IDENTIFICATION, CHARACTERIZATION, AND EVALUATION …srac.msstate.edu/Project PDFs/Trematodes.pdf · Identification, Characterization, and Evaluation of Mechanisms of Control of Bolbophorus-like

Identification, Characterization, and Evaluation of Mechanisms

of Control of Bolbophorus-like Trematodes and Flavobacterium

Columnare-like Bacteria Causing Disease in Warm Water Fish

16 SRAC Eighteenth Annual Progress Report, December 2005

IDENTIFICATION, CHARACTERIZATION, ANDEVALUATION OF MECHANISMS OFCONTROL OF BOLBOPHORUS-LIKETREMATODES AND FLAVOBACTERIUMCOLUMNARE-LIKE BACTERIA CAUSINGDISEASE IN WARM WATER FISH

Reporting PeriodMarch 1, 2003 - August 31, 2005

Funding Level Year 1 . . . . . . . . . . . . . . . . . . . . . . . . $224,800Year 2 . . . . . . . . . . . . . . . . . . . . . . . . $227,377Year 3 . . . . . . . . . . . . . . . . . . . . . . . . $146,770Total . . . . . . . . . . . . . . . . . . . . . . . . . $598,947

Participants Louisiana State University (Lead Institution) . . . . . . . . . . . . . . . John Hawke (Project Leader),

Richard Cooper University of Tennessee . . . . . . . . . . Andrew Mathew, Richard J. StrangeUniversity of Arkansas at Pine Bluff . Andrew GoodwinUSDA/APHIS/WS (Starkville) . . . . Brian Dorr, D. T. KingUSDA/ARS (Stuttgart) . . . . . . . . . . Andrew J. MitchellMississippi State University College of Veterinary Medicine (Starkville) . Linda Pote, Larry Hanson, Mark

LawrenceAuburn University . . . . . . . . . . . . . . John Grizzle, Joe NewtonMississippi State University, Delta Research and Extension Center (Stoneville) . . . . . . . . . . . . . . David WiseSouthern Illinois University . . . . . . . Anita KellyNorth Carolina State University . . . . Michael Levy, James Flowers

Administrative Dr. Jerald Ainsworth, Associate Dean Advisor College of Veterinary Medicine

Mississippi State UniversityMississippi State, Mississippi




SRAC Eighteenth Annual Progress Report, December 2005 17

PROJECT OBJECTIVES

1. Identify and characterize all of the life stages of the digenetic trematode (tentatively identifiedas Bolbophorus sp.) that infects channel catfish using both classical and molecular methods.

2. Evaluate integrated methods for snail control in catfish ponds.

a. Monitor populations of catfish infected with Bolbophorus spp. to document theeffect of parasite loads on growth and survival of the fish.

b. Examine the efficacy of chemical control methods on snail populations.c. Examine the efficacy of biological control methods (snail-eating fish) on snail

populations in ponds.

3. Develop and implement standardized methods for the isolation, culture, and antimicrobialsusceptibility testing of strains of columnaris-like bacteria isolated from diseased fish.

4. Characterize archived strains of columnaris-like bacteria based on the followingconventional and molecular techniques.

a. Morphologyb. Enzyme analysisc. Biochemical analysisd. Sequencing 16S ribosomal RNA and ribotyping

5. Develop challenge models for columnaris-like bacteria isolated from major warmwateraquaculture species in the southeast.

6. Using the challenge model for each species, correlate virulence with biotype and/orgenotype of columnaris-like bacteria.

PROGRESS AND PRINCIPAL ACCOMPLISHMENTS

Objective 1. Identify and characterize all of the life stages of thedigenetic trematode(tentatively identified as Bolbophorus sp.) thatinfects channel catfish using both classical and molecular methods.

Confirmation of Bolbophorus life cycle

Mississippi State University and USDA/APHIS/WS. Two studies were conducted to

confirm the life stages of Bolbophorusdamnificus in American white pelicans and itssnail host, Planorbella trivolvis. Three pelicanswere pretreated with praziquantel, challenged





with B. damnificus metacercaria to establishpatent infections, and were subsequently usedto artificially infect P. trivolvis. Catfish wereexposed to these infected snails, metacercariafrom this challenge were fed to parasite freepelicans, and patent B. damnificus infectionswere established. Each life stage of thisparasite was confirmed to be B. damnificusmorphologically and molecularly. Data arebeing analyzed on cercaria and ova shedding.

A second study was conducted to determinepotential snail hosts for B. damnificus and itslife cycle in the snail. Ova from pelicansinfected in Study 1 were used to artificially in-fect several snail species housed in aquaria at80°F. Snails were checked weekly for cercariashedding, and checked daily when they werepositive. Time and number of cercaria shedwas recorded and data are being analyzed.

