedingoccurs year-round. Nests aremarked only by very small piles ofsticks, and they are found in barns,hot 1 ow 1 ogs, and subterranean chambersunder stumps or the bases or roots oftrees.The golden mouse is most plentifulin areas having a dense thicket orshrub layer and a sparse herbaceousground cover. Nests are in denseshrubs or subterranean chambers.Golden mice use shrubs, hollow logs,and underground tunnels as escapecover (Pearson 1954). It is importantto note that although Pearson (1954)found golden mice only in adjacentrnesic hammocks, the Florida Game andFresh Water Fish Commission (1976) andthe U.S. Fish and Wildlife Service[1980] found them in hydric hammock.The cotton mouse is the most abundantmammal in Gulf Hammock (Pearson1953, 1954). Cover and nest sites aremore common in hydric hammock than inadjacent mesic hammock. Males havelarger home ranges than females. Homeranges of males overlap one another,but those of females do not. Homeranges are small er when populationdensity is high than when it is low.Breeding takes place during most ofthe year, but females seldom are pregnantin summer (May through August).Nests are in logs, stumps, and basesof trees; often they contain caches of1 ive oak and swamp l aurel oak acorns.The most important factor affectingpopulations probably i s the quantityof acorn mast--high densities of cottonmice decline after a mast failure.Potential competitors in poor mastyears include wild turkeys, blue jays,common grackles, gray squirrel s, easternwoodrats, opossums, whi te-taileddeer and (most importantly) feralhogs.bobcats,Potent i a1 predatorsbarred owl s, andincludeseveralspecies of snake. Parasitism bycuterebrid f1 y 1 arvae (Cuterebridae)in this habitat is heavy and may befatal. Cotton mice readily swim,climb vegetation, and jump to theground from considerable heights.The raccoon is an opportunist andgeneralist in both habitat and diet.Raccoons occur in every terrestrialand wetland habitat within their overallrange. Plants (most1 y nuts,drupes, and berries) make up 50%-80%of the raccoon diet. The drupes andberries have seeds that probably aredispersed rather than destroyed, includingbeautyberry, blackberry, bl ueberies,cabbage palm, palmettos, el -derberry, grapes, greenbri ars, hol -lies, pepper vine, persimmon, redcedar, sugarberry, swamp tupelo, andviburnum (F. Ctarper 1927; Ivey 1947;Caldwell 1963; Johnson 1970; Hal 1 s1977).Ihe black bear populat~on of witHammock was exterminated by about1950, because the local people consideredbears destroyers of property(Pearson 1954). The remaining fragmentsof bear distribution include extensiveareas of hydric hammock alongthe gulf coast of Pasco and HernandoCounties and Taylor and Wakulla Counties(Brady and Maehr 1985). Extantpopulations in the Osceola and OcalaNat io~ldl Tot ests also itiiltide tiiialhydric hammocks in their ranges. Likeother large mammals, the black bearhas broad habitat requirements and can1 ive wherever sufficient foraging areas,denning sites, and escape coverare available. Areas in Florida occupiedby black bears consist of largetracts of undeveloped forests containingdiverse vegetation types (Harlow1961). No seasonal movements amonghabitats by Florida black bear areknown, but they probably occur, especiallyin spring when mast suppliesare exhausted.The black bear is an omnivore, butmost of its diet is plant material,and mast is the prominent component.Examinat ion of stomach contents andscats of black bears in a variety ofFlorida habitats (Maehr and Brady1984) showed that the diet of black

ears in spr.ing is dominated by cabbage. .palm hearts, early growths of al-P'-- ll-L-7,. - rrl\n;r,17>f 3) ??C(I lYdtLJi ! lay [ r s w >L,r......saw palmetto, and other non-fruitplant parts, plus honeybees (Apis me7-7 ifera) and carpenter ants(Campanotus). In summer the dietshifts from soft vegetative parts toripening soft mast and early hardmast, including blueberries, gal 1 berries,blackberries, saw palmettoberries, honeybees, bess beet1 es(Odontotaenius disjunctus), carpenteraiits, w;? ki ngsti cks (Anf somorphabuprestoides) , paper wasps (Pol istes) ,and bumblebees (Bombus bimacul atus).In autumn and winter the diet consistsof hard mast and fruits of oaks, sawpalmetto, swamp tupelo, cabbage palm,needle palm, gal 1 berries, honeybees,and yellow jackets (Vespula). Bearsfeed on acorns both on the ground andarboreally. Vertebrates account foronly about 5% of the diet; species include gopher tortoi se (Gopheruspolyphemus), armadi 11 o, feral hog, andwhite-tailed deer. Black bears activelymaintain certain species in theplant communities they occupy by dispersingundigested seeds of the fruitsthey eat (Rogers and Applegate 1983;Maehr 1984). Major species invol vedin this mutualism ,include saw palmetto,cabbage palm, needle palm,swamp tupelo, blueberry, and raspberry,Probably numerous otherspecies (see Maehr and DeFazio 1985)also gain this advantage as minor dietarycomponents of bears.Sites preferred by black bears forwinter denning are cavities in largetrees, which provide protection fromweather and disturbance (Hami 1 ton andMarchinton 1980; Pelton et a7. 1980).Denning is especi a1 ly important forsows with cubs. The smaller size ofsows and the tendency of sows to denearlier than boars (Pelton et a7.1980) may give females access tosmall er cavities and the best-protectedden sites. Large, dense thicketsprovide escape cover from mostdangers (U.S. Fish and Wildlife Service1978), but hunters' dogs are deterredonly by very 1 arge water-fill edareas. In Florida the best such escapecover is bayheads, titi swamps,and hardwood swamps (Layne 1976;Williams 1978). The home ranges ofblack bears over1 ap broadly, but indi -vid~~al c avoid one another. Subadul tmales may be killed or driven away byadult ma1 es . Di spersi ng subadul ts of -ten move out of suitable habitat andare shot by humans. Home ranges ofblack bears in the Osceola NationalForest (which contains many very smallpatches of hydric hammock) are largeand variable relative to those inother areas of the United States,probably because of low quality of thehahi tat (James Mykvtka. Reynolds,Smith and Hills, Tampa; pers. comm.).The Florida panther inhabited hydrichammock near the town of Gulf Hammock(Levy County) and east of Cedar Keyuntil about 1950 (e-g., Pearson 1954).Now it is widely thought that allbreeding popul ati ons have been ext i r-pated from northern Florida. However,occasional reports 1 i ke the confi rmedsighting near the northern edge of theOcala National Forest (Table 12) andregular reports in the 1970's and1980's in and near hydric hammock onthe west bank of the St. Johns Riverin Orange and Seminole Counties indi -cate that a few individuals remain inthis habitat.Domestic hogs were first introducedinto Florida in 1539 by Hernando DeSoto (Lewis 1907). Although now abundantin hydric hammock, feral hogshave not been studied there. In SouthCarol ina, feral hogs usually avoidsalt marsh but make heavy use offresh- and bracki sh-water marsh and ofcypress-gum swamps (except during autumn);they use upland pine habitatsin proportion to avail abil i ty, andthey use upland hardwood forestslightly except when acorns are available(Wood and Brenneman 1980). Feralhogs feed in oak stands in autumn andwinter as long as acorns are available,and at other times they feed ongrasses, roots, and tubers on the marginsof marshes and swamps (Wood andRoark 1980). Where fewer habitats arepresent and individual hogs competeintensely for food (e.g., on OssabawIsland, Georgia), feral hogs frequentlyuse salt marsh (Graves and

