Florida Scientist, QUARTERLY JOURNAL of the FLORIDA ACADEMY OF SCIENCES VOL 51-3-4-1988

com-The neurological activity of crude PBTX was enhanced by HPLC purification.. , 1982 Dueto thehealth dangers associatedwith PBTX, it is of interest toknow the actualcompound responsibl

Pocket Gophers (Geomyspinetis) ina SandhillCommunity

CyndiA Gates and George W Tanner 129

Toxin AssociatedwithRespiratoryProblems

MiquelangeloJ.Perez-Cruet,JosephJ. Krzanowski

and DeanF.Martin 140

gosypinus anastasae)May BeExtinct

Stephen R Humphrey, William H Kern,Jr.,

and MarkE.Ludlow 150

A Modified Live TrapfortheCaptureofSoutheasternPocket

Gophers

CyndiA Gates, George W. Tanner,and BrianK Gates 156

WeightinBobwhite Quail(Colinus virginianus)

P. A.Skewes, H. R Wilson, andF. B Mather 159

YoungSnook, Centropomusundecimalis, inAquaria

Mary Higby andAlfred Beulig 163

Archaeological TestingattheAndersSite: A Weeden

Island-RelatedMidden on Boca CiegaBay, St.Petersburg, Florida

RobertJ.Austin 172 RangeExtensionofIpomoeaturbinataLag (Convulvulaceae) to

SouthernFlorida

DanielF. Austinand RichardK Jansson 182

Fungal FilamentsinMilleporaComplanata Lamarck, 1816

(Cnidaria: Hydrozoa) AfterMass ExpulsionofZooxanthellae

Volume

FLORIDA SCIENTIST

Quarterly JournaloftheFloridaAcademyofSciences

Copyright ©bytheFloridaAcademyofSciences, Inc 1988

Editor: Dr DeanF Martin Co-Editor: Mrs BarbaraB Martin

Chemical andEnvironmentalManagement Services (CHEMS) Center

DepartmentofChemistryUniversityofSouth Florida

Tampa, Florida33620

TheFloridaScientistispublished quarterlybytheFloridaAcademyofSciences,Inc., a non-profit scientific andeducational association Membership isopen toindi-

viduals or institutions interested in supportingscience in its broadest sense

Applica-tionsmaybe obtainedfromtheExecutiveSecretary.Both individualandinstitutional

membersreceive asubscriptiontotheFloridaScientist Direct subscriptionis

avail-ableat$20.00 per calendaryear

arewelcomed in anyfield of Science as represented by the sections of theAcademy,

viz., Biological Sciences, Conservation, Earth and Planetary Sciences, Medical

Sci-ences, Physical Sciences, Science Teaching, and Social Sciences Also, contributions

willbe consideredwhichpresentnewapplications ofscientificknowledgetopractical

problems within fieldsofinterest to theAcademy. Articles must not duplicatein any

substantial way material that is published elsewhere Contributions are acceptedonlyfrom membersoftheAcademyandsopapers submittedbynon-memberswillbe

manuscripts are inside theback cover

FLORIDA ACADEMY OF SCIENCES

Founded 1936

President:Dr.MarvinL.Ivey

HammondandHolman,P.A.

Suites 11and12, HillBuilding

4930Park Boulevard

PinellasPark,FL34665

President-Elect:Dr.Ernest D.Estevez

MoteMarineLaboratory

1600 CityIslandPark

Sarasota,FL33577

Secretary:Dr.PatrickJ.Gleason

1131NorthPalmway

Lake Worth,Florida33460

Treasurer:Dr.AnthonyF.Walsh

5636SatelDriveOrlando, Florida32810

Executive Secretary

Dr.AlexanderDickison

Departmentof Physicial Sciences

SeminoleCommunityCollegeSanford,FL32771

ProgramChairs:Dr GeorgeM Dooris

Dr PatriciaM Dooris

P.O.Box2378

St.Leo, Florida33574

PublishedbytheFloridaAcademyof Sciences,Inc.

810EastRollins Street

Printedbythe Storter PrintingCompany

Florida Scientist

QUARTERLY JOURNAL OF THE FLORIDA ACADEMY OF SCIENCES

DeanF. Martin, Editor BarbaraB Martin, Co-Editor

Volume51 Summer/ Autumn, 1988 Numbers3/4

BiologicalSciences

CyndiA Gates and George W Tanner

DepartmentofWildlifeandRangeSciences,Universityof Florida, Gainesville,FL32611Abstract:Effectsof season of prescribedburningandtime elapsedsince lastburn onherba-ceousbiomassandspeciescomposition, rootcarbohydrateandnitrogencontent, andsoutheast-

ernpocketgopher(Geomyspinetis) moundingactivityandbodyconditionwere examined ona sandhillcommunity Treatments includedlate-winterandlate-summerburnsin1983,1984, and

1985,and an unburnedcontrol.Post-burndatacollected inthefallof1985showednosignificant differences(P>0.05)fortotalbiomass ofabovegroundherbaceousvegetation.Abovegroundbio-

massofgrasseswashigherthanforbbiomass forall treatments. Forbbiomasswaslowestonthe

unburned control. Shannon diversity indices were consistently lower on unburned than onburnedtreatments. Carbohydrate levels in roots ofselectedherbaceousspecies collectedfrom

each treatmentin thewinter of1986werevariableandnosignificant differences(P>0.05) were

detected. Nitrogenlevels in rootstendedto behigheronunburnedthanon burnedtreatments.

Moundingactivity variedwidely bothamongand within treatmentsandrangedfrom0 73 to

5.19 moundsproducedperday duringthe winter of 1986 Nosignificant correlationbetweenmoundingactivityandcarbohydrate or nitrogenlevelsofselected plantswasdetected.Nosignifi-

cantdifferences(P>0.05)weredetectedinweight:bodylengthratiosof pocket gophersamong

treatmentsduringlatewinter-earlyspring.

Fire plays a major role in maintaining Florida sandhill communities oflongleafpine (Pinuspalustris) andturkeyoak (Quercuslaevis). Increasedfire

protection over the years has resulted in succession of these xeric, park-like

savannas toward hardwood hammock or xeric hardwood-mixed pine

com-munities, with a decrease in herbaceous species diversity (Laessle, 1942,1958; Vogl, 1972;Myers, 1985)

Gre-len and Epps, 1967; Hulbert, 1969) Litterremoval also alters the

microcli-mate of emerging vegetation through warmer soil temperatures and

in-creasedlight which leads to rapid growth of grasses (Old, 1969; Peet et al.,

1975)

Changes in species composition frequently accompany burning Arata

more common on burned than unburnedsites in a longleafpine-turkey oak

and legumes had greater biomass andspecies richnesson one-year-oldburns than ontwo-,three-,orfour-year-oldburns (Harlow andBielling, 1962).Nutritional status of plants also is affected by fire. Chapin (1981) noted

that new growth of species that resprout immediatelyfollowing fire is

sup-ported bystoredcarbohydratereserves and anycurrent photosynthate Ona

wetflatwoods sitein south Florida, higher total nonstructural carbohydrate (TNC)levelswere found in creepingbluestem (Schizachyriumstoloniferum)

littlethroughoutthegrowingseasonwhilesimilarparts of plantsfrom areas

burnedtheprevious springshowed significantdecreasesin percentN during

themiddleofthegrowingseason (LeukelandStokes, 1940)

A native herbivore in Florida's sandhill communities is the southeastern

nota-tion), dogfennel (Eupatoriumspp), lambsquarters (Chenopodiumspp), and

false moneywort (Alysicarpus vaginalis) in food caches ofthis species rington (1940) foundthat captiveanimals consumednutgrass (Cyperusescu-

Sporobolus spp) Burrowing activity, indicated by the presence of soil

mounds, variesseasonallyduetochangesinfeedinghabits, breedingactivity,

Hick-man and Brown, 1973) Mounds produced by Geomysattwateriwerelarger

on burned areaswheredicots (preferredfoodplants) had decreased, as pared to unburned areas (Spencer et al., 1985) They suggested that this

com-indicated increased foraging activity Tilman (1983) estimated responses of

G. bursariusto fertilizertreatmentson rangegrassesby conductingcountsof

newly-produced mounds and foundpositiverelationshipsbetweentotalplant

biomass and moundingactivity

This studywasdesignedtofurtherinvestigatethepremisethatprescribed

ofherbaceousplantsandthat thesevegetative responseswill affectbody

DescriptionofStudy Area—Thestudywas conducted on asandhillcommunitysite intheKatherine OrdwayPreserve-Swisher Memorial Sanctuary, Putnam County, Florida. Longleafpine-turkeyoaksavannasmakeupabout one-thirdofthe3,640haPreserve (Humphreyet al.,

1985). Major components of the herbaceous vegetation include pineland threeawn (Aristida

stricta)and pineywoodsdropseed (Sporobolusjunceus), aswellasanumberofbluestemgrasses

(AndropogonsppandSchizachyriumspp)anda variety of forbs Soil typesincludeCandlerfine

sand,Apopkasand,andAstatulafinesand(PutnamCountySoil Survey,unpublished)

Loggingof pinesandfrequentwinterfires that killedturpentinedtrees prior to1970 allowedturkeyoakto increase considerably Prior to 1974,sandhill areas inthe western portionofthe

Nos.3/4, 1988] gatesand tanner —effectsofprescribedburning 131the easternportionwere burned annually (T Perry, 1985). ThePreservewasleased for cattle

grazingprior to 1979. In 1980, the UniversityofFloridaacquired the property; subsequently, cattlegrazingwasexcludedandfirefrequencyaltered(Humphreyet al ,1985)

Methods—Vegetation parameters and pocket gopher activity were measured on six scribedburntreatments;late-winterburnsinMarchor early Apriland late-summerburnsin late

pre-July,August,or earlySeptemberof1983,1984,and1985.Twounburnedareasservedas controls.