Collections to Evaluate the Avian HostRange for Bolbophorus

USDA/ARS. A total of 106 aquatic birds havebeen collected and trematodes harvested fromtheir alimentary canals for identification.Some of these trematodes have potential to betransmitted to cultured fish species. In 2003,25 aquatic birds were collected including5 pelicans, 10 cormorants, and 10 great egrets.In 2004, 54 birds were collected including17 great egrets, 12 great blue herons, 11 snowyegrets, 6 cattle egrets, 6 green herons, 4 beltedkingfishers, and 1 little blue heron. In 2005,27 birds were collected including 6 beltedkingfishers, 5 white pelicans, 2 great egrets,9 black-crowned night herons, and 5 little blueherons. It appears that Bolbophorus spp. havebeen recovered only from white pelicanscollected in 2003 and 2005. The trematode

Clinostomum spp., one species of which isresponsible for the yellow grub in fish, wasfound in great egrets, great blue herons,snowy egrets, black-crowned night herons,little blue herons, and cattle egrets. The gilltrematode Centrocestus formosanus was re-covered from green herons and great egrets.Identification of the trematodes is ongoing.

Confirmation of the Definitive Final Host of Bolbophorus

North Carolina State University. AdultBolbophorus damnificus and immature Bol-bophorus sp. type 2 have been recovered andidentified from the American white pelican.Mature ovogenous Bolbophorus sp. type 2have not been recovered from any avianspecies and identification of its definitivehost remains a priority.

Mississippi State University. Birds (twoeach of American white pelicans, double-crested cormorants, great blue herons, greategrets) were live-captured in the MississippiDelta. They were individually housed inpens with recirculating water tanks and fedcatfish ad libitum daily until challenge.Birds were acclimated for at least 1 week.Fecal samples were collected daily startingat 48 hours prior to anti-helmintic treatmentand continued until necropsy. At 7 dayspre-challenge, birds were administeredpraziquantel at 34 mg/kg BW per os toeliminate all trematodes. At 7 dayspost-treatment birds were fed live fishnaturally infected with Bolbophorusdamnificus metacercariae (confirmed by aB. damnificus-specific polymerase chainreaction, PCR). Birds were necropsied 21days post-challenge, intestinal contents of





each bird were examined; all parasites wereremoved, examined microscopically, identi-fied and enumerated. A sub-sample of eachparasite type was processed for electronmicroscopy and DNA analysis.

The only bird species that shed B. damnificusova (confirmed by PCR) during the trial was theAmerican white pelican. Adult B. damnificuswere found in pelican 1 (one adult trematode)and pelican 2 (five adult trematodes). All otherbird species were negative for B. damnificus andother trematodes.

This study confirms that the American whitepelican is a host for B. damnificus. Resultsfrom this study demonstrate that artificialinfections of B. damnificus could not beestablished in double-crested cormorants,great blue herons, and great egrets.

Confirmation of Intermediate Hosts of Bolbophorus spp.

North Carolina State University,USDA/ARS, Mississippi State University.Planorbella trivolvis snails collected fromcatfish ponds in Mississippi experiencingoutbreaks of Bolbophorus-associatedmorbidity/mortality were screened for theshedding of forked-tailed cercariae in snailsshipped to North Carolina. Two morpho-logically distinct types of bolbophorid cercariaewere confirmed morphologically andgenetically utilizing species-specific PCR.These were 1) Bolbophorus damnificus, aserious pathogen of channel catfish, Ictaluruspunctatus, and 2) Bolbophorus sp. type 2, aspecies not recovered from catfish but present inseveral other fish hosts. Interestingly, severalsnails were shown to be shedding both

bolbophorid species simultaneously orsequentially. This indicated that both specieswere present in aquaculture ponds and theyutilized the same molluscan host. A manuscript"Morphological description of the cercariae ofBolbophorus damnificus and Bolbophorus sp.with notes on North American Bolbophorids"by J. R. Flowers, M. F. Poore, L. M. Pote,R. W. Litaker and M. G. Levy was submittedto Comparative Parasitology in June 2004.Information in this manuscript will allowidentification and speciation of bolbophoridcercariae based on light microscopic details.

North Carolina Planorbella duryi snails weresent to Dr. L. Pote at Mississippi StateUniversity who was successful in infectingthem with B. damnificus, indicating that theNorth Carolina snails are a permissiveintermediate host. This indicates that in thepresence of a suitable avian host, thisinfection is capable of further spread to thesoutheastern United States. Dr. Pote alsoprovided several shipments of P. trivolvispositive for B. damnificus and Bolbophorus-

RESULTS AT A GLANCE...

Ú Studies on the various life stages of

Bolbophorus damnificus have

revealed that the adult trematode

resides in the gut of the American

white pelican. The parasite has not

been found in wild cormorants, great

egrets, great blue herons, snowy

egrets, cattle egrets, green herons,

belted kingfishers and little blue

herons. Attempts to artificially infect

cormorants, great blue herons and

great egrets failed whereas the white





type 2 to Dr. Michael Levy at North CarolinaState University, and has maintained theP. trivolvis snail colony which providednegative snails for other cooperators in thisproject.