Graves 1977). These observations suggestthat the supply ;C riast, par t iiu-1 arly acorns, is a key resource forferal hogs in hydric hammock. In someseasons or during mast failures, rootsare a dietary staple; feral hogs a1 soeat tree seedlings, chewing the roots,swallowing the sap and starches, andrejecting the woody tissue (Wood andRoark 1980). This destruction ofroots and seedlings can deter forestregeneration. Feral hogs are intensecompetitors with native wi?d:ife forthe supply of fallen mast. During ayear of mast failure, they may haveserious impacts on turkey, deer, and(to a lesser extent) bear populations.White-tailed deer were studied in avery general way in Gulf Hammock bySwi ndel 1 (1949). According to Bartram(1791) and early residents, deer werevery abundant in the region. Markethunting was intense in the late1800's, and by 1925 the population wasso reduced that sightings of deer signin Gulf Hammock were unusual. Subsequentlythe deer popul ation increasedto an estimated 100-325 animals forthe entire area by 1948-49. The legalharvest that year was documented asone deer per 900 acres of suitablehabitat. Swindell judged that hunting(which legally allowed harvest ofbucks only) continued to be the pri-mary limiting factor, based on thepr-evaiiing buc~:doe ratlo of 1:2.5 to1:4, with illegal "fire huntingw (withhead1 ights at night) af both sexes accountingfor at least 25% of the legalki 11 . Legal hunting usual 1 y employeddogs; this method was very effectiveand caused survivors to abandon theirhome ranges temporari ly to move to remoteportions of the hammock.Swindell considered the habitat to beexcel 1 ent in qua1 i ty but underused bybeer. Deer forayed rnaiiily in hjdrichammock in the dry springtime andmoved to the higher mesic hammocks insummer when hydri c hammock f 1 ooded .In autumn, deer sought acorns whereverthey occurred, preferring acorns fromlive oak > swamp chestnut oak > laureland water oaks. Diet of deer in theGulf and Richloam Hammocks in latefall and winter was 58% acorns, 30%woody plants, 6% forbs, and 6% mushrooms(War1 ow 1965).Subsequently 122,000 acres of Gul fHammock became the Gulf HammockWildlife Management Area. Ki 11 datafrom the period 1953-56 indicated apopulation of 1,400 to 1,800 deer forthe area. Along with an estimated3,000 cattle and 6,000 feral hogs, thetotal density of ungulates was approximatelyone per 11 acres--among thehezviest concentrations in Florida(Harlow 1959).

CHAPTER 5. PLANTANIMAL INTERACTIONSR primawy value of hydric hanlmock~for wild1 ife is the numerous tree cavitiesused by many species of wildlifefor nesting or cover. Den trees (livecavi ty-bearing trees) remain standingmuch longer (Carey 1983) than snags(dead cavity-bearing trees). In asurvey of den trees in various habitatsin Florida and South Carolina,McComb et a]. (1986) found that oldforest stands (>60 years) had 3 to 20times more dens than in young stands.The survey classified forest typesinto groups found in both states(hydric hammock does not occur inSouth Carol ina; Florida hammmocks ei -ther were included in another group orwere not sampled). The oak-tupelobaldcypress and palm categories hadrespectively the highest and secondhighestnumber of den trees and densamong 12 forest types in Florida.Loblolly pine plantations, to whichhydric hammocks often are converted,had among the lowest number of dentrees and dens--roughly 10% as many asin palm forest and 5% as many as inoak-tupel o-bal d cypress forest.Another feature of hydric hammock ofgreat significance to wildlife is theabundance and character of the seedsand fruits it produces. Seasonallyabundant seeds and fruits are importantsources of protein and phosphorus(Short and Epps 1976), whichfrequently are in 1 imited supply inforage available in southern forests(Blair and Epps 1969). The phosphorusrequirement in the diet of growing anddeveloping wild turkeys is 0.75%-0.80%, 1.0% for the northern bobwhite,0.5% for tree squirrels, and >0.25%for whi te-tai led deer (Hal 1 s 1970).Average values show that legume seedsare the best. source of both nutrients(31.2% and 0.54% of dry matter, respectively;Short and Epps 1976), butthe only common species of legume inhydri c hammock is water 1 ocust, whichoccurs on the marsh edge of somestands. Water locust seeds are consumedby black bears, cattle, andferal hogs. Numerous other seeds, includingthose of pines and sweetgum,average lower in protein (17.6% of drymatter) and phosphorus (0.37%) thanlegumes, but higher than acorns andfleshy and dried fruits. Pine seedsare used by squirrels and apparentlywere a dietary staple of Carolinaparakeets. Sweetgum seeds are a mainstayfor overwintering Americangoldfinches. Dried fruits, includingthe samaras of maple, elm, and ash,also are intermediate in content ofprotein (11.9% of dry matter) andphosphorus (0.24%), and they are importantfoods when fresh. At thattime, they are heavily eaten by animalssuch as the northern cardinal andgray squirrel . However, when dried,the high fiber content of these seedsmakes their digestibil ity very low.Despite relatively low content of proteinand phosphorus, acorns (5.9% and0.09% of dry matter, respectively) andfleshy fruits (8.4% and 0.22% of drymatter, respectively) are extremelyvaluable because of their pal atabil -ity, seasonal abundance, and accessibilityto a wide variety of wildlifespecies. In parti cu1 ar, acorns areavailable for a long period in autumnand winter. Acorns and fleshy fruitsare dietary staples of numerousspecies of birds and mammals, includingthe important game species (wildturkey, gray squirrel, raccoon, blackbear, feral hog, and white-taileddeer).