Onecontrolareahad been burnedevery 3to4yearsandtheotherburnedannuallyprior to 1974.

There were tworeplications ofburns inwinter 1984, winter 1985, andsummer 1985. The 11

studysitesrangedin sizefrom25hato170ha Sitesburnedduring wintersof1983, 1984,or1985

aredesignatedasW83,W84, orW85,respectively. Treatmentsitesburnedduring thesummers

of 1983, 1984,or1985aredesignatedS83, S84, orS85,respectively. Unburnedcontrol sites are

designatedCONTROL.

Pre-andpost-burnherbaceousspeciescompositionand abovegroundbiomasswereestimatedusing the quadrat-clip technique (National Academy of Sciences-National Research Council,

1986). Pre-burnmeasurements weretaken1 to2monthsprior toburning on W85 (W85Aand

W85B)andS85(S85A andS85B)sites.Ninety-nine0.5-m2quadratswereplacedat25mintervals

alongthreerandomlylocatedtransects of800mlengthonsitesW85A,S85A, andS85B; 33

0.5-m2quadratswereplaced alongonetransect of800mlengthonsiteW85B.Post-burnherbaceous

speciescompositionand abovegroundbiomasswereestimatedduringOctober through

Decem-ber1986 Sixty0.5-m2

quadratswereplacedat25m intervalsalongrandomlylocatedtransects

withineach studysite. Vegetationwashand-clipped atgroundlevel, separatedbyspecies andweighedinthefield.Aportionofthesamplesfromeachsitewasoven-driedat70°Cfor72 hours

and re-weighedto establish dryweightconversions. Thenumberofsamplescollected for

dry-weight conversions was dependent on variability of weather conditions during thesampling

period.TheShannonindexwasusedasthemeasureof diversity(ShannonandWeaver,1949).

Moundcountsweremadetodeterminerelativemoundingactivity levels ofpocket gophers(Tilman,1983).Three, square 1.0-haquadratswereestablishedoneachofthe11studysites.The

size ofquadratselectedwasbased onBrown and Hickman's (1973)estimateofaverage tunnelsystem lengthinsouthFlorida.InSeptember1985,newmoundsshowinglittle ornoevidenceof

weatheringwereselected as center points foreach quadrat ContinuingthroughSeptemberand

partofOctober,all existingpocketgophermoundswithin thequadratswereflaggedandthedaterecorded CountsofnewmoundsweremadeinDecemberandJanuary, monthsgophersprevi- ouslyhad beenreportedtobemostactive, to serve asanindexof activity(Hickman and Brown,

1973).Anindexofmoundingactivity ineachquadratwascalculatedbydividing thenumberof

newmoundscountedin32to39dayintervalsduringDecemberand January bythenumberof

dayssincetheprevious count

From lateJanuary to early March 1986, belowground parts of six plant species, pinelandthreeawn, pineywoodsdropseed, creeping bluestrem, grassleaf goldaster(Heterothecagramini-

folia), falsedeertongue(Carphephoruscorymbosus),andwildbuckwheat (Eriogonumsum), werecollected inthepocketgophermoundactivityquadratsforcarbohydrateandnitro-

tomento-genanalyses.Belowgroundplantparts, hereinafter referred to as roots,includedroots,rhizomes,

and belowgroundcrownportions ofsomespecies.Rootsof at least five plants ofeachspecieswere

collectedand combinedto give acompositesampleofeachspecies foreach quadrat Extractionprocedures fordetermining carbohydrate content ofthe root samples followed those ofRowe

(1980) withminormodifications (Gates, 1986). Microkjeldahldigestion of root tissues followedNelson and Sommers (1973) and ammonium in the digest was determined using proceduresoutlinedbyBremner andEdwards(1965).

A minimumofthreepocket gophers wascapturedineach ofthe 11 studysites between 2

Februaryand9April 1986 using a modifiedversion oftheShermanpocket gophertrap

(Sher-man,1941). Capturedanimalswereweighed,sexed,andbodylengthmeasurementstakenprior

to releasebackintothetunnel system.Bodyweight-to-body lengthratioswereusedasanindexof condition.

Data wereanalyzed using a generallinearmodel(GLM)for leastsquaresregression cal Analysis Institute, 1982).When GLMprocedures detectedsignificant differences(P<0.05),

(Statisti-t-tests of least significant differences (LSD) were carriedouttodeterminedifferencesbetween

specific treatments. Contrastanalyses (StatisticalAnalysis Institute, 1982) were usedto detect

seasonandyeareffectsandseasonbyyearinteractions.Vegetation datawerelog-transformedfor

Results and Discussion — Herbaceous Vegetation—Fifty-two

threeawn, piney woods dropseed, and purple threeawn (Aristida

common broadleaved species Overall, pineland threeawn had the highest

frequencyofoccurrence (73.3%) andthe greatest mean biomass (35.2 g/m2

)

(Ta-ble 1). However, no trend in forb frequency was apparent among the burn

than on unburned plots but differences were less dramatic than for grass

biomass among burn treatments (Table 1). Although biomass generally was lower in the CONTROL for both plant groups, increases in forb biomass wereproportionatelymuchhigherinburntreatmentsascomparedtooverall

consumed mostofthat year'sgrowth.

Table1 Frequencyofoccurrence(%)and abovegroundbiomass(g/m2 of forbsandgrasses

onsandhillcommunitytreatments,Ordway-SwisherPreserve,PutnamCounty,Florida.

sum-procedures detected nosignificant differences amongtreatments for forb or

forestswashigherwhen burnedinMarchorAprilthaninJuneorAugust but

thatforbproduction wasgreatestfollowingAugustburning

Forbs generally respond rapidly to fire. Annual species have reduced

competition from perennials forsunlight, at least temporarily, andlitter

re-moval provides a more favorable seedbed (Hodgkins, 1958; Vogl, 1972).Plantswith wind-dispersedseed, such as composites, quickly invade and in-

while someforbs,including anumberoflegumes,benefitfromseed

scarifica-few annual

Nos 3/4, 1988] gatesand tanner —effectsofprescribedburning 133

Pineland threeawn was the dominant herbaceous species present,

Statisticalcomparisonsofpineland threeawn biomass amongthe treatments,

however, showed nosignificantdifferences(Table2) Greatestbiomassofthis

respec-tively Harlow and Bielling (1962) reported that wiregrasses (Aristida spp)

reached a peak in abundance 3 years after burning It is of interest that

pineland threeawn hadgreaterbiomass at8and 20 monthspost-burninthis

effect, highvariationobscured anysignificance in differences

ences between thesesites in post-burn vegetationbiomass may be primarily

from eachotherprior toburning with S85B havinggreatermean herbaceous biomass pre-burn and post-burn Interestingly, burning reduced herbaceous biomass on both S85sitesbyca50 %.

Table3. Comparisonofpre-burnandpost-burnmeansofabovegroundherbaceousbiomass(g/m2

) forW85andS85 studysites,Ordway-SwisherPreserve,PutnamCounty,Florida.

* Indicates post-burnmeanis significantly different (P<0.05)fromits respective pre-burnmean;meanswith

thesamesubscript letter are not significantly different.

Post-burn diversityindices usingboth biomass and frequency data from winter and summer treatments of 1985 were larger than pre-burn indices(Table4) Thisindicatesthatdiversitytendstoincreaseshortly afterburning

on biomass and frequency data

134 FLORIDA SCIENTIST [Vol.51

ments and allyears indicate that diversitywasgreater on burnedareas paredtotheunburned CONTROL(Table5)

com-Table4. Comparisonofpre-burnandpost-burnShannondiversity indices forbiomassand

frequencyofoccurrenceofherbaceousplants in1985ona sandhillcommunity, Ordway-Swisher

Preserve,PutnamCounty,Florida.

Other authors also have reported increased diversity due to fire (Vogl,

1972, Shafiand Yarranton, 1973; Mooreet al., 1982) Diversity increasedinturkeyoaksandhillsprotectedfromfirefor40years (Veno, 1976) Thegreat-

estchangein diversitylevels wasattributedto anincreaseinmesic hammockspecies Veno concluded that with long-term protectionfrom fire, anumber

of fire-intolerant speciescould invadethesesites.

