ACKNOWLEDGEMENTSThe author would like to thank Dr. JosephS.Davis for hiscontinued academic and research guidance, and Dr. William Hallerfor his financialsupport and research guidance.The author would also like to thank his loving wife, Sharon.The completion of the author's education and this study wouldhave been impossible without her constant understanding, guidance,financial support, and excellent typing.

TABLE OF RPOSE3MATERIALS AND METHODS5Descriptions of Study AreaField Sampling ProcedureChemical and Physical MeasurementsPhy top ankton and Zooplankton Collectionand CountingPeriphyton DiatomsWater Hyacinth LengthWater Hyacinth BiomassSediment AccumulationStatistical Analysis57131151720212125RESULTS25Water Hyacinth ControlChemical and Physical MeasurementsPhytoplanktonZooplanktonPeriphyton DiatomsWater Hyacinth LengthWater Hyacinth BiomassSediment LUSIONS257BIBLIOGRAPHY260BIOGRAPHICAL SKETCH.262

Abstract of Dissertation Presented to the Graduate Council ofThe University of Floridain Partial Fulfillment of the Requirements for theDegree of Doctor of PhilosophyBIOLOGICAL AND PHYSICAL INVESTIGATIONS OF BODIESOF WATER BENEATH DENSE WATER HYACINTHPOPULATIONS BEFORE AND AFTER CHEMICAL TREATMENTByWi11iam W.BrowerAugust 1980Chairman:JosephS.Major Department:DavisBotanyDense infestations of water hyacinthsMart.(Eichhornia orassipesSolms), or other free-floating, mat-forming aquatic weedspecies, dramatically affect the water quality below them.Dissolved oxygen concentrations, summer water temperatures,the annualare allrange of water temperature, and light penetrationreduced below actively growing hyacinthsuspended organic material, C0„, andincreased.HCO.mats,whileconcentrations arePhytoplankton populations were fairly abundantbeneath actively growing hyacinth mats, and composed mostlyof blue-green algae(e.g.Oscillatoria limnetica) and phyto-planktonic diatoms (e.g. Fvagilaria aacpuoina var. mesolepta

and Melosira granulata var. angustissima).Under such conditionssurface periphyton diatom populations were highly variable andminimal while periphyton diatom densities at greater depths werevery low.Zooplankton populations were large, being composedmostly of rhizopodal protozoans(such as Arcella spp.andCentropyxis spp.), and also various rotifers and crustaceans.Maximal hyacinth fresh weight biomass was found to be 57-7 kg/Mwhile the maximal dry weight biomass was 6.1 kg/M22.The hyacinth mats were chemically treated with 2,4-D,which proved to be an efficient means of aquatic weed control.After chemicaltreatment,floating mats of dying hyacinths werepersistent on the pond surfaces for several months.Waterquality under the decaying mats was worse than under activelygrowing hyacinths.Phy toplankton and zooplankton populationswere much reduced while the periphyton diatom populations werevariable.The zooplankton was mostly composed of rhizopodalprotozoans which were feeding on the decaying vegetation.Open water conditions were established approximately sixmonths after herbicide application, and water quality slowlyimproved.Dissolved oxygen concentrations, pH,tration, water temperatures, and annualatures allincreased, while COand HCOsuspended organic material decreased.light pene-range of water temper-concentrations, andPhy topi ankton populationswere large after hyacinth treatment and disappearance, beingmostly composed of blue-green algae, but with occasionalseasonal green algae populations.largePeriphyton diatom populations,

.increasedinboth the pond surfaces and depths.Zooplanktonpopulations were reduced after open water conditions wereestablished and were largely composed of various rotifers andcrustaceansSediment accumulation studies showed that the dry weightof sediments deposited below actively growing hyacinths averaged0.218 kg/M2during an eight month period(3-6 percent of theoriginal hyacinth biomass available), while that below chemicallytreated hyacinths was 0.626 kg/Mbiomass).(20.9 percent of originalTherefore, sedimentation was much greaterafter herbicide application to hyacinths.of the hyacinth biomassthe sediments.(79.1(5-8 times)However, the majoritypercent) was not accounted forin