Consumption of fleshy fruits by residentand migrant birds has not beenexamined irl ityCir i~ iia~tltlt~~k, but it hasbeen studied in mesic hammock inFlorida (Skeate 1987). Most speciesof frugivorous birds are mutual isticseed-di spersers ; only three species(northern cardinal, summer tanager,and rose-breasted grosbeak, Pheucticus7udovicianus) are fruit "thieves" thatate the pulp and dropped the seed,undi spersed, be1 ow the parent plant.Consequently the support of this segmentof the dvifauna has re~ipro~dlimp1 icati ons for the reproduction andresilience of a subset of the plantcommunity. In mesic hammock this mutualism involves 22 species of birdsand 45 species of plants, dominated bymigrant species of birds and plantsthat fruit in fall or fall -to-winter.The greatest proportions of fruits areeaten by wintering American robins(57%) and cedar waxwings (20%); thegreatest diversity of fruits are eatenby wintering robins (18 species) andhermit thrushes (Catharus guttatus, 16species). The pace of seed-dispersalquickens in September and October, beginningwith arrival of four transientspecies of thrushes and continuingwith the arrival of wintering robins,waxwings, hermit thrushes, and ye1 1 owrumpedwarblers (Figure 39). By thepeak of seed-dispersal activity in December,28 species of bi rd-di spersedplants are in fruit, being consumed by10 winter and 4 year-round residentspecies of bird. Seed-di spersal activitydiminishes in March, and byApril -May only 3 species of bird-dispersedplants are in fruit, and thedisperser association diminishes to 4summer and 4 year-round residentspecies of bird. This forest has proportionatglyless summer fruiting (atotal of 9 species) and more winterfruiting (4 species in addition to the19 fall-winter fruiting species) thanin a forest in Illinois.Maintenance and elaboration of theassociation of bird and plant mutualist~is thought to be driven by afew pairs of strongly interactingbirds and plants (Skeate 1987). Theseare the veery (Catharus fuscescens)with Virginia creeper (Parthenocissus1BIRD-FRUIT DISPERSAL IN HAMMOCKSFigure 39. Phenology of bird diversity and abundanceand the number of species of plantsproducing fruits edible to birds in a mesic hammockin San Felasco State Preserve, AlachuaCounty, Florida (Skeate 1987).quinquefo7 ia) , Aral ia spinosa, andflowering dogwood (Cornus f 1 orida) ;robins with dogwood, red bay (Perseaborbonia), and cherry laurel (Prunuscaro7iniana) ; and waxwings withmistletoe (Phoradendron serotinum) .All these plants provide excellentfood because their fruits are unusuallyrich in lipids. Consumption of

these primary dietary items indirectlyc..,,I uvors seed dispersa: of other specie:that are less common or have less rewardingfruits, because the primaryfrugivores have varied diets. At thesame time, less abundant or effectivespecies of frugivorous birds, eachwith its own varied diet, are encouragedin the association by the predictablytimely resource of a varietyof fruits. Wide-ranging species ofprimary plants with high-1 ipid fruit(creeper, dogwood) supply rnigrati ngfrugivores with a high-qua1 i ty supplyof food continuously during migrationsouthward through the Atlantic flyway,and similar species with ranges restrictedto the southeastern UnitedStates offer resources to support thebird popul ati ons during winter.Another mutualism that may be importantin succession or invasion of hydrichammock is an association of treeswall ows and ye1 1 ow-rumped warblerswith wax-myrtl e and a microorgani sm(the soi 1 bacteri um Frankia) . Thisinteraction is important because ofthe aggressive pioneering of waxmyrtle,made possible by the nitrogenfixingof its actinomycete-nodul atedroots, its allelopathy to young competi tors, and its broadcast dispersalby large flocks of overwintering, omnivorousbirds (Schnoes and Humphrey1987).General i zed consumers (such as raccoonsand bears) and scatter-hoarders(such as blue jays and gray squirrels)also are prominent dispersers of seedsof plants of hydric hammock.Seedlings of cabbage palm and red bayhave been observed growing in bearscats (Maehr 1984). The diversity ofdrupes and herrips eaten bv bears andraccoons suggests the size of the assemblage of mutual i sts involved. A1 -though general ized consumers usuallykill nuts by eating them and hence donot benefit oak trees, acorns nonethelessare crucial to the interactiveroles of general i zed consumers by at -tracting them to the sites where drupesand berries also are eaten anddispersed. In contrast, jays andsquirrel s store nuts (especial 1 yacorns) individually in the soil for1 ater consumption. Nuts forgotten orleft because of death or departure ofthe consumer contribute to reproductionof the mature forest or successionafter disturbance or lumbering ofthe forest.The diversity of mast-producingspecies in southern forests produces acomposite phenol ogy yielding foodyear-round for generalized consumerscapable of eating both hard and softmast (Harris et a7. 1979). However,the continuity of this supply is incompletewithin specific types of forest.For example, little hard mast isproduced in hydric hammock in winter,and duration of the hiatus betweenconsumption of the last acorns andavailability of the first elm andmaple seeds varies according to thevolume of the year's acorn crop. Tobridge discontinuities in the seasonalfood supply, juxtaposition or interspersionof two or more types of forestis critically important to manymobile or large species of wild1 ifethat move tc find food. Documentedexamples are given in the discussionof individual species of mammals andbirds, Chapter 4.

CHAPTER 6. LINKAGES WITH OTHER ECOSYSTEMSHydric hammocks are interconnectedwith up1 and and downs1 ope communi ti esby flows of water and movements of animals. Because of high ground-waterlevel and, sometimes, discharge fromdeep aquifers, a hammock may be connectedto recharge areas, such as thesandy uplands of the central Floridaridge, which are located far beyondthe watershed defined by surfacedrainage. Hydric hammocks often interceptsurface and ground water beforeit enters rivers and estuaries,perhaps altering the pattern and qualityof flow. Some animals merely passthrough hydric hammocks en route topreferred habitat, but for others,this forested wetland provides criticalshelter and food during part oftheir life cycles. Movements of animals,and their transport of fruitsand seeds, 1 ink both the plant and animalcomponents of hydric hammocks toadjacent and distant communities,6.1 WITH ESTUARIESThe extensive band of hydric hammocksalong the gulf coast of Florida(Figure 1) links upland and saltmarsh/estuarine systems. The most irnpor.tantfeature of this coupling isthe modification, by passage of waterthrough hydric hammock, of the quantity,timing, and quality of freshwaterthat reaches the estuaries.Changes in one or more of these parameterscan greatly alter estuarinestructure and productivity.The flow of freshwater into estuarieshas numerous consequences(Snedaker and deSylva 1977), of whichthe major ones are the reduction ofsal ini ty; the induction of freshwaterand saltwater mixing; the establishmentof horizontal and vertical gradi -ents of salinity; and the addition ofnutrients, dissolved and particulateorganic matter, and sediments derivedfrom the watershed. The quantity andtiming of freshwater del ivery determine,in part, seasonal changes inphysico-chemical regimes that in turnare reflected in regular patterns ofestuarine use by animals. Year-roundinhabitants, mainly invertebrates, areplentiful, but the great number andvariety of part-time residents distinguishestuaries. In the southeasternUnited States, most fish and invertebratespecies of commerci a1 importanceuse estuaries and their fringing saltmarshes as nurseries, entering andleaving the shallow waters in a regularseasonal progression from 1 atefall through early summer (Sheridanand Livingston 1979; Rogers et a].1984). Temporal and spati a1 di stributionsof species within estuaries areassociated with physico-chemical andbiotic factors, many of which are influenced by freshwater inflow (e.g.,sal ini ty and food availability) . Incoastal waters, annual freshwater inputi s correlated with production ofsome invertebrate and fish species,but the mechanisms generally are notwell understood (Meeter et a7. 1979;Sinclair et a7. 1986). The effect offreshwater discharge on the abundanceof oysters is mediated by changes insalinity. The optimal salinity rangefor oyster growth in Apalachicola Bayis 15%-25%, but a more important roleof salinity is indirect in that lowsal ini ty excludes important predatorsof oysters (Menzel et a7. 1966).How forested wet1 ands influence thedelivery of freshwater to estuaries