The Bolbophorus trematode has been found inwild fish species including channel catfish andseveral centrarchids in Lake Chicot, Arkansas.Metacercarie recovered from a variety of fishdemonstrated the following distribution: OnlyB. damnificus was recovered from catfish inaquaculture ponds. Bolbophorus speciestype 2 was recovered from white crappie andlongear sunfish and largemouth bass. Thefathead minnow was found to harbor bothB. damnificus and species type 2. This is thefirst finding of a B. damnificus in a fishspecies other than catfish.

Both patent and pre-patent infections ininfected snails were identified using PCR.Using PCR we also identified snailsshedding either B. damnificus or type 2exclusively. Cercariae were then fixed in hot10% neutral buffered formalin. Ten cercariaeof each type were examined for body length,body width, tail-stem length and width,furcae length and width, and oral sucker size.An additional large number of livingcercariae were held under a cover slip andexamined for the following characteristics:penetration glands, flame cells, organprimordial and tegumental spine arrange-ments. Differences between the two speciesstrongly suggest that cercariae havedistinguishing morphologic characteristics.Confirmation of these observations will beaccomplished by examining additionalcercariae during the coming season in orderto rule out individual snail variation.

Fish Challenge Trials with Bolbophorus spp.

North Carolina State University. Thepotential pathogenic effect of both trematodespecies was investigated in a series ofpreliminary experiments. Hybrid striped bass(Morone saxitalis × M. chrysops), andchannel catfish fingerlings were obtainedfrom commercial farms in North Carolinawhere Bolbophorus is not known to bepresent. Snails were divided into two groupsbased on PCR identification of theBolbophorus species that they shed. Infectionrates were based on available numbers ofcercariae less than 2 hours after emergencefrom the snails. Catfish were 2- to 3-inchfingerlings and hybrid striped bass were1.5-inch fingerlings. An aliquot of cercariaewas retained from each infection time and thechallenge species reconfirmed using PCR.These results were not available until afterchallenge was completed due to the timeinvolved in running the PCR assay.

Five groups of five bass were infected with300, 500, or 550 B. damnificus, and twogroups of bass were infected with either 40or 285 Bolbophorus sp. type 2. Eight groupsof five catfish were infected with 175, 350,637, or 700 B. damnificus cercariae/fish.Three groups were infected with 300, 700, or1,000 Bolbophorus species type 2. One groupof fish was infected with 700 cercariae of amixture of the two species due to a “switch”in the species shed by one or more snails inthis group. All fish were necropsied andmetacercariae removed and identified as totype using PCR.

All catfish infected with any dose of





B. damnificus developed the typical hem-orrhagic lesions and most died beginning onday 4 post-infection. Several fish exposed tothe lowest numbers of cercariae survived andwere euthanized 6 weeks post-infection.Although catfish exposed to only Bol-bophorus sp. type 2 failed to exhibit obvioussigns of infection such as hemorrhagic lesionstypical of a B. damnificus challenge, expo-sure to B. sp type 2 cercariae did result inthese fish going 'off feed' for several weeks. Afew degenerate metacercariae, none con-taining intact immature adult worms, wererecovered. These were identified as type 2 byPCR.

Hybrid striped bass challenged with type 2cercariae exhibited hemorrhagic lesionssimilar to those observed with B. damnificus-challenged catfish and mortality rates weresimilarly high. No morbidity or mortality wasobserved with hybrid striped bass challengedwith B. damnificus. Only Bolbophorus speciestype 2 metacercariae were recovered fromhybrid bass.

In Year 2 experimental infection of fish wascontinued with the two bolbophorid species.The potential pathogenic effect of bothtrematode species was investigated in a seriesof additional experiments complementingthose performed in Year 1.

Laboratory-reared hybrid striped bassfingerlings and raceway-reared channelcatfish fingerlings were purchased fromcommercial fisheries. Snails that hadidentified bolbophorid infections were placedin separate groups by species of cercariae asdetermined by species-specific PCR andcercariae were allowed to escape from their

snail host into the water column. Samples ofcercariae from each snail group were countedand cercarial yields were calculated. Within4 hours of escape from the snails, cercariaewere added to fish tanks containing experi-mental fish.

Groups of ten hybrid striped bass wereseparately exposed to Bolbophorus damnifi-cus cercariae and Bolbophorus sp. cercariaeat the following exposure rates: 500, 250, or100 cercariae/fish. Groups of ten channelcatfish were also separately exposed toBolbophorus damnificus cercariae andBolbophorus sp. cercariae at the followingexposure rates: 250, 100, or 25 cercariae/fish.