Table5.Post-burnShannondiversity indices forbiomassandfrequencyofherbaceousplants

onsandhillcommunitytreatments,Ordway-SwisherPreserve,PutnamCounty,Florida.

Table6 Totalnonstructuralcarbohydrate(TNC)levels(mg/g)in roots of selectedherbaceous

species,Ordway-SwisherPreserve,PutnamCounty,Florida.

Control

YearofBurn SeasonofBurn

Species 1983 1984 1985 Winter Summer

Grassleafgoldaster 147 152a 131a 160a 145a 153a

Falsedeertongue 142 153a 148a 176a 159a 163a

* Indicates yearly or seasonalmeansignificantly differentfromcontrol; for a given species,meanswith the

samesubscript letter are not significantly different.

(TNC) concentrationsinmostspeciesgenerallywerenotaffectedby burning when compared tothe CONTROL (Table 6). TNC generallywas higher in

TNC valuesforpineland threeawn were significantlylower for all burn

Nos.3/4, 1988] gatesand tanner —effectsofprescribedburning 135

burn-ing TNC washigherwith winterburning, which wasa generaltrend forall

grasses TNC in forbstendedtobe somewhathigherwith summerburns It is

than anyotherspecies

discern-ible However, defoliation through burning may have differential effects

withinthe relatively shorttimeperiodssinceburning Trlica and co-workers

west-ernrangespecies Somespecies, such asfringedsagewort (Artemisiafrigida)

and rabbitbrush (Chrysothamnusvicidiflorus), showed lower TNC levels in

microsite ormacrosite conditionsatthetimeofsampling

Table7. Nitrogenlevels (%) in roots of selectedherbaceousspecies, Ordway-Swisher serve,PutnamCounty,Florida.

Pre-Control

YearofBurn SeasonofBurn

Species 1983 1984 1985 Winter Summer

Pinelandthreeawn 0.45 0.30a* 0.31a

*

0.34a

*0.33a* 0.31a

Grassleafgoldaster 1.15 0.76a 0.70a 0.83a 0.69a 0.87aFalsedeertongue 0.99 0.64a 0.58a 0.59a 0.48a* 0.72b

indicates yearly or seasonalmeansignificantly differentfromcontrol; for a given species,meanswith the

samesubscript letter are not significantly different.

Nitrogen Concentrations — Average nitrogen (N) levels were somewhat lowerfor all species after burning (Table 7). Differences generallywere not

treatment means for the 3 years combined were significantly less than the

CONTROL also PercentN infalsedeertongue wassignificantlyhigherwith summer burns than with winter Pineland threeawn and pineywoods drop-

Nlevelsthanforbs

Woods and co-workers (1959)studied seasonal variationinnitrogenlevels

ofpineland threeawnroots insandhillcommunitiesinwesternFloridawhere

levelswere about0.48% duringthemonthsofJanuary, February and March.

This was comparable to the 0.45% N levels for pineland threeawn in our

CONTROL undersimilar conditions Old (1969) reported thatN content of

lower thaninunburnedplots andattributedthis to rootN depletionby idlygrowing abovegroundvegetation Leukel andStokes (1940)foundthatN

rap-percentage of roots of Florida range grasses decreased during the middle ofthe growing season in areas burned the previous spring, with percentages

pocket gophersonsandhillcommunitytreatments,Ordway-SwisherPreserve, PutnamCounty,

usuallywas greater on burned treatments except for the 1985 winter burn

Hickman and Brown (1973) notedseasonal variationinmoundingactivityof

individual variationinnumberofmounds produced andstated that"patterns

ofmound production wereentirely irregular,varyingfrom oneportionofthe

samesystemtotheother."Thusthedifferences inmoundingactivityobserved within our quadrats and treatments may have been dueto variation among

individualanimals

To determine ifa relationship existedbetweenplantnutritionallevels and pocket gopher mounding activity, scatter diagrams of carbohydrate values

num-berofmounds per day per quadrat No relationship between mounding

ac-tivityandnutrientlevelswasapparent

Pocket Gopher Body Condition — Threeto fiveindividuals were trapped

on eachstudysitebetween2February and9 Aprilof1986 Ofthe45animalstrapped,30 werefemales and 15weremales Weights ranged from 90to 265

gand bodylengthranged from 14.0cmto20.3cm Average weight and body

length ofmalepocket gophers was 207 g and 18.0 cm, respectively Females had an average weightof139gand average bodylengthof16.8cm.

To compensate for differences due to sex and age, a ratio of weight to

treatments (Table9). There were nosignificant differences in meanweight:

bodylengthratios amongtreatments,years, orseasons, although ratiosfrom winter burns wereslightlyhigherthan from summer.

per treatment However, it is also possible thatdespitetreatments these

ani-mals areabletoselecthome rangeswithineachsitethat allowthem tomeet

Determination numbers pocket gophers

Meanratio Numbercaptured

Table9. Meanweight:bodylengthratios(g/cm)andnumberofpocket gophers capturedon

sandhillcommunitytreatments,Ordway-SwisherPreserve, PutnamCounty,Florida,between2

Februaryand9 April 1986.

*M/F=Numberof malesandnumberof females.

on eachsitemay morepreciselyindicate qualityofthe habitat forG. pinetis

underdifferentburnregimes

Conclusions —Differences among burn treatments were difficult to

following fire with no major shifts in frequency of occurrence or biomass

biomass andhigher frequenciesofoccurrenceofforbsandgrasses Each

mea-sure ofresponse toburning generallyexceeded the level on plots with

long-termprotectionfrom burning It ispossiblethatsomeactivelygrowingplant

All burn treatments had greater diversity than the unburned control

Fur-thermore, forbswere more diverseon recently burned areas as compared to

unburned sites or sites burned 2 years earlier Impacts of burning on root

carbohydrate and nitrogen levels were minor While periodic prescribed

burning has beneficial effects on some sandhill plant species, resulting inincreasedbiomass, no apparenteffects, either positiveornegative, ofburning

on pocket gophers were observed at the successional stages examined From

information gatheredin this study, itappears thatpocket gophersmay have a

fairlywide range oftolerancetopresent vegetative conditionsinsandhill

unim-peded toward ahardwood hammock community, leadingto a gradual loss of

Acknowledgments— We thank Brian Gates, John Wood, Candace Cantlin, and Mike

Drummondfor their assistance in fieldandlaboratory.WethankCliffLewisand Dick Franzfor

reviewing the manuscript.TheNatureConservancyprovidedpartialfundingfor this research.

LITERATURE CITED

Arata,A A 1959 Effects ofburning onvegetationandrodent populationsinalongleaf

pine-turkeyoak

Barrington,B A., Jr 1940.Thenaturalhistory ofpocket gophers M.S Thesis,Univ.Florida, Gainesville,FL.

Bremner,J.M, andA R.Edwards 1965.Determinationandisotope-ratio analysis of different

formsofnitrogenin soils: I.Apparatusandprocedurefor distillationanddetermination

ofammonium.Soil Sci Soc Proc.1965, pp 504-507

Brown,L. N., and G C Hickman 1973.Tunnel systemstructure ofthe southeasternpocket

gopher.FloridaScient.36:97-103

Chapin,F S., III 1981.Plantnutrientabsorptionandretentionunderdiffering fire regimes In: FireRegimesand EcosystemProperties(Mooney, H.A., T.M.Bonnicksen,N.L.Chris-tensen,J.E.Lotan,andW.A Reiners).USDAForest ServiceGen Tech.ReportWO-

26.

Ewel,K C 1970.Annualreproductivecycle ofthemalepocketgopher(Geomyspinetis). Ph

D.Dissertation,Univ.Florida, Gainesville,FL

Garren, K H 1943 Effects of fireonvegetationofthesoutheasternUnited States Bot. Rev.9:617-654

Gates,C A 1986 Effects ofprescribedburningonherbaceous vegetationandthesoutheasternpocket gopher, Geomyspinetis, in sandhill communities M S Thesis, Univ Florida, Gainesville, Fl.

Grelen,H E., andE A Epps 1967.Herbageresponses to fireandlitterremovalonsouthernbluestemrange.J.RangeMgt 20:403-404

Harlow, R.F, andP. Bielling 1962. Controlledburningstudies in longleafpine-turkeyoak

associationsontheOcalaNationalForest Proc FifteenthAnn S. E Assoc. Gameand

1958.Areportonsuccession studies of selectedplantcommunitiesintheUniversity of

FloridaConservationReserve,Welaka,Florida.Quart.J.Fla.Acad.Sci. 21:101-112

Lemon,P. C 1949 Successional responses ofherbsin longleaf-slashpineforests after fire.

Ecol-ogy.30:135-145

Leukel, W A., and W E.Stokes 1940. Growth, behavior, andrelativecompositionofrange

grasses as affectedby burningandtheeffect ofburningon maintenanceofnaturalgrass stands and upon the establishment of improved grasses. Florida Agr. Exp Stn. Ann

Report,1940.pp.50-51.

Lewis,J. B., D A Murphy, andJ.Ehrenreich 1965 Effects ofburningdates onvegetative

productiononOzarkforests Proc S. E.GameandFishComm.18:63-72.