.INTRODUCTIONThe water hyacinth, Eichhornia arassipes Mart. Solms,isaperennial, free-floating, mat-forming aquatic plant, whichoften accumulates to form dense infestationssubtropical areas(Holm et a).1United States of America circa 1890,environmentalnative to South America.has caused significant economic andconstitutinga(A)(2)impeding drainage,(3)destroyingreducing water related recreation andhazard to life.However, perhaps the mostserious effect of dense water hyacinth infestationsenvironmentalAccording(19 8), water hyacinths cause problems byobstructing navigation,wildlife resources,(5)isintroduced into theimpact on Southeastern U.S. Penfound and Earle(l)ittropical and969)The water hyacinth, an exotic speciesSince its introduction,inistheirimpact upon the water quality and species compo-sition of aquatic ecosystems.Itiswellknown that waterhyacinth infestations deteriorate water quality by reducing thepH,dissolved oxygen concentrations, nutrient availability,andlight penetration of the underlying bodies of waterand Boyd,1975)-(McVeaBut what effect do water hyacinths have uponthe naturally occurring aquatic producer communities, which arethe basis for the entire aquatic food web?How are the phyto-plankton and periphyton communities affected, and what effect

willthis have upon the zooplankton communities?These are allquestions which have not been adequately investigatedinthepast, and to which this study was partially addressed.Control of aquatic weed populations has been thoroughlyinvestigatedinthelast severaldecades.Mechanical harvestingof water hyacinths has been attemped time and time again, buthas proved uneconomical.These failures are due to severalfactors, all of which continue to make mechanical harvestingunfeasible onalarge scale basis.expensive harvesting machinerydown often,isisFirstly, elaborate andneeded.difficult and costly to repair, and requiresfairly skilled employees to operate.vegetationisThis machinery breaksMore importantly, aquaticcomposed mostly of water.time and energy expended on mechanicalTherefore, most of theharvestingiswasted onmoving water, not plant material.During the last two decades there have been extensiveinvestigations into the use of various herbicide formulationsas a controlchemically dense growths of aquaticThis appears to be the most economically soundmeans of aquatic weed control available at this time.The most commonly used herbicide to control water hyacinthsis2,4-D (2,4-Dichiorophenoxyacetic acid).This herbicide hasbeen successfully used for many years throughout the South-eastern U.S., and especially Florida, where aquatic weed problemscan become quite severe.However,little emphasis has been placedupon the in vivo conditions which arise due to herbicide applicationand the effects that they have upon water quality and aquaticbiological communities.

.PURPOSEThe objectives of this study were as follows:1.Determine the effort required to eradicate waterhyacinths from smallfarm ponds using a herbicideformulation of the oil soluble amine of 2, -D(Emul sami ne E 3)2.Study the aquatic biological communities(phyto-plankton, periphyton diatoms, and zooplankton)associated with naturally occurring dense infestations of water hyacinths, populations of chemicallytreated hyacinths, and thoseinopen water conditionsafter hyacinth treatment and disappearance.3.Compare water quality parameters(chemicalandphysical) associated with natural hyacinth pop-ulations and open water conditions after hyacinthtreatment and disappearance.k.Investigate the interrelationships and inter-actions existing between aquatic producer (periphytonEmulsamine E 3 is a registered product of Union CarbideAgricultural Products Incorporated.Emulsamine E 3 contains50.7 percent of the dodecyl and 12.7 percent of the tetradecylamine salts of 2 -Di chlorophenoxyacet c acid.The productis formulated as an oil soluble salt and contains 359.5 gmacid equivalent per liter.,i

diatom and phytopl ankton) and consumer (zoopl ankton)communities, along with the physical and chemicalparameters which affect them.5-Compare the environmentalimpact of herbicideapplication to water hyacinths, to the impactof naturally occurring populations of waterhyacinths on the aquatic environment.