At 5-days post-exposure, hemorrhagic lesions,lethargy, and decreased appetite were noted inthe hybrid striped bass exposed to theBolbophorus sp. cercariae at the rates of 500,250, and 100 cercariae/fish. Mortality of thehybrid striped bass exposed to 500 cercariae/fish began at 6 days post-infection and all fishwere dead by 11 days post-exposure.Hemorrhagic lesions in hybrid striped bassexposed to 250 and 100 cercariae per fishdisappeared by 19 days post-exposure;however, large bumps under the skin werenoted. None of the hybrid striped bassexposed to Bolbophorus damnificusdeveloped lesions.

Conversely, channel catfish reacted to theBolbophorus damnificus cercarial exposures.Hemorrhagic lesions and bumps wereobserved on catfish exposed to B. damnificuscercariae at the rates of 250 and 100cercariae/fish in the morning of day 9 post-exposure. In the afternoon of day 9 post-exposure, three catfish exposed to 250





cercariae/fish and one catfish exposed to 100cercariae/fish had died. Later, one catfishexposed to 250 cercariae/fish and one exposedto 100 cercariae/fish died at 14 and 23 dayspost-exposure, respectively. Some of thecatfish exposed to 250 cercariae/fishdeveloped exophthalmia and abdominaldistension. None of the catfish exposed to theBolbophorus sp. cercariae developed lesions.Also, lesions were not seen in the catfishexposed to Bolbophorus damnificus cercariaeat the rate of 25 cercariae/fish.

The pathological consequences ofB. damnificus infection in channel catfishwere less severe compared with those seen inpast experiments. This may be due to thelarger size of fingerlings or the presence offewer non-trematode pathogens. In previouschallenges, fingerlings were collected fromponds and may have harbored other path-

ogens, whereas for this experiment wecollected swim-up fry hatched in well waterand laboratory-reared the fish.

These results demonstrate a high degree ofspecificity for the intermediate hosts for thesetwo bolbophorid trematodes. Bolbophorusdamnificus caused lesions only in catfishwhere bolbophorus sp. caused lesions only inhybrid bass.

Fish growth rates at the different parasite-exposure rates are currently being statis-tically analyzed. The number of encysedmetacercariae for each fish (for eachchallenge parasite species and number) willalso be determined. Development of methodsfor quantification and estimation of parasiteloads in molluscan populations are also inprogress.

Objective 2. Evaluate integrated methods for snail control in catfish ponds.

Objective 2a. Monitor populations of catfish infected with Bolbophorus sp.to document the effect of parasite loads on growth and survival of the fish.

Mississippi State University. Laboratory andfield studies were conducted to evaluate theeffects of sub-lethal trematode infections ongrowth, performance and disease resistance ofchannel catfish fingerlings. Trematode in-fections were established in populations offish stocked in four, 0.1-acre ponds. Areservoir of trematode-infected snails wasmaintained in recirculating 300 gallon tankslocated on the bank of each pond. Pond waterwas recirculated through each tank at a rate of2 gallons per minute. The effluent (containingBolbophorus cercariae) from the tank was

directed back into the pond and served as thesource of infection. Four additional pondswere used as control ponds. After 40 days,each population of fish was sampled toevaluate health status and 120 fish from eachpond were transferred to 30-gallon aquaria toevaluate growth rates under controlled labora-tory conditions using well water free ofBolbophorus cercariae. Only fish containingvisible cysts (1 to 5 cysts per fish) wereselected and used to evaluate growthpotential. Fish were acclimated to laboratoryconditions at 31 to 32°C for 3 weeks before





the start of the study. Following the con-ditioning period, fish were fed once daily for9 weeks. Total weight gain, percent weightgain, specific growth rate, and feed efficiencywere used to evaluate growth. Evaluation of health and growth ofchannel catfish continually exposed to thecercarial stage of Bolbophorus damnificusthroughout a production cycle Mild trematode infections were established inpond populations of experimental fish byexposing fish to trematode cercaria. The percentof infected fish in each pond ranged between20.4% and 1.6%. Mortalities directly related totrematode infections were not observed.Edwardsiella ictaluri and Flavobacteriumcolumnare infections were diagnosed from allpopulations of fish and no differences inmortality were observed between trematodeinfected and non-infected fish. At the end ofthe production cycle, trematode infected fishconsumed approximately 40% less feedcompared to fish in control ponds. Evaluation of health and growth of fish that have been infected with Bolbophorusdamnificus cercariae by a single-pulseexposure At the start of the acclimation period,trematode infected fish were significantlysmaller compared to fish collected fromcontrol ponds. No differences in any of themeasured parameters were observed betweentrematode-infected and non-infected fish atthe end of 9 weeks. Although the final weightof trematode-infected fish was numericallysmaller than control fish, percent weight gainand specific growth rate demonstrated a

tendency towards compensatory growth oftrematode-infected fish. Feed efficiency(0.86) was identical between treatmentgroups. Data indicates that once fish areremoved from the source of infection, chronictrematode infections do not affect the growthpotential of channel catfish.