can be inferred from the consequencesof timber harvesting in swamps withinthe drainage area of Apalachicola Bay,in northwestern Florida (Livingstonand Duncan 1979). Normally theApal achicola River's flow into the baypeaks in winter-spring, whereas 1 ocalrainfall is highest in summer. Thediscrepancy in seasonal patterns ofriver flow and rainfall is due to thehigher evapotranspiration of watershedvegetation in summer. In contrast,runoff from clearcut swamps is pulsedand timed with bouts of rain. Hydrichammocks also are likely to stabilizethe timing of freshwater inflow to adjoiningestuaries and to attenuatepeak flows that, unchecked, causeabrupt changes in sal ini ty. Howforested wetl ands influence the qua1 -i ty of freshwater inflow to estuariesis less clear. Both the magnitude andtiming of flooding affect the amountsand forms of materials transportedfrom forested wetl ands to estuaries(Livingston 1984). Nutrient and pol -lutant assimilation by wetlands is enhancedby slow passage of water overthese flat, densely vegetated areas.On the other hand, materials can betransferred from wetl ands to receivingwaters during periods of high flow. Astrong positive correl ation betweenrate of flow and concentration of detrituswas observed during winter andspring in the lower Apalachicola River(Livingston 1984). No correlation wasfound in summer; then, even occasionalhigh flows contained low levels of detritus. Particulate organic matterwas transferred from the floodplain tothe estuary primarily during winterand spring floods because river flowrates were high and because the floodsoccurred soon after the 1 i tter-fallpeak (Elder and Cairns 1982). Consideredover an annual cycle, the floodplainforest of the Apal achicol a Riverexported nutrients to the estuary(Mattraw and Elder 1982). Hydric hammockis flooded less frequently andseverely than ri ver-dominated floodplainforest, so net export of nu-trients is likely to be small. Floodsdue to hurricanes may occasionallyflush detritus from coastal hydrichammocks into estuaries, a1 thoughtheir timing (late summer) is suchthat the standing crop of litter ,islikely to be small.6.2 WITH ADJACENT AND DISTANT HABITATSMany kinds of animals move amonghabitats to find essent i a1 supportduring the annual cycle or in differentstages of their life cycles. Localmovements of resident animal swithin their home ranges can link hydrichammock with iipsl ope habitats.However, this phenomenon has seldombeen documented because of the paucityof research on animals inhabiting hydrichammocks. The reports of movementsof eastern indigo snakes betweenhydri c hammocks and cl earcuts [Mol er19851 and of seasonal shifts in density of gray squirrel s between hydri cand mesic hammock (Jennings 1951) areexceptions that prove the rule.Longer movements of non-sedentarypasserine birds that spend much of thewinter in hydric hammock likewise linkthis habitat with upland forests inFlorida. While on winter range, thesebirds are nomadic and non-specific intheir habitat requirements. In latewinter and early spring, large numbersof normally forest-dwel 1 i ng passeri nesappear in various upland and disturbedhabitats to take advantage of isolatedand newly avai 1 able food resources.Unfortunately, this movement is socomplex and variable among places andyears as to seem chaotic; it probablywill never be documented adequatelywithout a massive investment of fundingfor research. Some of the mostcommon species involved are seed-di s-persi ng mutual i sts (tree swallow,cedar waxwing, American robin, ye1 1 owrumpedwarbler) ; this l inkage presumablyalso maintains similarities incomposition of the various plantcommunities.Finally, long-distance, seasonal mi -grations of these same species betweenbreeding and non-breeding ranges(Figure 40) 1 ink otherwise unrelatedhabitats through support of the birdsover their annual reproductive cycles.Especially for strongly mutualistic

R Amerlcan swallow-tailed kite b. tree swallowC. hermit thrush d. American robinspecies, these linkages set up interdependencies,the scope of which canonly dimly be envisaged. The treeswall ow, for example, i s omnivorousduring winter and strikingly involvedin plant succession of Florida habi -tats because of its great numbers andwide dispersal of seeds of the pioneeringwax-myrtle. On breedingrange, the tree swallow population islimited by a declining resource, cavitiesmade in trees by other animals(Erskine 1979). Possibly a decline innesting of tree swallows in easternNorth America may affect succession ofhydri c hammocks.Breedng year-round WnterFigure 40. Breeding and winter ranges ofselected migrant birds that live in hydric hammockfor part of the year: (a) American swallowtailedkite; (b) tree swallow; (c) hermit thrush; (d)American robin; (e)cedarwaxwing; and (f) yellowrumpedwarbler.

REFERENCESApplin, P.L. 1951. Preliminaryre~ort on buried pre-Mesozoic rocksin Florida and adjacent states.U.S. Geol. Surv. Inf. Circ. 47. 15PP.Barry, J.M. 1980. Natural vegetationof South Carolina. University ofSouth Carolina Press, Columbia. 214PP.Bartram, W. 1791. Travels throughNorth and South Carolina, Georgia,East and West Florida. Q.M. vanDoren, ed. Dover Pub1 ications, NewYork (1928 reprint). 414 pp.Bechtold,. W.A., and H.A. Knight.1982. Florida's forests. U.S. Dep.Agric. For. Serv. Resour. Bull. SE-62. 84 pp.Bedinger, M.S. 1978. Re1 ati onbetween forest species and flooding.Pages 427-435 in P.E. Greeson, J.R.Clark, and J.E. Clark, eds. Wetlandfunctions and values: the state ofour understanding. American WaterResource Association, Minneapolis,Minn.Blair, R.M., and E.A. Epps, Jr. 1969.Seasonal distribution of nutrientsin plants of seven browse species inLouisiana. U.S. Dep. Agric. For.Serv. Res. Pap. SO-51. 35 pp.Brady, J.R., and D.S. Maehr. 1985.Distribution of black bears inFlorida. Florida Field-Nat. 13: 1-7.Brinson, M.M., A.E. Lugo, and S.Brown. 1981. Primary productivity,decomposition, and consumer activityin freshwater wetl ands. Annu. Rev.Ecol . Syst. 12: 123-161.Brown, K.E. 1973. Biological lifehistory and geographicaldistribution of the cabbage palm,Sabal palmetto. Ph.D. Dissertation.North Carol ina State University,Raleigh. 101 pp.Brown, M.T., and E.M. Starnes. 1983.A wetl ands study of Seminole County:identification, evaluation andpreparation of development standardsand guidelines. Cent. Wet1 andsTech. Rep. No. 41. University ofFlorida, Gainesville. 284 pp.Brown, S. 1981. A comparison of thestructure, primary productivity, andtranspiration of cypress ecosystemsin Florida. Ecol. Monogr. 51:403-427.Caldwell, J.A. 1963. Aninvestigation of raccoons in northcentralFlorida. M.S. Thesis.University of Florida, Gainesville.107 pp.Carey, A.B. 1983. Cavities in treesin hardwood forests. Pages 167-184in J.W. Davis, G.A. Goodwin, andR.A. Ockenfels, eds, Snag habitatmanagmen t : proceedings of thesymposium. U.S. For. Serv. Gen.Tech. Rep. RM-99.Clewel 1, A. F. 1981. Natural settingand vegetation of the Floridapanhandle, Contract No. DACW01-77-C-0104. U.S. Army Corps ofEngineers, Mobile, Ala. 773 pp.Clewell, A.F., J.A. Goolsby, and A.G.Shuey. 1982. Riverine forests ofthe south prong Alafia River system,Florida. Wetlands 2:21-72.