McNab, B. K 1966. Themetabolism of fossorial rodents: a studyof convergence Ecology.47:712-733

Miller, R S., and H E Bond 1960. The summerburrowingactivity ofpocket gophers. J.

National Academy of Sciences-National Research Council 1986. Range Research: Basic

ProblemsandTechniques.Publ 890.Washington,D.C.336p.

Nelson, D W, andC E Sommers 1973. Determinationof totalnitrogeninplantmaterial.

AgronomyJ.65:109-111

Old, S 1969. Microclimate, fire and plant production in an Illinois prairie Ecol. Monogr

Nos.3/4, 1988] gatesand tanner —effectsofprescribedburning 139

Peet, M.,R.Anderson,and M.S. Adams 1975 Effect of fireonbigbluestem production.Am.

Midi.Nat 94:15-25

PerryT 1985 {CatherineOrdwayPreserve-SwisherMemorialSanctuary,pers.commun

Ross,J.P 1976. Seasonalenergy budgetsofafossorialrodentGeomyspinetis. Ph D

Disserta-tion,Univ.Florida, Gainesville,FL

Rowe, M.1980.Theeffect ofparaquattreatmentonthestarchcontentof loblollypineseedlings (PinustaedaL.).Mastersthesis,Univ.Tennessee, Knoxville,TN.57p.

Shafi, M I., and G A Yarranton 1973 Diversity, floristic richness and species evennessduring a secondary(post-fire) succession.Ecology 54:897-902

Shannon, C E.,and W Weaver 1949.TheMathematicalTheoryofCommunications.Univ

Illinois Press,Urbana,IL.

Sherman, H.B 1941.Aboxtrapforthecaptureof liveGeomys.J.Mamm.22:182-184.Spencer, S R., G N Cameron, B. D Eshelman, L. C Cooper, andL R Williams 1985.

Influenceofpocketgophermoundsona Texascoastal prairie.Oecologia 66:111-115.StatisticalAnalysisInstitute.1982.SASUser'sGuide:Statistics.SASInstitute,Cary,NC

Stoddard, H.L., Sr 1931.TheBobwhiteQuail: ItsHabits,PreservationandIncrease. Charles

Scribner's Sons,NewYork,NewYork.559p.

Tilman, D 1983. Plant succession and gopher disturbance along an experimental gradient.

Vogl, R.J.1972 Fire insoutheasterngrasslands TallTimbersFire Ecol.Conf.No 12:175-198

Weaver, J. E., and N W Rowland 1952 Effects of excessive natural mulch development,

yield,andstructure of native grassland Bot.Gaz 114:1-19.

Woods,F. W., H L. Harris,andR E Caldwell 1959. Monthlyvariations ofcarbohydrates

andnitrogenin roots of sandhilloaksandwiregrass.Ecology 40:292-295

FloridaSci 51(3/4):129-139 1988

Accepted: January9, 1988

PURIFICATION AND ACTIVITY OF

ASSOCIATED WITH RESPIRATORY PRORLEMS

MlQUELANGELOJ.PEREZ-CRUET1

JOSEPHJ. KRZANOWSKI2

and DeanF. Martin1

iChemicaland EnvironmentalManagementServices(CHEMS) Center,DepartmentChemistry,

UniversityofSouthFlorida,Tampa,Florida33620

2DepartmentofPharmacologyandTherapeutics,University ofSouthFlorida,

CollegeofMedicine.Tampa,FL33612

Abstract: Thisstudy wasinitiated to investigate the compound(s) ofPtychodiscus brevis toxin (PBTX) that causes contraction of canine trachealsmooth muscle The components of

PBTXwereseparatedby reverse-phasehighperformanceliquidchromatography(HPLC)using

achloroform-methanollinear solventgradient system PurifiedPBTXfractionsweretestedfor

neurological activity (asmeasured bycaninetrachealsmoothmusclecontraction)andhemolytic

activity (as measured bykineticsof hemolysisofcaninebloodcells in ablood-buffersolution).

Chemical analysiswas conductedto determinethe structure of the neurologically activepoundofPBTX

com-The neurological activity of crude PBTX was enhanced by HPLC purification. A 6-fold increase in neurological activitywasobservedafter the firstHPLCpurification, while a 75-fold increase inneurologicalactivity was demonstratedafter thesecondHPLCpurification ascom-pared to crudePBTX Three primarypeakswere observed in the HPLC chromatogram: one(PBTX-2a) had neurological activity, a second (PBTX-1) hadhemolytic activity, and a third

(PBTX-2b) did not have neurologicorhemolytic activity In addition, tachyphylaxis (rapidly

developingtolerance)occurredincaninetrachealsmoothmusclerepetitivelyexposedto

PBTX-2a, whichwas foundtohavea molecularweight ofabout 900daltonsand anultraviolet

absorp-tion maximumof208nm Thesechemicalcharacteristicswereobserved foraneleven member

transfused ether ringcompoundthatwasneurologically activeandwhosestructurewas

eluci-datedbypreviousresearchers.

Overblooms ofplanktonic organisms (redtides) can occur when optimal growth conditions exist in the marine environment. Red-tide blooms in the

GulfofMexico have beenassociatedwithlargefishkills forat least 100years(Walker, 1884) In 1948, Davis identified Gymnodinium brevis as the caus-

Steidinger(1979) laternoticedanindentationonthethecalridgeofthe

blooms, sudden appearanceoflargenumbersofPtychodiscusbrevis, impart

a reddish tint to the seawater, thus the name red tide The red color seen

during bloomsofPtychodiscusbreviscouldbe caused byabacterium (Evans

etal.,1973)

Ptychodiscus brevis toxin (PBTX) poses serioushealth problems during a

red tide outbreak when PBTX becomes aerosolized in the sea-spray

nonproduc-tivecough, sneezing,andrhinorrhea, which mayleadto difficulty in

Nos.3/4, 1988] perez-cruet etal.—activityof toxin 141

shown in a clinicalsurvey where 12 out of 15 asthmatic patients exposedto

PBTX had anasthmatic attack (Asaietal , 1982)

Dueto thehealth dangers associatedwith PBTX, it is of interest toknow

the actualcompound responsibleforthe asthma-likesymptoms, particularlybronchoconstriction

Researchhasbeen undertakento determinethe physiological responseto

crude and purified Ptychodiscusbrevis toxin in various in-vitro and in-vivo

al., 1985; Sasner et al., 1972), isolated cells (Catterall et al., 1984), and

understanding of the neurological mechanism of PBTX causing canine

tra-cheal smooth muscle contraction, identification or prediction of compounds with neurological activity, a clearer understanding ofthe asthmatic abnor-

usedasaresearchtool

Materials and Methods—Culturing Ptychodiscus brevis—The dinoflagellate was grown

usingbatchculturetechniques(Brydonet al., 1971;Asaiet al., 1982).

Extraction ofCrudePtychodiscusbrevisToxinFromCultures—TheprocedureofAsaiet al.,

1982 was used. After 3to 7 weeksof cell growth, cultures were extractedwith 1 mlof 12Mhydrochloricacidperliter of cultureand50mlofpure chloroform perliter of culture(Martinet

al., 1970).Theresultingcrudetoxinwasthenstored inthedarkat-17°C

Purificationof PtychodiscusbrevisToxin—CrudePBTXwasdissolved inchloroformto give a

concentrationof 10 mg/ml Thesolution was then passed through a 3 cc C-18 Bond Elut®

column(Analyti-chemInternational,HarborCity, CA) Samplesrangingfrom5to100jAwere

injected into asolventprogrammable HPLC (Model 110A, Altex) equippedwith 2 pumps A LKB Bromma UVdetectormeasuringabsorbanceat254nm (Model 2238UvicordS2) wasused

andsampleswereseparatedthrough a reverse-phasecolumn(AltexUltrasphere,octadecylsilane,

5/*,4.6mminner diameter).Thecolumnpressurewasapproximately1000psi at 1ml/minflow

rate.TheHPLCwasprogrammedtoruna lineargradientfor10minutesstarting at0%

metha-noland endingat30% methanol inchloroform PBTXfractionswerecollected manually and

stored inthedarkat -17°C.Thesolventphasewasthenevaporatedunderreducedpressureusing

arotaryevaporatorandstored inthe darkat-17°C

TestingNeurologicalActivity ofPBTXonCanineTrachealSmoothMuscleStrips—

Neurologi-cal activitywas measured using methodspreviously described (Andersonet al., 1979).

Mixed-breed dogsweighing from 15 to 22kg wereanesthetized with intravenousinjections ofsodium

pentobarbital(30mg/kg) Thetracheawasthensurgicallyremoved and immediatelyplacedin

aerated (95% oxygen, 5% carbon dioxide) Krebs-Ringer solution ofthe following composition

(raM):NaCl, 117.0; KC1, 4.0; NaHC03 , 25; MgS04 , 2.4; NaH2P04 , 1.2;CaCl2 , 2.5; and

dex-trose, 11.0.Beforeobtainingsmooth musclestrips,mucous membraneandextraneoustissuewasremoved fromthetrachea.Smooth musclestrips (2 to3by15to20mm,weighingapproximately

30mg) weredissectedfromthecartilage.Eachstripwastiedwithsurgicalthreadandmounted

in an isolated water jacketed bath containing 10 ml of Krebs-Ringer solution, aerated (95%

oxygen,5% carbondioxide)and maintainedat37 °C,pH7.4.