Evaluation of health status and growthpotential of channel catfish fingerlingsinfected with Bolbophorus damnificusunder controlled laboratory conditions

Trematode infections were established underlaboratory conditions by placing fingerlings intriplicate tanks containing Planorbellatrivolvis snails shedding cercariae. Fish wereleft in the tank for 24 hours and snails wereshedding cercariae at a rate of 770 ± 82 per24 hours. Unexposed fish were maintained inthree tanks under similar conditions. Fromeach tank, trematode-infected or non-infectedfish were transferred to six aquaria (30fish/aquaria). Three aquaria from eachreplicate treatment tank received 7.5 × 105

CFU E. ictaluri/mL of water for 30 minutes(Bolbophorus-ESC and ESC-only groups).Fish in the remaining three aquaria were notexposed to E. ictaluri and served asBolbophorus-only and negative controlgroups. The later two treatment groups wereused in a second study and were challengedwith E. ictaluri 28 days after exposure toBolbophorus sp. cercariae.

No mortalities were observed in theBolbophorus-only and negative control groups.Twenty-one days following exposure toE. ictaluri, the percent cumulative mortalitieswere 84.1 ± 16.2% in the Bolbophorus-ESCtreatment and 45.9 ± 3.2% in the ESC-only





treatment. Mortalities were significantlydifferent between the two groups. In thesecond study, when E. ictaluri exposure wasdelayed 28 days following Bolbophorus sp.infection, there was no difference inmortalities between the ESC-only (17.8 ±4.0%) and combined Bolbophorus-ESC (21.5± 1.7%) exposed groups. Apparently, oncefish are removed from the source of additionalinfections, chronic trematode infections donot increase the susceptibility of fish to ESC.

Findings in these studies have significantimplications for management strategies tocontrol losses associated with trematodeinfections. Data collected from laboratory andfield trials indicated that mild sub-lethalactive trematode infections, commonlyobserved in channel catfish productionsystems, can significantly reduce productionby reducing feed consumption and increasinglosses associated with ESC. These studies alsoindicated that the presence of fully developedmetacercariae does not compromise growthand health status of fish. These data supportthe contention that the deleterious effects oftrematode infection are associated withpenetration of the parasite and initial stages ofencystment. Findings point to the need forincreased surveillance for this disease and thebenefit of initiating management protocols at

the earliest stages of infection. In addition,breaking the trematode’s life cycle by movingfish to non-infested water or by eradicatingsnails in the pond will eliminate the adverseeffects associated with this disease.

Evaluation of the farm-wide economicimpact of Bolbophorus damnificusinfections of channel catfish

The economic impact of trematode infectionswas evaluated by conducting a disease-monitoring and production-efficiency studyon a commercial catfish operation with pondscontaining trematode-infected fish. Fish weresampled from each food fish production pondand examined for the presence of cysts thatcontain the metacercariae. Each pond wasplaced into one of four categories based on thepercentage of infected fish in the sample.Ponds with trematode-infected catfish wereplaced into categories of light, moderate, orsevere when the percentage of trematode-infected fish in the sample ranged from 0% to33%, 33% to 66% or 66% to 100%, respec-tively. Ponds that did not contain trematode-infected fish were categorized as negative.Production records were grouped by in-festation level for analysis. Of the 40 pondpopulations sampled, 17 were categorized as


Ú Mild, sub-lethal trematode infections can

significantly reduce catfish growth by

reducing feed consumption and

increasing mortality associated with

concurrent bacterial infections.


Ú The presence of fully developed

Bolbophorus metacercariae does not

affect growth or health of catfish. The

deleterious effects of this infectious

agent are therefore associated with

penetration of the parasite and initial

stages of encystment.





negative, 6 as light, 6 as moderate, and 11 assevere. Fish from trematode-positive pondsconsumed significantly less feed compared tofish from ponds that were categorized astrematode-negative. Fish from ponds in thetrematode-negative category consumed onaverage 73.4 pounds/acre per day, and fishfrom ponds categorized as light, moderate andsevere consumed 62.2, 47.5, and 47.2pounds/acre per day, respectively. Similarly,production decreased as severity of infectionincreased. Compared to trematode-negativeponds, ponds in the light, moderate and severecategories produced 16.8%, 36.4%, and

44.5% fewer pounds of fish per acre,respectively. Net returns from ponds in thelight category were reduced by 80.8% andproduction from ponds in the moderate andsevere categories were not shown to covervariable costs of production. Ponds in themoderate category produced a net loss of$506 per acre and severe ponds produced anet loss of $631 per acre. These data from acommercial setting support results fromexperimental ponds and demonstrate thatBolbophorus infestations, regardless of sever-ity, are a significant risk to commercialproduction of channel catfish.

Obective 2b. Examine the efficacy of chemical control methods on snailpopulations.