Conner, W.H., and J.W. Day, Jr. 1976.Productivity and composition of abaldcypress-water tupelo site and abottomland hardwood site in aLouisiana swamp. Am. J. Bot.63:1354-1364.Conner, W.H., and J.W. Day, Jr. 1982.The ecology of forested wetlands inthe southeastern United States.Pages 69-87 in B. Gopal, R.E.Turner, R.G. Wetzel, and D.F.Whigham, eds. Wet1 ands--ecology andmanacjement. National Institute ofEcology and International ScientificPub1 ications, Jaipur, India.Conover, C.S., J.J. Geraghty, and G.G.Parker. 1984. Ground water. Pages36-53 in E.A. Fernald and D.J.Patton, eds. Water resources atlasof Florida. Florida StateUniversity, Tallahassee.Cooke, C.W. 1939. Scenery of Floridaas interpreted by a geologist. Fla.Geol. Surv. Bull. No. 17. 118 pp.Cowardin, L.M., V. Carter, F.C. Golet,and E.T. LaRoe. 1979.Classification of wetlands anddeepwater habitats of the UnitedStates. U.S. Fish Wildl. Serv.,Bi 01. Serv. Program FWS/OBS-79/31.103 pp.Cox, J. 1987. The breeding birdsurvey in Florida: 1969-1983. Fla.Field Nat. 15:29-56.Davis, J.H., Jr. 1943. The naturalfeatures of southern Florida. Fl a,Geol. Surv. Bull. No. 25. 311 pp.Davis, J.A., Jr.. 1957. A new shrew(Sorex) from Florida. Am. Mus.Novit. 1844:l-9.Dohrenwend, R.E., and L. Harris.1975. A climatic change impactanalysis of peninsular Florida 1 ifezones. Pages 107-122 in Impacts ofclimatic change on the biosphere.Climatic Impact Assessment Program,Monogr. 5, Part 2, U.S. Dep.Transp., Washington, D.C.Dunn, W.J. 1982. Plant communitiesand vascular flora of San FelascoHammock, A1 achua County, Florida.M.S. Thesis. University of Florida,Gainesville. 134 pp.Elder, J.F., and D.J. Cairns. 1982.Product i on and decomposition offorest litter fall on theApal achicol a River flood plain,Florida. U.S. Geol. Surv. WaterSupp. Pap. 2196-0. 42 pp.Erskine, A.J. 1979. Man's influenceon potential nesting sites andpopulations of swal 1 ows in Canada.Canad. Field-Nat. 93:371-377.Ewel, K.C., and W.J. Mitsch. 1978.The effects of fire on speciescomposition in cypress domeecosystems. Fla. Sci. 41:25-31.Faul kner, G. L. 1973. Geohydrol ogy ofthe Cross-Fl orida Barge Canal areawith special reference to the Ocalavicinity. U.S. Geol. Surv. Water-Resour. Invest. 1-73. 117 pp.Florida Game and Fresh Water FishCommission. 1976. Cross FloridaBarge Canal restudy report.Wild1 ife study. Vols. 1-5.Wild1 ife Research Laboratory,Gainesville, Fla.Florida Game and Fresh Water FishCommission. 1987. Official listsof endangered and potenti a1 1 yendangered fauna and flora inFlorida. Tallahassee, Fla. 19 pp.Florida Natural Areas Inventory.1984. Florida natural communities.The Nature Conservancy, Tall ahassee,Fla. 12 pp.Fowells, H.A. 1965. Silvics offorest trees of the United States.U.S. Agric. Handb. No. 271. 762 pp.Franz, R. 1976. Freshwatercrustaceans: their distribution inthe Cross Florida Barge Canal area.Pages J-1 to J-19 in Florida Gameand Fresh Water Fish Commission,Cross Florida Barge Canal RestudyReport. Wildlife study, Vo1. 5.

Harper, R.M. 1927. Natural resourcesof southern Florida. Fla. Geol.Surv. Annu. Rep. 18:27-192.Harris, L.D., D.H. Hirth, and W.R.Marion. 1979. The development ofsilvicul tural systems for wild1 i fe.Pages 65-82 in C.L. Shilling andJ.R. Tolliver, eds. Recreation inthe South's third forest. Proc.28th Annu. La. State Univ. For.Symp., Baton Rouge.Healy, W.G. 1975. Terraces andshore1 ines of Florida. Fla. Bur.Geol. Map Ser. No. 71.Hook, D.D., and R.M.M. Crawford, eds.1978. Pl ant 1 i fe in anaerobicenvi ronments . Ann Arbor Sci encePubl., Ann Arbor, Mich. 564 pp.Hosner, J . F. 1960. Relativeto1 erance to complete inundation offourteen bottoml and tree species.For. Sci. 6:246-251.Hosner, J.F., and S.G. Boyce. 1962.Tolerance of water saturated soil ofvarious bottoml and hardwoods. For.Sci. 8:180-186.Huffman, R.T., and S.W. Forsythe.1981. Bottom1 and hardwood forestcommunities and their relation toanaerobic soil conditions. Pages187-196 in J.R. Clark and J.Benforado, eds. Wetlands ofbottom1 and hardwood forests.Developments in Agricultural andManaged-forest Ecology, Vol . 11.Elsevier Scientific, New York.Humphrey, S.R., and S.A. Nesbitt.[1989]. the community approach towildlife conservation: characteristicsof bird communities innorthern Florida.Humphrey, S.R., R.W. Repenning, andH.W. Setzer. 1986. Status surveyof five Florida mammals. FloridaCoop. Fish and Wildlife Res. Unit,Tech. Rep. No. 22. 117 Newins-Ziegler Hal 1, University of Florida,Gainesville. 38 pp.Ivey, R.D.W. 1947. The mammals,excluding bats, of Palm Valley,Florida. M.S. Thesis. Universityof Florida, Gainesville. 129 pp.Jennings, W.L. 1951. A study of thelife history and ecology of the graysquirrel (St' il;rus c. caro7inensisGael in) in Gulf Hammock. M.S.Thesi s. University of Florida,Gainesville. 151 pp.Johnson, A.S. 1970. Biology of theraccoon (Procyon lotor varius Nel sonand Goldman) in Alabama. A1 a. Coop.Wildl. Res. Unit, Auburn Univ. Bull.No. 402. 148 pp.Jones, R.H. 1981. A classificationof lowland forests in the northerncoastal plain of South Carolina.M.S. Thesis. Clemson University,Clemson, S.C. 177 pp.Jordan, C. L. 1984. Florida's weatherand cl imate: imp1 ications forwater. Pages 18-35 in €.A. Fernaldand D.J. Patton, eds. Waterresources atlas of Florida. FloridaState University, Tall ahassee.Kozl owski , T. T. 1984a. Extent,causes, and impacts of flooding.Pages 1-7 in T.T. Kozlowski, ed.Fl ooding and plant growth. AcademicPress, Or1 ando, Fl a.Kozl owski, T.T. 19846. Responses ofwoody plants to flooding. Pages129-163 in T.T. Kozlowski, ed.Fl ooding and pl ant growth. AcademicPress, Or1 ando, Fl a.Kurz, H., and K. Wagner. 1957. Tidalmarshes of the Gulf and Atlanticcoasts of northern Florida andCharleston, South Carolina. Fl a.State Univ. Stud. 24. 168 pp.Laessle, A.M. 1942. The plantcommunities of the We1 aka area.Univ. Fla. Biol. Sci. Ser. 4(1).143 pp.Laessle, A.M., and C.D. Monk. 1961.Some live oak forests ofnortheastern Florida. Q. J. Fla.Acad. Sci. 24:39-55.Landers, J.L., A.S. Johnson, P.H.Morgan, and W.P. Baldwin. 1976.