Isometrictension was measuredwith a Grass FT03 force-displacement transducer (GrassInstrument Co., Quincy, Mass.) coupledto aGrassModel7 polygraph.Eachtissuewasinitially

loadedwith 2 gramsof tension,and30 minuteswasallowed beforetesting toallow thetissue to equilibrate.Thisinitial tensionwastested to assure thatresponseswere measuredat theLmaxforthis tissue.

Cumulativeacetylcholineconcentration-response curves(108 to103M)wereobtained before

andafterexposureofthetissue toPBTXto test tissue responsiveness.Crudeor purifiedPBTXwas

142 FLORIDA SCIENTIST [Vol.51

Aconcentration-responsecurvewasdeterminedforthecrudePBTXandthepurified fraction

ofPBTXhavingneurological activity.

Thelogconcentration of PBTX neededtocause50% of themaximal contractile response

(EC50)was determinedforthecrudeandpurifiedPBTXfractions.

Tachyphylaxiswas measured byconsecutive additionofPBTX tothetissuebath containing

tissue, washingandallowing thetissue to equilibrate for 15-minutesbetweeneach exposureto

thetoxin.

All contractile responseswerecalculated ingramstensiongenerated pergramtissue.

TestingHemolyticActivity ofPBTXon Canine ErythrocyteSolution—Hemolytic activity

wastested with the procedure of Martin et al (1972). Heparinized canine blood (5 cc) waswashed twice with 20 mL of phosphate-saline buffer of the following composition (mM):

NaH2P04H20, 4.0; Na2HP04 , 210, in 0.9% NaCl solution, pH 7.5. Absorbances of canineerythrocyte suspensionswere measuredat5-minintervalsusinganIBMmodel9420 spectropho-

tometer Test samples used 10/J of 10/ig/mLmethanolictoxin (crudeor purified) per 3 mlof

erythrocytesuspension.Controlsamples used 10/d ofmethanol

Spectra—The UV-visiblespectrum of 0.1 mg/ml ofa purifiedPBTX fraction dissolved in

methanolwas determinedusinganIBM UV-visiblespectrophotometer (model 9420). The

spec-trum wasscannedfrom185.0to450.0nm

Mass Spectrum—CrudePBTXwasinjected intothe HPLC andfraction 2awas collected.

Theresulting fraction was then re-injected into theHPLC and recollected. Thesolution was

evaporatedandthe residue,labeledtwice-purifiedPBTX-2A-2x,wassent toFloridaState

Uni-versityMassSpectrometryLaboratoryformolecularweightanalysis. Thesamplewasanalyzedusingpositivechemicalionization mass spectrometryona Finniganmodel 4510GC/MSspec-

trometerandscanningfrom to1050E/M.Isobutanewasthe reactantgasanda solidprobe

(80-240/20)wasusedtointroducethesampleintothe system

StatisticalMethods—TheStudent's t-testwasusedforcomparingunpaireddata.The

geomet-ricmeanwasusedto calculate theEC50 valuesfromtheconcentration-response curvesbasedon

Fleming's(1972) method.Dataareexpressedasameanplus orminusthe standard deviationor

standarderror.

Results —Purification of PBTX — The liquid chromatogram (Fig 1)

showsthe threefractionselutedfrom crude PBTX Thefirstpeak(fraction 1)

appeared at 6-8% methanol and 94-92% chloroform Fraction 1 was the

second peak (fraction 2a) eluted at 15-16% methanol and 85-84%

chloroform Theretentiontimeforeachfractionwasconsistentfrom one run

2.6 ±0.2, 5.1±0.2, and5.6 ±0.2, respectively

Neurological Activity ofPBTX — Datacollected were usedtoconstruct aconcentration-responsecurveforcrude andpurifiedPBTXfractions (Fig 2)

Fraction 2a had neurological activity (induced smooth muscle contraction);

con-centration-responsecurve (108 to 103 M Ch) was obtained foreach tissue to

test the consistency ofthetissue responsiveness before and after exposure to

PBTX An EC50 value of 3.6 x 106 mol/L (n=6) was calculated from the

initialcumulativeacetylcholineconcentration-responsecurve whilethefinal

EC50 value was 4.8 x 106 mol/L (n=6, P =0.25-0.10). The acetylcholine

unchangedinthreshold,EC50 ,ormaximalcontractile response

Nos.3/4, 1988] PEREZ-CRUET ETAL.—ACTIVITYOF TOXIN 143

Elution Time (min.)

Fig 1.Representative reverse-phasehighperformanceliquidchromatogramofcrude

Ptycho-discus brevis toxin.Thechromatographwasprogrammedtoruna lineargradientfor10minutes

starting at0% methanoland endingat30% methanolinchloroform Three primarypeaksare

seenontheelution profile. PBTX-1 hadhemolyticactivity, PBTX-2ahad neurologicalactivity,

andPBTX-2bdid nothavehemolyticorneurologicactivity.

the maximum contractile response (mean±S.E = 170 ±49 gram tension/

gram tissue) occurred at a concentration of 5.0 /ig/ml Following the first

0.1 ^g/ml and a maximal contractile response (mean±S.E = 169 ±56 gram

concen-tration-response curve of fraction 2a, after the second purification (Fig 3)illustratesathreshold concentrationof0.01 /ig/ml andthemaximalcontract-

ile response (mean±S.E =281 ±40 gram tension/gram tissue) occurred at a

PBTXconcentrationof0.06/xg/ml

Analysis of the data obtained from contractile responses of canine

tra-cheal smooth muscle strips showed an increase in the neurological activity

15 20 25 30

Crude PBTX (ug/ml)

Fig 2.Concentration-responsecurveofcrudePBTXoncaninetrachealsmoothmusclestrips.

Thethresholdconcentration ofcrudePBTX was 1.0 /ig/ml, and the maximum contractile

re-sponsewas induced bya concentrationof 9.0 jig/ml. Dataillustrated aremeanvaluesplus or

minusthe standarderror.

Fig 3. Concentration-response curveofPBTXfraction 2a, afterthe secondHPLC

purifica-tion,oncaninetracheal smoothmusclestrips. The threshold concentrationof fraction2awas

0.01 /xg/ml, andmaximum contractileresponsewas induced byaconcentrationof0.06/tg/ml.

Nos 3/4,1988] PEREZ-CRUET ETAL.—ACTIVITYOF TOXIN ]4,

Table1 Dataanalysis ofneurologicalactivity ofcrudeandpurifiedPBTXoncaninetracheal

smoothmusclestrips.

1

6b75

ameanplus orminusone standard error.

b ratio ofEC50values.

cPBTXFraction 2a after the firstHPLCpurification.

dPBTXFraction 2a after the secondHPLCpurification.

eTheneurological activity of Fraction 2a,andFraction 2a-2x, relative to crudePBTXwascalculated.

75-fold increase in neurological activity occurred with thesecond purification(Table1).

Consecutive AdministrationofPBTXto Canine TrachealSmooth Muscle

timestocrudeorpurifiedPBTXtotestfortachyphylaxis (rapidlydeveloping

withthe additionofincreasingconcentrationsofpurifiedPBTX-2a

Adminis-showed

contractile-response (gtension/gtissue) ofthesecond (mean±S.E.=41±33),

com-paringat5fig/ml of fraction 2a

Hemolytic Activity of PBTX — The rate of hemolysis of canine

was taken astheslope of thelinear relationship, -ln(At-A00) as a functionof

theendofthestudy period(typicallyseveralhours). The data weresubjected

10%, min1

are: crude (8.5±0.5), fraction 1 (16.1±0.9), fraction 2a

(2.8±0.06), fraction2b (2.9±0.09) and control (2.6±0.9) Analysisofthese

data show thatonly crude PBTX and hemolytic rate constants of fraction 1

Analysis of Purified PBTX — The ultraviolet-visible spectrum ofpurified

frac-tions The spectrum did not have any other significant peaks in thevisible

regionabove 450 nm.

weightofapproximately 900daltons, which is consistentwith Brevetoxin B,reported previouslyby Lin and co-workers (1981)

Discussion —Various HPLC solvent systemshave been utilized topurify

crude extracts ofPBTX. Pierce and co-workers (1985) used isocraticelution

with methanol andwater Also gradientelutionsystemshave beenutilizedin

which the solvent system becomes increasingly more polar Success was achievedin separating the ichthyotoxiccomponent from the hemolytic com- ponentofPBTXusingmethanolinchloroform (Padillaetal., 1979); asimilarseparationwasdescribedbyAsaiand co-workers (1982)

In the present study, a chloroform-methanol solvent gradient gavethree

samplesinjected intotheHPLC.