USDA/ARS. A total of seven trials wereconducted from 2003 through 2005 to test theeffectiveness of pond-shoreline treatments incontrolling aquatic snails. Initially, four trialswere conducted to compare a slurried-hydrated lime treatment with an establishedcopper sulfate treatment. Copper sulfate andhydrated lime were applied at 4 pound and 80pounds, respectively, per 100 feet of shore-line in a 6-foot swath. Trials were run underconditions of variable wind speed (0 to 16mph) and treatment temperature (24 to 32°C).Both treatments effectively lowered the snailpopulations in the test cages. It appears thatcopper sulfate was more effective than limein most trials, that hydrated lime treatmentsappeared to increase in effectiveness athigher temperatures (32°C vs 25°C), and thatstrong winds negatively impacted bothtreatments. Snail survival under all con-ditions in the four trials ranged from anaverage of 3.4% to 27.8% and 10.4% to

41.5% for copper sulfate and hydrated lime,respectively.

The goal of the second part of the studychanged from comparing lime and coppersulfate to optimizing the hydrated limetreatment. Trials 5, 6, and 7 were conductedusing only hydrated lime treatments attemperatures of 24 to 26°C and under lowwind conditions. In Trial 5, the rate ofhydrated lime was increased to 175 poundsper 100 feet of shoreline in a 6-foot swath. Atthat rate, effectiveness was increased and snailsurvival was less than 2%. Further optimi-zation of the hydrated lime treatment wasmade by narrowing the treatment swath to3 feet and reducing the amount of lime to 80and 100 pounds per 100 feet of shoreline.Average snail survivals for the 80 and 100pound treatments were 13.3% and 2.7%,respectively. The optimum pond shorelinetreatment with slurried hydrated lime is





100 pounds of lime per 100 feet of shorelineapplied in a 3-foot swath.

Mississippi State University. The toxicity ofcopper sulfate to ram's horn snails wasevaluated by establishing the 24-hour LC50using Spearman-Karber analyses. Tests wereconducted in 300-mL glass containerscontaining 200 mL of pond water (alkalinity =

3235 mg/L as CaCO , hardness = 300 mg/L as

3CaCO , pH = 7.5, temperature = 23°C). Testconcentrations of copper were arranged in ageometrically spaced dilution series. Each testconcentration consisted of 4 replications with 5snails per replication. Copper sulfate granuleswere dissolved in distilled water and deliveredas a solution. After 24 hours, snails wereremoved from the test solution and placed infresh, untreated water. End-points for the testswere death of the snails as determined by anadditional 96-hours post-test observation periodto confirm mortality. The effect of temperature(15, 20, 25, and 30°C) and alkalinity (0, 50, 100,

3and 200 mg/L CaCO ) on the toxicity of copperto snails were also evaluated.

Laboratory tests showed copper sulfatecrystals had a 24-hour LC50 of 0.6 mg/L Cuand, based on the alkalinity of the test water,was below the level considered toxic to fish.Alkalinity at the levels tested (0 to 200 mg/L

3CaCO ) was not shown to effect to thetoxicity of copper to snails. The LC50concentration at an alkalinity of 0 mg/L

3CaCO was 0.52 mg/L Cu versus 0.67 mg/LCu at an alkalinity of 200 mg/L. Althoughthere appeared to be a trend in the LC50values toward decreasing toxicity with in-creasing alkalinity, these differences were notstatistically significant. Analysis of data fromthis study showed a significant linear rela-tionship between temperature and LC50values for copper. As temperature increasedfrom 15°C to 30°C, the LC50 valuesdecreased from 1.1 mg/L to 0.18 mg/L Cu,representing a ten-fold increase in toxicity.This would be an important considerationwhen treating ponds with copper sulfate withrespect to both snail and fish toxicity.

Three dose-titration trials were performed todetermine the copper concentration required tokill snails under field conditions. Two trialswere conducted in plastic tanks containing 200gallons of pond water. Fish and snails wereplaced in three replicate tanks and dosed with asolution of copper sulfate at 0 (control) 1.25,2.5, 5.0, and 10.0 mg/L Cu during the first trialand 0 (control), 0.375, 0.75, 1.25, and 2.50 mg/LCu during the second trial. Each tank consistedof 20 snails confined in wire mesh cages and 10channel catfish fingerlings to evaluate thetoxicity of the treatment dose to fish.

The third trial was a non-replicated studyconducted in 0.25-acre ponds containingapproximately 1 acre-foot of water. In eachpond, three sample sites were evenly dis-tributed along the center long axis of thepond. At each sample site, 20 snails wereconfined in cages located at the surface andbottom of the pond. The toxicity of copper to


Ú A shoreline treatment with slurried

hydrated lime applied in a 3-foot swath

at 100 pounds of lime per 100 feet of

shoreline reduced snail populations by

over 95%.





fish was evaluated in the 0.25-acre ponds byconfining 10 channel catfish fingerlings infloating net-pens in close proximity to each snailsampling site. Each pond received a dose of 0(control), 1.25, 2.5, 5.0, and 10.0 mg/L Cu byevenly applying the copper sulfate solution (200gallons/pond) around the margins of the pond.Twenty-four hours after treatment, test snailswere transferred to 1-L containers (containingthe appropriate test water) and transported to thelaboratory and observed for 72 hours. Twicedaily, dead snails were removed and placed in aseparate container containing untreated pondwater. Dissolved oxygen concentrations in thetest water and fish mortality were observed for96 hours.