,MATERIALS AND METHODSDescription of Study AreaThe three ponds studied areGroves,alocatedinSilver Springscommercial citrus grove near Macintosh, Florida.These three ponds are shallow bodies of water used for irri-gation of the surrounding groves during dry periodsfrost protection during cold weather.in1961.and forThe ponds were constructedThey originally had sand bottoms and clear water,being fed by severalsprings and ground water runoff.smallWater hyacinths eventually invaded the areaand totally covered the surface of allThe ponds are highly eutrophic,of nutrientsinthelatethree ponds(Fig.1960's1).receiving an abundancefrom springs and surface runoff of the fertilizedcitrus groves.The once sand-bottomed ponds developed thickdeposits of dead and decaying organic material on the pondbottoms which originated from naturally dying water hyacinthsand other vegetation.was poor.Water quality under these conditionsThe water contained much suspended and dissolvedorganic matter,little(if any)dissolved oxygen, abundantnutrients, and smelled strongly of hydrogen sulfide.The three ponds willand Three,be referred to as Ponds One, Two,for convenience.with the highest pondsAllthree ponds are(Pond One and Pond Two,interconnected,respectively)

4m*A*AM.* 44H l! HH H 4H4H H. H. H H'H "H V* V J,h;&%.W&-f*rJr \\ - 'v WJ, . r?A'*-f* fg?'xslU *f'-XrA A 4 /& A v''*»"r-'.' * *4 4 4 4 4 4 4 4 4 4 "44 *44A A A A A A A A"a"»"a A"» A »"a* a * 4 Aa A4 4 * a. a - a --, -a, -a,4 4 4 4 4 4 4 4 4 " 4 4 4 4 C4 4 4 4 4 4 4 4 4 4 4 4 4 4 4I,,5.,4 4 4 44 4 * 4 *4 4 4 4 4 4 4 44 4 4 4 4 4? 44. 4 4 4 0. 4 4 *?4 4 4 4 4 44 * * *'4 4"44'44 44 4 4 4 4 A * 4 4 4 4 4 4 4 4444 .H c a ,* -,,H H4444444O H 'HH H M H »H.444444 4 4 o 44 4 0. 4 4 0. 44 0.4 4 4 4(J. .' !* 4 4 4 4 4 4 * 4 4 *4 4 4 4 4 4 4 4 4 4 44 4 4 4 4 4 4 4 "4 "4 4 4 44 4 4 4 4 4 4 4 4 4 4 4 4 44 4 4 4 4 4 4 4 4 4 '4 4 44 4 4 4 4 4 4 4 4 4 4 4 4 44 4"4 4 4 4 4 4 4 4 4 4 4 4 44 4 4 4 4 4 4 4 4 4 4/44"4 4 4 4 4 4 4 4 4 4 4 4 4 444 4444444444444444MMMMM4444444444f.'«.l i.'Ia a.r'T*'*"*; a»4AM444444444M M 'M M//*4 4 4 4 4 44 4 4 4MMMM MM M M MHMMMM /MM.MMMM"MMM4444 M MM'4 V4 4 4 * 4 4 4 4 4 4 4 4 4 a,4 4 '4 "a, \j " * 'A,//§ « "A " " A. 4 A* // 0. A A ' i A A//«4 4 4 * 44 4 44 4 440. [.//«3 4 *"'4'*,.0.4//"4"A *%***.***A 4« 44 4 4 4 4 4//v4 44 4// // //4 4 4 4 4 4 4-44 § 4 4 4 4 5 4 4 44 4 4 44 4 4 4'}'J4 4 4"444 4 44 4 4 44 4 4V 4 4 44 4 44 4 4 44 4 45 4 4 44 4 4 44 4 4 44 4 4 44 4 4 4 4A44 4 4 44 4 44 4 4 4 4 4 4 4 4 "4 4 44 4 4 4 4 4 4 4 4 4 4/44 4 4 4 4 4 4 4 "4 4 44*444444444444444HOJ4 4 4 4 4 4 4 4 4 4 4 444444444 4 4 "4"MOM4 4 4 4 4 4 4 4 4 4 4 44 4 4 4 4 4 4 4 4 4 44 4 4 4 4 4 4 4 4 4 44 4 4 4 44 4 4 4 4 44 4 4 44 4 MMMU44MM4 A 4Figure1.444444444444 4 4 4 4 44444444.4'. 444 4 4 4 4 4 4 4 4 64 4 4 4. 4 4 4 4 4 44 4 4 4 4 4 4 4 4 4 44 4 4 4 4 4 4 4 4 4 4"4 4 4 4 4 4 4 4 4 4 4 44 4 4 4 4 4 4 4 4 4 4"*'//ft4 4 4 4 4 4. 4 4 4 4 44 4 4 4 4 4 4 4 4 4"44444"4 4 4 4 4 4 4HHM HHM44"4 Map of Silver Springs Groves study areamixed hardwood forest).(4 4 4 4 citrus trees, ./'///