Duck foods in managed tidalimpoundments in South Carol ina. J.a,..,wt 141. k~2~j2. 4":7?1-778Laurent, P. 1906. Bird notes from aFlorida porch. Bird-Lore 8:67.Layne, J.N. 1970. Terrestri a1wi ldl jfe of the Okl awaha regionalecosystem. Pages 33-41 inEnvironmental impact of the Cross-Florida Barge Canal with specialemphasis on the Oklawaha regionalecosysteg. El orSda Defenders of theEnvironment, Gainesvil le. 115 pp.Layne, J.N. 1976. Introduction.Pages 8-11 in J.N. Layne, ed.Mammal s . Rare and endangered biotaof Florida, Vol. 1. UniversityPresses of Florida, Gainesvi 1 le.Leitman, H.M., J.E. Sohm, and M.A.Frank1 in. 1983. Wetl and hydrologyand tree distribution of theApalachicola River flood plain,Florida. U.S. Geol. Surv. WaterSupp. Pap. 2196-A. 52 pp.Lewis, T.H. 1907. Spanish explorersin the southern United States 1528-1543. Char1 es Scri bner's Sons, NewYork. 411 pp.Lichtler, W.F., W. Anderson, and B.F.Joyner. 1968. Water resources ofOrange County, Florida. Fl a. Geol .Surv. Rep. Invest. No. 50. 150 pp.Little, E. L., Jr. 1978. Florida.Atlas of United States trees, Vol.5. U.S.D.A. Misc. Publ . 1361. 22pp., 268 maps.Little, E.L., Jr. 1979. Checklist ofUnited States trees (native andnatural izedj. Agric. Handb. No.541. U.S.D.A. For. Serv.,Washington, D.C. 375 pp.Livingston, R.J. 1984. The ecologyof the Apalachicol a Bay system: anestuarine profile. U.S. Fish Wildl.Serv. FWS/OBS 82/05. 148 pp.Livingstone, R.J., and J. Duncan.1979. Short- and long-term effectsof forestry operations on waterqua7 i ty and epi benthic assemblagesof a north Florida estuary. Pages339-382 in R.J. Livingstone, ed.Ecological processes in coast.al andmarine systems. Plenum Press, NewYork.Maehr, D.S. 1984. The black bear asa seed disperser. Fla. Field Nat.12:40-42.Maehr, D.S., and J.R. Brady. 1984.Food habits of Florida black bears.J. Wildl . Manage. 48:230-235.Maehr, D.S., and J.T. DeFazio, Jr.1985. Foods of black bears inFlorida. Fla. Field Nat. 13:8-12.Martin, A.C., H.S. Zim, and A.L.Nelson. 1951. American wild1 ifeand plants. McGraw Hill Book Co.,New York. 500 pp.Martin, R.A., and S.D. Webb. 1974.1.at.e Pl ei stocene mammal s from theDevil's Den fauna, Levy County.Pages 114-145 in S.D. Webb, ed.Pl ei stocene mammal s of Florida.University of Florida Press,Gainesville.Mattraw, H.C., and J.F. Elder. 1982.Nutrient and detritus transport inthe Apal achicola River, Florida.U. S. Geol . Surv. Water Supply Pap.2196-C.McComb, W.C., S.A. Bonney, R.M.Sheffield, and N.D. Cost. 1986.Den tree characteristics andabundance in Florida and SouthCarol ina. J. Wildl . Manage. 50:584-591.McKinley, D. 1985. The Carolinaparakeet in Florida. Fla. Orni tho1 .Soc., Spec. Publ. No. 2. 65 pp.McKnight, J.S., D.D. Hook, O.G.Langdon, and R.L. Johnson. 1981.Flood to1 erance and re1 atedcharacteristics of trees of thebottom1 and forests of the southernUnited States. Pages 29-69 in J.R.Clark and J. Benforado, eds.Wetl ands of bottom1 and hardwoodforests. Developments inAgricultural and Managed-forest

Ecology, Vol. 11. El sevi erScientific,. New York.Meeter, D.A., R.J. Livingston, and G.Woodsum. 1979. Short and 1 bng-termhydrological cycles of theApal achicol a drainage system withappl ication to Gulf coastalpopulations. Pages 315-338 in R. J.Livingstone, ed. Ecol ogi calprocesses in coastal and marinesystems. Plenum Press, New York.Menzef , R.W., N.C. Hul ings, and R.R.Mathway. 1966. Oyster abundance inApalachicol a Bay, Florida, inrelation to biotic associationsinfluenced by salinity and otherfactors. Gulf Res. Rep. 2:73-96.Milliman, J.D., and K.O. Emery. 1968.Sea-level s during the past 35,000years, Science 162:1121-1123.Moler, P.E. 119851. Home range andseasonal activity of the easternindigo snake, Drymarchon coraiscouperi, in northern Florida. Fl a.Game Fresh Water Fish Comm. Unpubl.MS. 25pp.Monk, C.D. 1965. Southern mixedhardwood forest of north centralFlorida. Ecol . Monogr. 35:335-354.Monk, C.D. 1966. An ecological studyof hardwood swamps in north-centralFlorida. Ecology 47: 649-654.Monk, C.D., and J.T. McGinnis. 1966.Tree species diversity in six foresttypes in north central Florida. J.ECO~ . 54 : 341 -344.Nash, G.V. 1895. Notes on someFlorida plants. Bull. Torrey Bot.Club '22: 141-147.Nehrl ing, H, 1889, 1896. Our nativebirds of song and beauty. Brumder,Milwaukee. 2 vols.Nesbitt, S.A., W.M. Hetrick, L.E.Williams, Jr., and D.H. Austin.1977. Foods of the nine-bandedarmadillo in Florida. Proc. Annu.Conf. Southeast. Assoc. Fish Wild1 .Agencies 31 : 57-61.Olmsted, I., and L.L. Loope. 1984.PI ant communities of Ever91 adesNasionaI Park. Pages IG:-l84 ;- I.111 r.Gl eason, ed. Environments of southFl orida--past and present. MemoirI I, Miami Geological Soci ety , CoralGables, Fla.Oxford English Dictionary. 1933.Vol. V. Oxford University Press,Oxford, England.Pasquier, R.F. 1982. Whose birds arethey? Nat. Conservancy News32(4):12-15.Pearson, P.G. 1951. Mammals o f GulfHammock, Levy County, Florida. M. S.Thesis. University of F1 orida,Gainesville. 155 pp.Pearson, P.G. 1952. Observationsconcerning the life history andecology of the Florida wood rat,Neotoma f loridana f 7oridana (Ord) .J. Mammal. 333459-463.Pearson, P.G. 1953. A field study ofPerornyscus popul ations in Gul fHammock, Florida. Ecology 34: 199-207.Pearson, P.G. 1954. Mammals aF GulfHammock, Levy County, Florida. Am.Mid1 and Nat. 51:468-480.Pelton, M.R., L.E. Beeman, and D.C.Eagar. 1980. Den selection byblack bears in the Great SmokyMountains National Park. Pages 149-151 in C.J. Martinka and K.L.McArthur, eds. Bears--their biologyand management. Bear Biol . Assoc.Conf. Ser. 3. U.S. Gov. Print.Off., Washington, D.C.Ponnamperuma, F.N. 1984. Effects offlooding on soils. Pages 9-45 inT.T. Kozlowski, ed. Flooding andpl ant growth. Academic Press,Or1 ando, Fl a.Puri , H.S., and R.O. Vernon. 1964-Summary of the geology of Floridaand a guidebook to the classicexposures. Fla. Geol . Surv. Spec.Publ. 5, rev. 312 pp.