Titration trials in tanks and ponds werecomparable, with laboratory toxicity trials andindicated the minimum effect dose (snail mor-tality >90%) of copper sulfate ranged between0.75 and 1.25 mg/L Cu. Fish mortality was notobserved at or below 1.25 mg/L Cu in the pondtank studies. Fish mortality (average mortality5.3%), however, was observed at one of thethree sample locations in one of the replicateponds following treatment with 1.25 mg/L Cu.

Based on the results of the dose-titration trials,triplicate 0.25-acre and duplicate 10.0-acreexperimental ponds were treated with 0.75 and1.25 mg/L Cu to verify the minimum effectivedose. Sample site configurations and treatmentapplications for tests conducted in the 0.25-acreponds are described above. Sample siteconfigurations for tests conducted in 10-acreponds varied to accommodate pond size. Each10-acre experimental pond contained 12 uni-formly distributed sample sites consisting of20 snails confined at the surface and bottom ofthe pond. An additional 20 snails were confined

along the pond bank at equal intervals. Each10.0-acre pond was managed as a commercialproduction pond and contained approximately3,500 (average estimated size = 1 pound) fishper acre. No mortality or signs of disease werenoted before any of the tests were conducted.Toxicity of the copper treatment to fish wasevaluated by observing the pond stock forbehavioral indicators of toxicosis and mortality.For each test conducted, a non-treated pondcontained similar sample site configurations andserved as a control.

Replicate single-dose pond treatments verifiedthat treatment doses of 0.75 and 1.25 mg/L Cuwere effective in killing snails. Average snailmortality in trials conducted in the 0.25-acreponds ranged between 98.0% and 95.5% at thelow treatment doses and was 100% at the hightreatment dose. Fish mortality was observed in1 of the 3 replicate ponds at each treatment dose.Similar results with respect to snail toxicity wereobserved in the single dose toxicity trialsconducted in 10-acre ponds. Average snailmortality of the replicate trials at each samplelocation ranged between 92 and 98% followingtreatment, with 0.75 mg/L and between 98 and100% following treatment with 1.25 g/L Cu. Incontrast to the 0.25-acre pond trials, no fishmortality or behavioral signs of toxicosis wereobserved following treatment. In all pond trials,dissolved oxygen depletions were not observedfor up to 168 hours after treatment. Fishmortality in the 0.25 acre ponds may be causedby exposure of confined fish to high concen-trations of the applied chemical before it wascompletely mixed with the pond water.

Treatment efficacy was then evaluated in a13.0-acre commercial channel catfish produc-tion pond. Moderate numbers of snails were





observed along the margins of the pond.Seven sample sites (surface and bottom cagescontaining 20 snails each) were placed aminimum of 50 yards from the pond bank andwere distributed randomly across the pond. Anequal number of cages containing 20 snailseach were placed along the margins of thepond. In addition, natural snail populationsalong the margins of the test pond weresampled at five locations before and aftertreatment. A 20-foot section of pond bankconsisting of uniform vegetation and leveeslope was marked and divided into 2 equalsections. Snails were collected from a 10-footsection of the sample site before the pond wastreated, and the remaining section wassampled 24 hours after the pond was treated.Snails collected from each section were placedin 4-L containers and transported to thelaboratory for observation. Live snails werecounted to estimate viable snail numbers andused to determine the number of snails perfoot of pond bank in the sampled area.Production fish were monitored for behavioralsigns of toxicosis and mortality. Snailsconfined in a similar configuration in anadjoining pond served as a control.

Results of the commercial field trial werecomparable to tests conducted in the experi-mental ponds where application of coppersulfate at 1.25 mg/L was shown to beeffective in killing snails. Average mortalityof snails confined in cages ranged between95.4 and 97.7%. The treatment was alsoshown to be effective against natural popula-tions of snails along the margins of the pond.The average number of snails per foot of pondbank decreased from 21.5 snails to 0.18 snails24 hours after treatment, representing a 99%

reduction in viable snail populations in thehabitat along the pond bank.