,feeding the lowestTwo bya(Pond Three).Pond Onethrough which runsdike,aisseparated from Pond30 cm culvertallows water to flow from Pond One to Pond Two.thatPonds Two andThree are connected by a standpipe which allows overflow fromPond Two to enter Pond Three.exits throughOverflow water from Pond Threesimilar standpipe, and drainsainto a smallwhich eventually enters Orange Lake.Pond One hasarea of 0.81 hectares, with depths toaPond Twoasurfacemaximum of 1.5M.located between Ponds One and Three andiscreekisthe deepest of the three ponds, with depths up to a maximumof 3- M andsurface area of 0.99 hectares.aPond Three with a surface area of 1.87 hectares hasand a canal-like arm leading towardsfairly circular main bodyPond Two.Pond Three hasaamaximum depth of 2.0M.Field Sampling ProcedureThe ponds were studied duringspan from April8,1975,untilathirty-four month timeJanuary 18,1978.Samples weretaken every three or four weeks throughout the study period,except duringtoadrawdown period(when the ponds were drainedremove nutrient-rich water and to consolidate and oxidizethe sediments)inthe summer and fallof 1976.Phy toplanktonzooplankton, and periphyton diatom samples were taken duringeach sampling period.Water samples and physical measurementswere obtained as often as possible.

Before herbicide application, all sampling had to be donefrom a small jonboat towed across Ponds One and Two, or fromadock built out into Pond Three, due to dense hyacinth popu-lations.A rowboat was used for sample collection after openwater conditions were established.Theinitialherbicide applications of 2, -D were done byprofessional field crews from the State of Florida Game andFreshwater Fish Commission.Using an airboat and sprayingequipment, 2,A-D was applied to Ponds Two (Fig.(Fig.3)on May 15,1975,at the rate of 3-3A kilograms ofactive ingredient/hectare.months laterto(July15,and Three2)1975,Respraying was necessary two and sixand November 19,insure complete control of hyacinths.ponds were resprayed after two months.respectively)1975,Due to regrowth,theseAt that time the pondsurfaces were covered with mats of organic detritus and deadand dying hyacinths, many of which had started to reproducevegetatively by sprouting.After six months the hyacinth matswere resprayed again, but by this time the mats only occupiedapproximately one-fourth of the pond surfaces.Ponds Two and Three were the experimental ponds treatedwith 2, -D, while Pond One was left untreated as an experimentalcontrolfor the first twelve months of the study.was selected as the controlpond since itisPond Onethe highest pond,and received no flow from either of the two experimental ponds.Pond One was chemically treated with 2,4-D on Aprilusing the same procedures as on Ponds Two and Three.15,1976,Efficient


12*1 I'1«v.-**. «?j-

13hyacinth control was achieved with this first chemicaltreatment,Retreatment (with Diquat) was not necessary until March 1977,to controlamarginalfringe of hyacinths, pennywort, andduckweed.Chemical and Physical MeasurementsThe determination of chemical parameters was done fromwater samples collected during each field sampling.Atleasttwo samples were collected from each pond at a depth of 30 cmbelow the pond surface.Water samples were collectedinclean,airtight, dark polyethylene bottles, which were labeled andimmediately refrigerated in the field.Samples were transportedback to the lab where they were analyzed within 24 hours todetermine the following chemicalparameter concentrations andphysical parameters:COpH,HCO,,and specific conductivity.,turbidity, soluble salts,These samples were fixed withapreservative (phenyl mercuric acetate), and later analyzed atthe University of Florida Soils Lab to determine the followingchemical nitrate-nitrogen, potassium,available phosphate-phosphorus,calcium and magnesium.Turbidity was measured using2100A, measuring turbidityinaHach turbidimeter, ModelNephelometric Turbidity Units(NTU's) which are equivalent to JTU's and FTU's.conductivity was determined byaYSISpecificconductivity bridge,