Puri, M,S., 3.W, Yon, and W.R.Oglesby. 1967. Geology of Dixieand Gilchrist Counties, Florida.Fla. Geol. burv. 6uii. 49. 155 pp.Putnam, J.A., G.M, Furnival, and J.S.McKnight. 1960. Management andinventory of southern hardwoods.U.S.D.A. For. Serv. Agric. Handb,No. 181. 102 pp.Riggs, S. R. 1984. Pal eoceanographicmodel of Neogene phosphoritedeposition, U. S. Atlanticcontinental margin. Science223:123-131.Rogers, L.L., and R.D. Applegate.1983. Dispersal of fruit seeds byblack bears. J. Mammal. 64:310-311.Rogers, S.G., T.E. Targett, and S.B.VanSant. 1984. Fi sh-nursery use inGeorgia sal t-marsh estuaries: theinfluence of springtime freshwaterconditions. Trans. Am. Fish Soc.113:595-606.Schnoes, R.S., and S.R. Humphrey.1987. Terrestrial plant andwild1 ife communities on phosphateminedlands in central Florida.Bull. Fla. State Mus., Biol. Sci.30:53-116.Shannon, C.E., and W. Weaver. 1949.The mathematical theory ofcommunication. University ofIllinois Press, Urbana. 125 pp.Sheridan, P., and R.J. Livingston.1979. Cyclic trophic re1 ationshipsof fishes in an unpolluted, riverdominatedestuary in north Florida.Pages 143-161 in R.J. Livingstone,ed. Ecological processes in coastaland marine systems. Plenum Press,New York.Short, H.L., and E.A. Epps, Jr. 1976.Nutrient qua1 i ty and digestibilityof seeds and fruits from southernforests. J. Wildl . Manage. 40:283-289.Shure, D.J., and M.R. Gottschalk.1985. Litter-fall patterns within afloodplain forest. Am. Midl. Nat.114:98-111.Shure, D.J., M.R. Gottschal k, and K.A.Parsons. 1986. Litterdecomposition processes in afloodplain forest. Am. Midl. Nat.115:314-327.Simons, R.W., and J.H. Hintermister.1984. Native plant and animalcommunities, northern phosphateresource district. Pages 76-81 inH. Hood and R.M. Palmer, eds.Phosphate mining in Florida, asource book. Florida Defenders ofthe Environment, Gainesvill e and theEnvironmental Servi ce Center,Tall ahassee.Simons, R.W., W.R. Marion, and J.H.Hintermister. 1984. Native plantand animal communities , central andsouth phosphate resource districts.Pages 52-58 in H. Hood and R.M.Palmer, eds. Phosphate mining inFlorida, a source book. FloridaDefenders of the Environment,Gai nesvi 11 e and the EnvironmentalService Center, Tall ahassee.Simons, R.W., S.W. Vince, and S.R.Humphrey. [1989]. Hydric hammock:a guide to management. U.S. FishWildl. Serv. In press.Sincl air, M., G.L. Bugden, C.L. Tang,J.C. Therriault, and P.A. Yeats.1986. Assessment of effects offreshwater runoff variabil i ty onfisheries production in coastalwaters. Pages 139-160 in S.Skreslet, ed. The role offreshwater outflow in coastal marineecosystems.Springer-Verl ag,Berl in.Skeate, S.T. 1987. Interactionsbetween birds and fruits in anorthern Florida hammock communi ty.ECO~ ogy 68: 297-309.Snedaker, S.C., and D.P. de Sylva.1977. Role of freshwater inestuarine ecosystems. Volume One:Summary. Southwest Fl a. WaterManage. Dist. Plann. Rep. No. 1977-2.Speake, D.W., J.A. McGlincy, and T.R.Colvin. 1978. Ecology andmanagement of the eastern indigo