In the present study, the neurological activity of HPLC purified PBTX

PBTX was increased six-fold, and a 75-fold increase in neurological activity

ofPBTX was observed after asecond HPLC purification Previous research

with HPLC-purified PBTX on isolated canine tracheal smooth muscle (Asai

the presentstudy: fraction1had hemolyticactivitybut noneurological

activ-Nos.3/4, 1988] perez-cruet etal.—activityof toxin 147

PBTX caused canine tracheal smooth muscle contraction by stimulation

of the sodium channels in the axons of the parasympathetic postganglionicnerves resulting inareleaseoftheneurotransmitter acetylcholineatthe neu-

Morespecifically, PBTX appearstoinfluencesodiuminfluxatthehgate (Asai

channel mechanism was proposed for rat vas deferens smooth muscle, an

adrenergicallymediated system (Sakamotoetal., 1985)

The consecutive contractile responses of canine tracheal smooth muscle

stripstoa constant concentrationofcrude PBTXor purifiedPBTX, showthattachyphylaxisoccurred whenrelativelyhighconcentrationsofpurifiedPBTX

were utilized A significant decrease in the contractile response, compared with crude PBTX, of consecutive second and third responses was observed

occur as a consequence of depletion of the transmitter acetylcholine;

how-ever, it is more likely that the red tide toxin which is highly lipid soluble isbindingtotissuesitesandinactivatingthemforaprolonged timeperiod

Otherresearchershave observedtachyphylaxisinbothinvitroandinvivosystems usingcrude andpurifiedPBTX. Asaiand co-workers(1982) observed

contractionofcaninetrachealsmooth muscle wasfollowedbyrapid returntobaseline tension and the smooth muscle strips were insensitive to additional

crude PBTX Responses to acetylcholine were unaffected Blockage ofnerve

nervecordpreparationsofcrayfish (Parmentieretal ,1978)

This studyindicatesthatthecontractileresponseofcanine smooth muscle

weightof900 The massspectral resultseliminate anotherstructurethathas

been observed Unfortunately, there are at least five polyethers with

structures, differing onlyin a functional group at the terminal "K" ring on

an alcohol or an aldehyde Two other functional groups are noted that areattachedtothe K-ring Thefivestructures,then, haveabasicallysimilarring

(Shimizuet al., 1986), has a 10-ring structure thatdiffersfromtheotherfive

polyethers both with respect to thenumber and size ofthe individual rings

Again, our results would not distinguish betweenthis structureand the

oth-ers.

enhancementofactivity, andidentified an approximatestructure, theactual

compound or compounds responsible for thecontractile response remain to

be more adequatelycharacterized Onthe otherhand, we maywell wonder about the receptor site, and suggest that it may be more logical topresume

researchtoproveordisproveitsvalidity

Acknowledgments — Wearegrateful toDr James N Layne, whoserved

asConsultingEditor AU.S.F Sigma Xi Grant-in-Aidtooneofus(M.J P-C.)

LITERATURE CITED

Asai,S.,J.J. Krzanowski,W.H Anderson, D F. Martin,J. B.Polson, R F. Lockey, S. C

Bukantz,andA Szentivanyi 1982 Effects ofthetoxin ofredtide,Ptychodiscusbrevis,

oncaninetrachealsmoothmuscle: apossiblenewasthma-triggeringmechanism J. lergy Clin.Immunol.69:418-428

Al-Asai, S.,J.J. Krzanowski,R F. Lockey,W H Anderson, D F. Martin,J.B.Polson, S. C

Bukantz, and A Szentivanyi 1984. The site of action ofPtychodiscus brevis toxin

within theparasympathetic axonalsodium channelh gate inairwaysmoothmuscle J.AllergyClin.Immunol.73:824-828

Anderson,W H.,J J.Krzanowski,J.B. Polson,andA Szentivanyi 1979 Characteristics of

histamine tachyphylaxisincaninetrachealsmoothmuscle:a possibleprostaglandinatedmechanism.NaunynSchmiedeberg's Arch.Pharmacol 308:117-125

medi-Baden, D C, G Bikhazi, S. J. Decker, F F. Foldes, andI. Leung 1984. Neuromuscular

blocking action of two brevetoxins from the Florida red tide organism, Ptychodiscus

Catterall,W.A.and M.Gainer 1985 Interaction ofbrevetoxinAwithanewreceptor siteon

thesodiumchannel.Toxicon.23:497-504

Davis, C C 1948.Gymnodiniumbrevis sp nov.,acauseof discoloredwaterandanimal

mor-tality intheGulfofMexico.Bot.Gzz 109:358-360

Evans,E E 1973.Therole of bacteria intheFloridaredtide.Environ.Lett 5(l):37-44.

Fleming, W W., D P. Westfall, I S. De La Lande and L B. Jellett 1972. Log-normal

distribution of equieffective doses ofnorepinephrineandacetylcholine in several tissues.J.

Pharmacol.Exp.Ther.181:339-345

Johnson, G L., J.J.SpikesandS Ellis 1985.Cardiovasculareffects of brevetoxins in dogs Toxicon 23:505.

Lin, Y Y., M Risk, S. M Ray, D V. Engen, J. Clardy, J. Golik, J. C James and K.Nakanishi 1981 isolationandstructure ofbrevetoxinBfromthe"redtide" dinoflagel-

latePtychodiscusbrevis(Gymnodiniumbreve).J.Am.Chem.Soc. 103:6773-6775

Martin, D.F.andA.B. Chatterjee.1970.Somechemicalandphysical properties ofapurified toxinfromthe redtideorganismGymnodiniumbreve Fish Bull.68:433

,G.M.Padilla,M.G HeylandP.A Brown 1972 Effect ofGymnodiniumbreve

toxinonhemolysis inducedbyPrymnesiumparvumtoxin Toxicon.10:285-290

Padilla, G M.,Y S. Kim,E.J.RauckmanandG M Rosen 1979 Physiological activities of toxinsfrom Gymnodiniumbreveisolatedbyhigh performanceliquidchromatography

Pp 351-354 In:Taylor,D L.andH H Seliger, (eds.), Toxic DinoflagellateBlooms,

Nos.3/4, 1988] perez-cruetetal.—activityof toxin 149

Parmentier,J.L., T.Narahashi,W A.Wilson,N M.Trieff, V.M Ramanujamand M.Risk.

1978 Electrophysiologicalandbiochemicalcharacteristics ofGymnodiniumbrevetoxins.

Toxicon.16:235-244

Pierce,R.H., R.C BrownandJ.R.Kucklick 1985 Analysis ofPtychodiscusbrevis toxinsby

reversephaseHPLC.Pp 309-314 In:Anderson, D.M., A.W.WhiteandD G.Baden

(eds.), Toxic Dinoflagellates, ElsevierSciencePublishing Co.,Inc.,NewYork.

Sakamoto, Y., J. J. Krzanowski, R Lockey, D F. Martin, R. Duncan, J. Polson and A.

Szentivanyi 1985.The mechanismofPtychodiscusbrevistoxin-inducedratvas deferens

contraction.J.AllergyClin. Immunol.76:117-122

Sasner,J.J.,M.Ikawa,F.ThurrergandM Alam.1972 PhysiologicalandchemicalstudiesonGymnodiniumbrevisDavisToxin.Toxicon 10:163-172

Shimizu,Y.,H-N Chou,H Bando, G.VanDuyneandJ.C Clardy 1986.Structureof

Breve-toxinA(GB-1toxin),themostpotenttoxin intheFlorida redtideorganism iumbreve (Ptychodiscusbrevis) J.Am.Chem.Soc, 108:514-515

Gymnodin-Shinnick-Gallagher,P 1980 Possiblemechanismof action ofGymnodiniumbrevetoxin atthe

mammalianneuromuscularjunction Br.J.Pharmac.69:373-378

Steidinger,K A 1979 Collection,enumerationandidentification of free livingmarine

dinofla-gellates. Pp 435 In: Taylor, D L and H H Seliger (eds.), Toxic DinoflagellateBlooms,ElsevierNorth HollandInc.,NewYork.

,M.A BurklewandR.M Ingle 1973 Pp 179-202.In:Martin, D.F. andG M.Padilla(eds.),Marine Pharmacognosy,AcademicPress,NewYork.

Walker,S T 1884.FishmortalityintheGulfofMexico.Proc.U.S.Natu.Mus.6:105.

FloridaSci.51(3/4):140-149 1988

Accepted: September8, 1987

150 FLORIDA SCIENTIST [Vol.51

THE ANASTASIA ISLAND COTTON MOUSE (Rodentia: Peromyscus

NaturalResources, 13610S.E.FederalHighway,HobeSound,FL, 33455

Abstract: TheAnastasia Island cotton mouseappears to have been gone from AnastasiaIsland forseveraldecades,butthis statushadnotbeenacknowledgedpreviously. Aclear state-

mentofthe survival statusofthis subspecies isconfounded bytheproblematicstate ofknowledge

ofitstaxonomy If its range is limited to Anastasia Island, the taxon is extinct If the range

includes Cumberland Island, Georgia, only theformer population is extirpated. Though the

mechanismoflossfromAnastasia Islandisunknown, muchofthe coastal foresthabitat ofthis

animalpersistsbutisoccupiedbyexoticblackrats (Rattus rattus).