.Modelmeasuring specific conductivity31.mhos/cmModel2The pH was measured with.ainterms of micro-Brinkmann pH meter,PH- 102.Light penetration was determined by Secchi disk disappearanceand byaThe Secchi disk used was 25 cmlight meter.and divided(aLI-COR ModelPhotometer) was equipped withaLI-185 Quantum/Radiometer/quantum type light probeInstrument Corp.), which measures quantasyntheticalnannometersdiameter,into four alternating black and white quadrants.The light meter(Lambdainactive radiation spectrum betweenlyicroe nste nsn miiThe photosynthetiic-2Min* 00the photoand 700-1secresponse of plants for which data areavailable approximates this 00 to 700 nannometer range.Readings obtained from this light meter must be multiplied by1kO when used(since theinwater to correct for the immersion effectlight entering the diffuser scattersinalldirections).Light penetration readings using the quantum sensor were madeat10cm depthintervals throughout the water column, and alsoabove the water surface at different levels within the plantcover(when present).Visual estimates of the percentage of pond surface areacovered by water hyacinths, pennywort{Hydrocotyle spp.), andduckweed {Lenina spp.), were made on each sampling date.Theseestimates, when added together, approximate the total percentageof pond surface areas covered by floating aquatic vegetation.

15Phy toplankton and Zooplankton Collection and CountingSurface quantitative phy toplankton and zooplankton sampleswere collected on most sampling dates from the deepest accessibleportion of each pond.These quantitative samples were obtainedby pouring measured amounts of surface pond water through aPrecautions(#25 standard nylon with 83 meshes/cm).plankton netwere taken to insure that all of the plankton poured into thenet was collected in the tube attached to the bottom of the net.This was done by lowering the netinto the water(leaving themouth of the net above water), and then letting the water drainout severaltimes.Immediately after collection, all phytop lankton and zoo-plankton samples were labeled and preserved withRose bengalformalin.solutionof zooplankters,percent(prepared according to Standard1965) was added to the samplesMethods,10to facilitate the countingespecially rotifers and crustaceans.The phytoplankton samples were thoroughly mixed beforeA subsample was removed with a Pasteur pipette andcounting.placed intoacounted usingPalmer counting cell.aPhytoplankton was thenbinocular microscope atamagnification of 312X.Ten to fifty random fields within the Palmer cell were thencounted.The number of fields counted depended on the densityof phytoplankton within the sample.Phy topi ankters were counted andto species when G.W.identified to genus, andThis identification was facilitatedPrescott's Algae of the Western Great Lakes Area (1962),

16and Whitford and Schumacher's Fresh-Water Algae in North CarolinaPlanktonic diatoms were usually identified only to(1969).genus(unless the species names were obvious), since the diatomsamples were not "cleaned" to aidinidentification.These plank-tonic species were usually similar to the members of the peri-phytonic diatom populations.The average number of the individualphytoplankters perfield and total phytopl ankton were calculated.were counted and recordedpond water.interms of cells per liter of originalWhole colonies and broken coloniesVolvox, Eudorina, Synura, ) were observed andterms of individual cells to determine an averagenumber of cells per colony.This average was then used toconvert from colonies to individual cells per liter.Zooplankton samples were similarly countedRafter Counting Cell atincluded protozoans,amagnification of 156X.rotifers,countedinwereterms of total zooplanktoninaSedgewick-Zooplanktonrotifer larvae, crustaceans,crustacean larvae, and crustacean eggs.terms ofinindividuals.ColonialThe totalforms werezooplankton countsindividuals per liter oforiginal pond water.Zooplankters were usually identified only to genus, exceptinthe case of certain cosmopolitanwhich were identified to species.rotifers and crustaceansThisidentification wasaided by Ward and Whipple's Freshwater BiologyThe Sedgewi ck-Raf ter counting cell(1959).serves for countingand estimating population sizes of microzoop lankton(e.g.