snake in Georgia: a progressreport. Pages 64-73 in R.R. Odom?E! !.. L2nd2r~, 94:. Pr~r~odinycrare and endangered wildl i fesymposium, Ga. Dept. Nat. Res., GameFish Div. Tech. Bull. WL 4.Stringfield, V.T. , and H. E. LeGrand.1966. Hydrology of 1 imestoneterranes in the coastal plain of thesoutheastern United States. Geol.Soc. Am. Spec. Pap. 93. 46 pp.Swi~d~ll, !I E . 1949 Pl antcommunities and other factorsaffecting the deer and turkeypoptrlations in Gulf Hammock. M.S.Thesis. University of Florida,Gainesville. 150 pp.Tanner, J. T, 1942. The ivory - bi 11 edwoodpecker. Nat. Audubon Soc. Res.Rep. No. 1. 111 pp.Tanner, W.F.. and I.M. Smith. 1981.Natural environment. Pages 14-73 inE.A. Fernald, ed. Atlas of Florida.Florida State University,Tall ahassee.Teskey, R.O., and T.M. Hinckley.1977. Impact of water level changeson woody riparian and wetlandcomauni t ies . Vol . 2 : The southernforest region. U.S. Fish Wildl.Serv. FWS/OBS -77/59.Thompson, S.K.S. 1980. Hammockvegetation in the northern Gulftlanimock region of Florida. M.S.Thesis. Florida State University,Tallahassee. 49 pp.U.S. Fish and Wildlife Service. 1978.Osceol a National Forest phosphateextraction and processing: impactson federally 1 i sted threatened orendangered and other species ofspecial concern. National EcologyCenter, Gainesville, Fla. 414 pp.U.S. Fish and Wildlife Service.119801. St. Marks National Wildl ifeRefuge: forestry management andnon-game wildl ife. Unpubl . rep.119 pp.U.S. Fish and Wildlife Service. 1985.Endangered and threatened wi ldl ifeand plants; review of vertebratewildlife; notice of review. Fed.Reg. 50:37958-37967.U.S. Soil Conservation Service. 1960.Soil survey of Orange County,Florida. Washington, D.C.U.S. Soil Conservation Service. 1974.Soil survey of Brevard County,Florida. Washington, D.C.U.S. Soil Conservation Service. l!l//.Soil survey of Hernando County,Florida. Washington, D.C.U.S. Soil Conservation Service. 1979.Soil survey of Marion County,Florida. Washington, D.C.U.S. Soil Conservation Service. 1980.Soil survey of Volusia County,Florida. Washington, D.C.U. S. Soil Conservation Service. 1981.26 Ecological communities ofFlorida. Washington, D.C.Vernon, R.O. 1951. Geology of Citrusand Levy Counties, Florida. Fla.Geol. Surv. Bull. 33. 256 pp.Vernon, R.O., and H.S. Puri . 1965.Geologic map of Florida. Fla. Geol .Surv. Map Ser. No. 18.Watts, W.A. 1971. Postglacial andinterglacial vegetation history ofsouthern Georgia and centralFlorida. Ecology 52:676-689.Watts, W.A. 1980. The lateQuaternary vegetation history of thesoutheastern United States. Annu.Rev. Ecol. Syst. 11:387-409.Watts, W.A., and M. Stuiver. 1980.Late Wisconsin climate of northernFlorida and the origin of speciesrichdeciduous forest. Science210:325-327.Wells, B.W. 1942. Ecologicalproblems of the southeastern UnitedStates coastal plain. Bot. Rev.8:533-561.

Wharton, C.H. 1977. The naturalenvironments of Georgia. Office ofBi dilirifi$ R?:c:r~h, ~coYCJ;?.Department of Natural Resources,Atlanta, Ga. 227 pp.Wharbon, C.H., H.T. Odum, K. Ewel, M.Duever, A. Lugo, R. Boyt, J.Bartholomew, E. DeBellevue, S.Brown, M. Brown, and L. Duever.1977. Forested wet1 ands of Florida- their management and use. Centerfor Wetlands, University of Florida,Gainejvilf 2.Wharton, C.H., W.M. Kitchens, E.C.Pendleton, and T.W. Sipe. 1982.The ecol ogy of bottom1 and hardwoodswamps of the Southeast: acommunity profile. U.S. Fish Wildl.Serv. FWS/OBS-81/37. 133 pp.White, W.A. 1970. The geomorphologyof the Florida peninsula. Fla.Geoi . Sur-v. Bull. 51. 163 pp.Whitlow, T.H., and R.W. Harris. 1979.Flood tolerance in plants: a stateof-the-artreview. U.S. ArmyEngineers Waterways Experiment Stn.Tech. Rep. E-79-2. Vicksburg, Miss.161 pp.Williams, L.E., Jr. 1978. Floridablack bear. Pages 23-25 in J.N.iayne, ed. F.idrntedl s . Rdre aridendangered biota of Florida, Vol. 1.University Presses of Florida,Gainesvil le.Winker, C.D., and J.D. Howard. 1977.Pl io-Pl ei stocene pal eogeography ofthe Florida gulf coast interpretedfrom re1 i ct shore1 i nes. Trans. GulfCoast Assoc. Geol . Soc. 27:409-420.Wirtz, W.O., D.H. Austin, and G.W.Delke. 1985. Food habits of thecommon 1 ong-nosed armad i 11 o Dasypusnovemcinctus in Florida, 1960-61.Pages 439-451 in G.G. Rontgomery,ed. The evolution and ecology ofarmadillos, sloths, andvermi 1 i nguas .SmithsonianInstitution Press, Washington D.C.451 pp.Wood, G.W., and R.E. Brenneman. 1980.Feral hog movements and habitat usein coastal South Carolina. J.Wildl . Manage. 44:420-426.Wood, G.W., and D.N. Roark. 1980.Food habits of feral hog in coastalSouth Carolina. J. Wildl. Manage.44:506-511.Woolfenden, G.E. 1967. Live oakcabbagepalm hammock. Audubon FieldNotes 21:635-637.Woolfenden, G.E. 1968. Live oakcabbagcpal n hazmock. Audubon FieldNotes 22:488.

PAGE4. Title and SubtitleThe Ecology of Hydric Hammocks:a Community Profile2. 3. Rac~piant's Accesston No.I7. Author(s)Susan W. Vince, Stephen R. Humphrey, and Robert W. Simons9. Performing Organization Name and AddrassAuthor Affiliation:Florida Museum of Natural HistoryUniversity of Florida--I8. Psrtormlng Organization Rept. No.10. Pmiect/Task/Work Unit No.t11. Contract(C) or Grant(G) No.(C)t ~ S ~ i r i Organiratlon n e Name and AddressU.S. Department of the InteriorFish and Wildlife ServiceNational Wet1 and Research CenterWashington. DC 2024015. supplementary Notes16. Abstract (Limit: 200 words)This publication describes the nature and dynamics of a type of forestedwetland called hydric hammock. Hydric hammock has not been carefullydefined or studied, and it often is ignored or lumped with other mixedhardwood forests. Hydric hammock occurs at low elevations along the gulfcoast of Florida from Aripeka to St. Marks and at various inland sites inFlorida. The largest contiguous tracts occur along the gulf coast and theSt. Johns River. Sites typically are gentle slopes between mesic hammockor pine flatwoods and river swamp, wet prairie, or marsh. Hydric hammocksoccur on soils that are poorly drained or have high water tables, andoccasional flooding saturates the soils long enough to restrict or modifyspecies composition. Hydric hammocks typically contain live and swamp1 aurel oaks (Quercus virginiana and Qoercus 1 aurifo 1 ia) , cabbage palm(Sabal pa7metto), southern red-cedar (Juniperus silicicola), sweetgum(Liquidambar styraciflua), and hornbeam (Carpinus caroliniana). Hydrichammocks lack certain tree species while sharing others with both upslopemesic forests and downslope swamp forests. The vertebrate fauna containsno unique species but has a high diversity of species and high abundanceof selected species. Hydric hammocks are particularly important habitatfor game species and for overwintering passerine birds.17. Document Analysis 8. Deacdptonforested wet1 and 1 imestone l inkageFl orida fire plant-animal interactionsinundationgame an i ma1 swater tablepasserine birds(G)!ic. COSATI FitldlGroup16 Ava~labllity StatementUnlimited distribution19. Securrty Class CThas Report)Unclassified20. Seeurtty Class Vhis Page)Unclassified21. No of Pagesix + 81n. Pnce(See ANSI-239.18) OPTIONAL FORM 272 (4-77)(Fomnrly NTIS-35)