Thecottonmouse onAnastasiaIsland, St.JohnsCo., Florida, was named

asPeromyscus anastasaeby Bangs (1898) and assigned toPeromyscuspinus anastasaeby Osgood (1909) Cotton mice onAnastasia Islandoccurred mostly inthickets ofwax myrtle andSpanishbayonet andoccasionallyinthe

else-where, maritime live oak forest probably also was occupied Osgood also

Co., Georgia, which had been named asP. insulanusby Bangs(1898)

The Anastasia Island cotton mouse is smaller and paler (more buffy in

theunderparts are grayish white, andthetail isdusky above and white below

(Bangs, 1898) To this description, Osgood (1909) added that thecombined

populations of Anastasia and Cumberland Islands have upper parts "pale

ochraceousbuff ratherlightlymixed withdusky, whichisslightlyornotatall

concentratedinthe mid-dorsalregion."

The purposeofthisstudywasto determinethe current geographical

populationwasconsidered "status undetermined" by Neuhauser (1978) It isunder reviewforpossiblelisting asthreatened orendangered (U.S Fishand

WildlifeService, 1985)

Methods—Samplingwas conducted onAnastasiaIslandandvicinityfrom22Mayto12lune

1986(Fig 1) Sites AI-9, 10, 13, 14, 15, and 17werewithin Anastasia StateRecreation Area

Sites AI-7, 8, 11, 12,and16andSiteRI-1werewithin FortMatanzasNationalMonument.Sites

CI-1,3,and4andAI-3, 8, 9, 10, 11, 15, 16,and17wereinwaxmyrtle,palmetto,xeric oak, or

mesicoakhabitatsjudgedsuitable forcottonmice Cottonmicealso arefoundoccasionally in

beachmousehabitat SitesCI-2; AM,4, 5, 6, 7, 12, 13,and14; RM; andML-1wereindune

grasslandhabitatjudgedsuitable forbeachmice.SiteAI-2wasindensegrassjudgedsuitable for

cottonrats. Samplingalsowasconducted onthebarrierbeach fromVilanoBeach northwardfrom 12-20lune1986 Sites SJ-3 to9werelocatedinGuanaRiver WildlifeManagementArea

Sites SJ-4, 7,and9wereinhabitatsuitable forcottonmice,andSites SJ-3, 5, 6, and8werein

habitatsuitable forbeachmice

Trappingwasdonewithlines of40largeShermantrapsbaited withrolled oats. To preventmortalityandsaturation of trapsbycottonrats, whichare active at alltimesof day, trapswere

set inthelateafternoonandcheckedandclosed inthe morning Traplineswererunfor 1 to3

nights.In reporting the samplingeffort, trapnightswere adjustedforsprungtraps, countedas

Nos.3/4, 1988] HUMPHREYETAL.— COTTON MOUSEEXTINCT? 151

r

SJ-3 SJ-4

SJ-6

O O

ST AUGUSTINE INLET

CI- 1-4

-AI-14

-AI-13 AI-10,18

AI-8,1 1,16

AI-7,12 ML-1

Fig 1. Locations of samplingsites onAnastasia Island and northward in St. Johns Co.,

Results and Discussion — We captured nocotton mice onAnastasia

Is-land in 880.5 trapnights in suitable habitat, or in over 2,000 trapnights of

beachnorthofVilano Beach(Table2)were darkincolorandnotreferable tothe subspeciesanastasae

Table1. SamplingofrodentswithShermantrapsbaitedwithrolled oatsonAnastasiaIsland,

nearbyConchandRattlesnakeIslands,andvicinity,22Mayto12June1986.

Adjusted

Site nights gossypinus polionotus hispidus musculus rattus

Cotton mice onAnastasia Island apparentlywerenotcommon whenfirst

encountered — Bangs(1898) took only nineindividuals Elliot (1901, cited inPournelleandBarrington, 1953) reported tenspecimenstakenat"Espanita",the homeofa Mr Middleton, located 2-3 miles(presumablynorth) from thetypelocality Since 1901, atleastfour unsuccessfulefforts have been madetofind this population Pournelleand Barrington (1953) failed tocatch anyin

306 trapnights in 1948, even though the purposeof their trip wasto obtaintopotypes Smith (1987) failed to catch any in 3,553 trapnights in October

1978; thiseffortwasdirectedatcottonmice andcottonrats, sotrappingwas

ap-Nos.3/4, 1988] Humphreyetal.— cotton mouseextinct? 153ering this history of trapping results, it is evident that the population on

Anastasia Island hasbeenextirpatedfordecades

As the Anastasia Island taxon wasoriginally conceived byBangs, it was endemic to the island, and the present interpretation of its survival status

would bethat thetaxon is extinct However, the taxonomic revision by

geo-graphicstatusofthis population (and hence its survival status) uncertain If

remains common elsewhere in its range (on Cumberland Island [Joshua

Laerm, 1987]) Ifonerejectsthecurrenttaxonomic arrangement, aswe

sug-gest,theAnastasia Island cottonmouse may beextinct

Osgood referred only 18 specimens from Anastasia Island and 36 from Cumberland Island toanastasae However, he examinedP. gossypinus from

other nearby areas, including St. Marys, Georgia; Amelia Island; Burnside Beach[on the coast ofDuval Co. northofAnastasia Island]; Carterville[onthemainland westofAnastasiaIsland]; and Summer Haven[southofMatan-

zasInlet] Osgood's remarks about anastasaecastdoubt onthevalidity ofhis

taxonomic arrangement: "Althoughthe paleforms fromAnastasiaand

Cum-berlandislands, respectively, are entirely isolatedfrom each other and from

themainland forms, theyseemto be absolutelyalike and also arenot

Moreover, themainland specimens mostsimilar to them arenot fromties immediately adjacent to the islands in question, specimens from St.

locali-Marys, Ga., Burnside Beach, Fla., etc., being typical gossypinus." Further doubt aboutthis arrangement can be inferredfrom Osgood's remarks under

P. g palmarius: "The type ofpalmarius and a very smallpercentage ofthelargeseriesoftopotypes areunusually paleandscarcelydistinguishablefrom comparable specimensofanastasae...thegreatpreponderanceofdarkspeci-

mens from the type locality [Oak Lodge, on East Peninsula at present-day

Floridana Beach, acrosstheintracoastalwaterway from Micco, Brevard Co.,Florida] tends to indicate that the type is probably an aberrant specimen

construed also to the effect that pale coast forms are undergoing parallel

onlyinitsincipiencyonthe peninsula opposite Micco."

arrange-ment given by Osgood implies a range forP. g anastasae extending far

mainland and excluding other portions of the mainland not known to be

—

thatsuchan arrangement mayrepresentthelumpingof relativelyunrelated

adaptation

Table2.SamplingofrodentswithShermantrapsbaitedwithrolled oatsonthebarrierbeach

in St.JohnsCounty fromVilanoBeach northward from12-20June1986.

Adjusted

Site nights gossypinus nuttalli hispidus musculus rattus

variationinP gossypinus Without good taxonomy,wildlifemanagers cannot determine what geneticentitiesareatrisk.

Pournelleand Barrington (1953) citeddevelopment by manasthe

Though much of the habitat of cotton mice on Anastasia Island has been converted tohuman uses, substantial amounts of suitablehabitat remain in

habitat now areoccupied byblackrats(Rattusrattus; Table 1). Though we

(Humphrey and Barbour, 1981) Research is needed on whether and under what conditions a strong competitive interaction of cotton mice and black

ratsispossible

Acknowledgments—This studywasmadepossiblebytheleadership ofpersonnelof several federalandstate agencies:Michael Bentzien,EdwinConklin,JohnParadiso,FranklinPercival,

James Stevenson,David Wesley, andDonWood The workwas funded bythe U.S Fishand

WildlifeServicethrough CooperativeAgreementNo 14-16-009-1544, WorkOrder No.37,withtheFloridaCooperative FishandWildlifeResearchUnit.Wethank JudyLamiaforhelpwith the

field work We appreciate theinterest and cooperation of numerous individuals inthe field,

including themanagersandstaff ofAnastasiaStateRecreationAreaandFortMatanzasNational

Caro-Nos.3/4, 1988] Humphreyetal.— cotton mouseextinct? 155Humphrey,S R.,andD.B.Barbour.1981 Statusandhabitatof three subspecies ofPeromyscus

polionotusin Florida.J.Mammal.62:840-844

Laerm,J.1987.UniversityofGeorgiaMuseumofNaturalHistory,Athens,GA., Pers. Commun.

Neuhauser, H N.1978. AnastasiaIslandcottonmouse.Pp.45-47 In:Layne,J.N.(ed.), Rareand endangeredbiota of Florida Vol 1.Mammals Univ.Presses of Florida, Gainesville.