1/larvae), but biases the estimationsrotifers, protozoans, naupliiof macrozooplanktonpopulations.on(e.g.For thisthe basis ofCyclops, Diaptomus, chi ronomids)reason, macrozooplankton was also determinedindividuals per liter of original pond water.These macrozooplankton counting wheel determinations were foundrealistic than counts made with the Sedgewi ck-Raf terto be morecounting cell when used for macrozooplankterslarger than naupliiAllzooplankterslarvae are reported in terms of the countingwheel population estimations, while naupliiorganisms are reportedcell.inlarvae and smallerterms of the Sedgewi ck-Raf ter countingpopulation estimations.Periphyton DiatomsPeriphyton diatoms were collected on artificial substrateby using diatometers(Biomonitor Inc., Ripon, Wisconsin) containingglass microscope slides.These diatometers are plastic frameseach holding eight glass slides which are heldwire frame with attached styrofoam floats.are heldininplace byaThe glass slidesvertical position in the water column and areaexposed to the water and available sunlight.the submerged diatometersinaThe floats keephorizontal position.The slideswithin remain vertical to reduce sediment accumulation, whichwould bury attached diatoms.Diatometers were installedon April22,were placed1975inineach of the three pondsSurface and one meter depth diatometersPonds One and Three, while Pond Two (the deepest)

received surface, one meter and two meter depth diatometers.The diatometers were secured by weights to the pond bottoms.Slides were exposed to the water column for periods ofthree or four weeks, after which time they were removed andclean slides reinserted.Removed slides were then labeledNo specialand allowed to air-dry.preserving technique wasnecessary for such dried slides since the siliceous frustulesof diatoms will not desiccate or deteriorateso obtained can be storedinair.Slidesindefinitely.Periphyton Diatom Density CountsDiatom dens ty counts were made to estimate the total numberiof diatom valves per cm2inthe three pondsThis was done by counting allthroughout the study period.diatom valves observedmicrons wide downaina(surface and depths)verticalthestrip 2.5 cm long and hGdried slide from the diatometers.Theseveral density counts obtained for each sample were averagedand numerically converted to diatom valve density per cm2ofartificial substrate.Periphyton Diatom Slide PreparationThe diatoms on one of the dried diatom slides from eachcollection were "cleaned" to oxidize all organic matter,ifnot removed, wouldidentification of the diatoms.1958) was accomplished by addinginterfere with the"Cleaning" (Van der Werff,100 mlof 30 percentHOThis

19to the air-dried diatometer slideslide was allowed to remainA smallamount of K Cr ?inina1000 mlThebeaker.this solution for 2A-*48 hours.was then added to the beaker initiating7an exothermic reaction.After the boiling subsided, this solutionwas poured intobeaker which was then filled to thea200 mltop with distilled water.of solutioninAfter four hours, the top 150the beaker was decanted.refilled with distilled water.the solutioninmlThe beaker was thenThis step was repeated untilthe beaker became colorless,indicating thatwas now absent from the solution and the diatomsthe K Crwere ready to mount on slides.The cleaned samples were then pipetted onto coverslips,using clean Pasteur pipettes and allowed to air dry.Thisprocess was repeated until visual observation indicated anadequate accumulation of diatom frustules on the coverslips.Microscopic investigation of the coverslips was necessary todetermine whether enough diatoms were on the coverslips, especiallyinsamples where sand or other siliceous deposits were present.The remaining cleaned diatom mixture left over after making theslides was put into vials and stored for future reference.After they were air-dried, the coverslips were placed diatomhot plate, and heated at approximately 5 0 C forside up ona30 minutesto evaporate moisture and to oxidize anyorganic matter.After 30 minutes, slides were removed from thehot plate with forceps andmedium onaremaininginverted onto a drop of Hyrax mountingclean glass slide.The slide (with coverslip and

23Hyrax) was then placed on the hot plate to drive the solventfrom the Hyrax.andThe slides were then removed from the hot platelabeled.Diatom Species Proportional AnalysisThe prepared diatom slides were then examined to obtainaspecies proportional analysis of the diatom populations in-habiting the three ponds throughout the sampling period.Thiswas done by counting 200 diatom valves per slide, and atleast100 valves on thoses1ides which were thinly populated with diatoms,Counting and identification of the diatom valves was done underthe oilimmersion objective(1162.5 magnifications).Identi-fication was aided by severaltaxonomic keys such as Patrickand Reimer(1930).(1966)and HustedtWater Hyacinth LengthThe length (meters) of water hyacinths was measuredPond One(controlpond) and Pond Two (treatment pond)first three months of the study1975).The(April8,1975,Severalfor theto June 29,length of hyacinth plants was measuredfrom petiole tips to root tips.ini isitu,such measurements weresimilarly obtained from each pond and averaged.

21Sediment AccumulationThe investigation of sediment accumulation beneath activelygrowing hyacinth populations, and chemicallywas nottreated populationsprime objective of this study, but an attempt wasamade to study this surface area ofSubmerged pans, each with an0.09 M2were suspended from buoys,Pond Onepond).1 ,(control pond),Ponds One and Two, oneTwo sediment pans were placedmeter below the pond surfaces.ininand three in Pond Two (treatmentThese sediment pans were placed1975,on Januaryinthe ponds on Maybefore herbicide application, and carefully removed12,1976,after all floating debris had disappeared.The sediment-water slurry taken from the pans was thenfiltered to separate the organic materialorganic material was then driedinfrom the water.Thisan oven at 60 C for 8 hoursand the dry weight determined.Stat st caii1Ana lysisThe data of this study were statistically analyzed usingtheonfacilities of the Northeast Regional Data Center locatedthe University of Florida CampusinGainesville, Florida.The procedures and programs of SAS 76-78(Statistical AnalysisSystem) were used to sort and analyze all data.Correlation coefficients were determined between all variablesusing the SAS CORR procedure.This procedure determines thecorrelation coefficient between two variables and approximates

22the significance probability of the correlation coefficient.The significance probability of a correlation coefficientprobability thatisthevalue of the correlation coefficient asalarge or larger in absolute value than the one calculated wouldhave arisen by chance,the two random variables were trulyifuncorrelated (Barr et al.1976).Only those correlations which had a significance probabilityof 10 percent or less are reported and are noted as being eitherpositively or negatively correlated.The SAS STEPWISE procedure was then used to determine whichof the collection of independent variables that showed significantcorrelations should most likely be includedThis procedureofaisusefulinaregression model.for determining the relative strengthsthe relationships between proposedindependent variables anddependent variable and involves the use of stepwise multipleregressions.This procedure first finds the single-variable model whichproduces the largest2Rstatisticmultiple correlation coefficient).2(Risthe square of theFor each of the otherindependent variables, STEPWISE calculates an F-statisticreflecting that variable's contribution to the model,were to be included.Ififitthe F-statistic for one or more variableshas a significance probability greater than the specified"significance levelfor entry"(50 percentthe variable with the largest F-statisticmodel.Inthe model.inisthis study),includedinthenthethis way, variables may be added one by one toHowever, afteravariableisadded, STEPWISE looks

23at allnothe variables alreadylonger producing a parialspecified significance levelisincludedthe model.inAny variableF-statistic significant at thefor stayingthen deleted from the model.inthe modelOnly after this check(50 percent)ismadeand any required deletions accomplished, can another variablebe added to the model(after F-statistics are again calculatedfor the variables stillevaluation processby one to the modelisremaining outside the model, and therepeated).Variables are thus added oneno variable produces a significantuntilThe process terminates when no variable meets theF-statistic.the model, or when the variable toconditions for inclusioninbe added to the modelone just deleted from1976).isit(Barr et al.The previously described procedure includes the STEPWISEoption of the STEPWISE procedure.Using the above mentioned statistical procedures, the bestpossible models for specific parameters were determined from listsof selectedindependent variables untilthe largest vari ancethe dependent variable was accounted for.inThe lists of independentvariables from which the STEPWISE procedure selected the bestpossible regression models were as follows:A.Totalphytoplankton and phy topi ankters:pH,bicarbonate,turbidity, soluble salts, potassium, available phosphate-phosphorous, calcium, magnesium, specific conductivity,water temperature, percent total vegetation cover,Secchi disk disappearance,

ACKNOWLEDGEMENTS TheauthorwouldliketothankDr.JosephS.Davisforhis Haller forhisfinancialsupportandresearchguidance .