U.S FishandWildlifeService.1985.Endangered andthreatenedwildlifeandplants;reviewof

vertebratewildlife; notice ofreview.Fed Reg 50:37958-37967

FloridaSci 51(3/4):150-155 1988

Accepted: April8, 1988

A MODIFIED LIVE TRAP FOR THE CAPTURE OF SOUTHEASTERN POCKET GOPHERS—CyndiA Gates1

, George W Tanner,andBrian K Gates,

Department ofWildlife and Range Sciences, University ofFlorida, Gainesville, FL

32611

Abstract: A modifiedlive trapfor the capture ofsoutheastern pocket gophers (Geomys

pinetis) wasdeveloped for usein sandhill populations. Trappingsuccess ratewas69%. Therate

ofpluggingof trapsbypocket gopherswaslowerandcaptureratemuchhigher than forSherman

pocketgophertraps.

Although anumberofauthorshavereportedonlive-trapping ofGeomys

has been published concerning live-trapping of the southeastern pocket

be-tween2February and9 April 1986,insandhillcommunities onthe Katherine Ordway-Swisher Preserve in Putnam County, Florida To accomplish thistask, livetrapswereplacedinthemaintunnel systemsofthepocket gophers

Main tunnels werelocated by clearing away soil of fresh mounds to expose

often were encountered and indicated that the excavator wasfollowing the

but ranged from8to91cm.

presenceoftheanimal could bedetected; i.e. the traphad been sprungorthe

burrowortrapplugged withsoil.

Initialtrappingeffortswhich employedthe useofShermanpocketgopher

yielded no capturedanimals Of 12 totaltrap sets, 3were active Ineachof

Subsequentefforts involvedsettingSherman-designedtraps asabove, but with boardsplaced over the top oftheexcavation andsoilplaced aroundtheedges ofthe boards to insure that nolight entered the burrow Under these

pluggingofburrows and/ortrapswithsoilwasstillacommonoccurrence In

3cases, animalscompletelyfilled the excavation, packingsoil flush withthe

bottomofthe boards

Completely covering the trap, as noted by Howard (1952), andrefilling

the excavation site were important factors in increased trapping success

However, the treadle design ofSherman's trap does not lend itselfto being covered with soil because of interference with the tripping mechanism.

Therefore, a modifiedtrap was designedto incorporate features thatwould

Nos.3/4, 1988] gatesetal.—modifiedgopher trap 157

allow complete coverage with soil (Fig 1). The dimensions ofthe trap (7.5

pivotingpaddle mechanism (A), similar in design to thatused by Baker and Williams (1972), was attached to the top rear portion of the trap, 3.5 cm

forward fromthebackwall The paddle was suspendedtowithin4 or 5 mm

ofthe trap floor Astainless steelpin (B) held thepaddlein place A3.2 mm

diameterstainless steelwire rod (C), about 14cmlong, wasinsertedthrough

thebottomcornerofthepaddle (D) and extended alongthe sidewall to thetop ofthetrap to a point about 11 cm from the front The rod was held in

placebya guideconsisting of3.2mm diameterstainless steelwire(E) insuch

paddle mechanism was pushedback, therebyreleasingthe trapdoor(F)

Fig 1.Sideviewofthemodifiedlivetrapforsoutheastern pocket gophers

served as thefloor and back wall ofthe trap (G). Thefloor of the trap was

held in place with a stainless steel pin (H) so that it pivoted at its point of

attachment to the trap walls The floor and back wall component could be openedtofacilitateremovalofcaptured animals (I)

Forty-one pocket gophers were captured with themodified trap without

mortality or injury Trapping success was 69% (41 of59 activesets), which

withthestandard Shermantrap When alltrap setswere counted (includingthose that were never revisited by the animal), the capture rate was 31%

Trapping success was greater when trapswere left in place for 12-16 hours(66% ofcaptures). There was afairlylow trap responserate overall (59 sets

Acknowledgments— WethankDr.RonLabisky,Dr.JohnEisenberg,Dr.MelSunquist,and

Ms.SuzanneBrandforreviewing the manuscript.PartialfundingforthestudywasprovidedbyTheNatureConservancy This paperisFlorida AgriculturalExperiment StationJournalSeries

Hart,E B 1973.Asimpleandeffective live trap forgophers.Am.Midi Nat 89:200-202

Howard,WE.1952.Alivetrapforpocket gophers.J.Mammal.33:61-65.

Sherman, H.B 1941.Aboxtrap forthecaptureof liveGeomys.J.Mammal.22:182-184

FloridaSci.51(3/4):156-158 1988

Accepted: April11, 1988

CORRELATIONS AMONG EGG WEIGHT,

P.A.Skewes2

,H. R Wilson andF.B Mather

PoultryScienceDepartment,University of Florida, Gainesville,Florida32611

Abstract: The relationships among egg weight, chick weight, yolk sacweight, and carcass

weight(chickweightminusy.s. weight)inBobwhitequail chicks athatchweredetermined Threeegg weight groupings(9.6,10.4and11.2g)wereused. Chickswereremoved fromtheincubatorat

4-hrintervalsfromthe start ofhatchingand body andyolk sacweightsweredetermined

immedi-ately. Largereggsproducedlarger chickswithlarger yolk sacsandlarger carcasses. Aseggsize

increased, the chickweightasa percentage of egg weightincreased Yolk sac as a percentageof chickweightandyolk sac asa percentage ofcarcassweightweresignificantlylower forchicksfrom

smalleggs. Thefollowingcorrelationswerefound: egg weightto chickweight, 0 88;eggweightto

yolk sacweight,0.38; chickweight to yolk sacweight, 0.37; egg weightto chick carcassweight,

0.73.

Early mortalityis aproblem in commerciallyrearedBobwhite quail and

appearstobe moresevere inthesmallquail chicks Inturkeys, earlymortality

reported that the amount of residual yolk found in poults at placement was

heavierin thepoults from larger eggs, but they alsorepresented a largerportionofthe poult weight

pro-In additionto theimportantroletheyolksacplaysin embryonic

develop-ment, it is also presumed to be the primary nutrient source for chicks until

placement (Romanoff, 1960) Harvey and co-workers (1955) deutectomized

two weeks but were comparablethereafter

Althoughthe influenceofeggsizeontheamountofyolk materialavailable

and on Bobwhite quail (Romanoff, 1944; Beane, 1981) suggest that as egg weight increases so does chick weight and yolk sac weight It is not known, however, if the chicks from small quail eggs have proportionately less yolk

insmall quail chicks This study wasinitiatedtodetermineifthe influence of

eggsizeonthe amountofyolk material available totheBobwhite quail chick

athatchisconsistent foralleggsizes.Additionally,therelationshipsamong egg

weight, chick weight,andyolkweightinBobwhitequailwere examined.

1 Florida AgriculturalExperimentStations Journal Series No 8005.

Methods—Eggsfrom 18-month-oldBobwhitequail breederswerecollected dailyandstored

from 1 to7days at 12.8°C with a relative humidityof80-85% Prior to setting, all eggs wereremoved fromthe cooler,weighedto the nearest 0.1g,andgroupedinto thefollowingthreeweight

categories: 9.6, 10.4,and11 2g.Atolerance of plus orminus0.1gwasallowedineach group.Eggs

betweenthesegroupswerediscarded.Theeggswereincubatedat37.8°Cdrybulband30°Cwet

bulb temperatures Incubation temperatureswere changedto37.2and33.3° C,respectively,on

the twenty-first day.

Afterhatchingwas underway, chickswereremoved from theincubatoratfour-hour vals, killedbycervical dislocation,and weighedtothenearest 0.1 g.Yolksacswereremoved by

inter-tearing the skinfromtheabdomenand making anincision intheperitoneumalongthemidline,

anterior tothe umbilicus.Theyolksacwasthen forcedfromthebodycavitybyapplyingpressure

tothesides oftheabdomen Theyolksac stalkwasseveredwithforcepsandthe yolksac(with

contents)was removed and weighedtothenearest0.01g.Carcass weightswerereportedaschickweightminustheyolksacweight

Data weresubjectedtothe generallinearmodel andcorrelationprocedures(Barr et al., 1976)

withsignificanttreatmentdifferencesdeterminedby Duncan'smultiplerangetest(SnedecorandCochran,1980).

Table1. Theinfluence ofegg weightonchickweight, yolksacweightandcarcassweightof

Bobwhitequail.

weight weight weightYolk as% as% of as%of

weight weight weight weight weight weight weight

• bcValueswithinacolumn bydifferent superscript letters are significantly different(P<0.05)

Results and Discussion — The average chick weight of 7.6 g (Table 1)

hatch averaged 71.4% of the original egg weight This is higher than the

66.2% value reportedforJapanesequail (Ghanyetal., 1966) andthenormal rangeof62 to68 % reportedfor chickens (Halbersleben and Mussehl, 1922;

values for Bobwhite quail: yolk sac weight of 0.49 g, yolk sac free carcass

weight of 4.95 g, and the yolk sac represented 9.9% of the chick weight

These values are considerably lower than the corresponding values in this

study 98 g, 6.6 g, 12.9%, respectively This discrepancymay be dueto

lim-itedsamplesizeoftheearlier study

Egg weight had asignificant effect onchick weight, yolk sacweight, cass weight, and chick weight as a percent of egg weight (Table 1). As egg

car-and