Treatment of the commercial pond resulted inchanges in fish behavior and mortality thatwas likely related to the copper treatment.Within 4 hours of treatment application, anincrease in the number of moribund fish wereobserved. Affected fish appeared lethargic orexhibited a spiraling swimming pattern.However, it is not thought that these ob-servations were solely related to the chemicaltreatment. Prior to treatment, moribund anddead fish were present in the pond that wasdiagnosed with bacterial (Edwardsiellaictaluri and Flavobacterium columnare) andparasitic infections (Bolbophorus sp.). Mori-bund fish also exhibited clinical symptomsconsistent with visceral toxicosis of catfish.Mortality rates in the pond were characterizedas chronic and were estimated to be 150 to200 fish per day. Farm management estimatedtotal mortality in excess of 20,000 pounds.Following treatment, fish mortality increasedwithin the first 24 hours. It was estimated thatapproximately 2,000 pounds of fish were lostfollowing treatment, however, a majority ofthe fish had clinical signs of ESC. In additionto infectious disease, analysis of water quality2 hours after treatment revealed low chloride tonitrite ratios and examination of fish gills andblood indicated acute nitrite toxicosis. On the


Ú A low-dose, full-pond treatment with

copper sulfate was developed that

safely eradicates the trematode's

intermediate host—the ram's horn snail.





day of treatment, salt was added to the pondwater and, after the first 24 hours, the daily

mortality rate returned to pretreatment levels.

Objective 2c. Examine the efficacy of biological control methods(snail-eating fish) on snail populations in ponds.

Effect of diet conditioning on prey selectionby blue catfish and redear sunfish

Mississippi State University, SouthernIllinois University. Four aquaria were stockedwith 300 juvenile blue catfish and four addi-tional aquaria were stocked with 300 juvenileredear sunfish. The fish were given an initial“conditioning diet” which consisted of only oneof the following: fish food – insect larvae, ram'shorn snails or red-rimmed melania snails. After2 weeks of feeding the “conditioning diets,” 100fish of each species were stocked into eightseparate aquaria (16 aquaria total) and offered aknown amount of their conditioning diet and aknown amount of one of the other conditioningdiets used above. Prey selection was determinedby the frequency of selection of the variousdiets. A Chi-square contingency test was usedto determine the influence of conditioning onfood selection. A two-tailed test of binomialproportion was used to examine the signifi-cance of food preference in each conditioningexperiment. Because fish that consume largeamounts of food may bias results, the percentageof the total number of prey items the percentageof food for each fish was determined either byvideo or by examination of the stomach con-tents. For video analysis, a video camera wasused to record a 2-hour segment of feeding. Thedata obtained was analyzed using the Observer(1997), a computer program specificallyprogrammed for behavior analysis. Dataobtained from this program was analyzed

using a Chi-square contingency test.

Prey selection studies with the blue catfishrevealed that regardless of the training orconditioning diet, blue catfish readily

converted to catfish feed when it was offered.Redear sunfish preferentially consumed snailswhen available. Even redear sunfish trained toeat commercial fish food readily consumedsnails and chironomids when available.

Determination of the ability of redearsunfish to withstand conditions ofcommercial catfish culture

Southern Illinois University. In the spring,redear sunfish (small, medium, and large)were stocked at a rate of 300 fish/acre intofour, 0.0-acre experimental ponds (12 pondstotal) stocked with channel catfish at aproduction rate of 8,000 pounds/acre. Waterquality variables such as dissolved oxygen,


Ú Ponds stocked with redear sunfish had

significantly fewer snails than ponds

without sunfish. Redear sunfish

preferentially consumed snails and

midge larvae when available, even

when trained on pelleted diets prior to

stocking ponds.





temperature, alkalinity, carbon dioxide,ammonia, and pH were measured on a dailybasis. The number of dead fish (catfish andsunfish) was recorded. In September, theponds were seined and harvest size channelcatfish removed. The number of survivingindividuals of each species was recorded. Theponds were under stocked with 5-inchfingerlings to replace the fish removed formarket. Survivability of redear sunfish wasanalyzed to determine if any correlation withwater quality variables exist. The number ofredear sunfish that died due to seining wasrecorded. In the spring of 2004, the pondswere seined and all redear were counted andweighed. Harvest size catfish were removedand 5-inch fingerlings were stocked to replacethose removed.

In the fall of 2004, redear sunfish and channelcatfish were removed from ponds. Thenumber and type of snails within a one metertransect were counted and recorded. Onehundred channel catfish from each pond weresampled and analyzed for the presence oftrematodes on gills and in the flesh. Thisstudy was repeated in 2005.

In the 2004 study, ponds containing redearsunfish had significantly fewer snails thanponds without sunfish. The number and typeof snails remaining in the ponds did not differsignificantly when medium size or large

sunfish were stocked. Redear sunfish trainedto eat snails did not remove significantlyhigher numbers of snails than fish not trainedor conditioned to snail diets. Survival of theredear sunfish was 100 percent in all ponds.The incidence of trematode infestation wasstill evident on channel catfish. Approxi-mately, 25% of the catfish had trematodes onthe gills and in some cases within the flesh.

In the 2005 study, the water temperatures in theponds at Southern Illinois University averaged5°C higher than in 2004. Catfish survival in allponds was greater than 98%. Survival of redearsunfish was strongly correlated with thedecreased dissolved oxygen and increasedtemperature in the ponds. Ponds that routinelyhad dissolved oxygen concentrations of 3 mg/Land temperatures of 32°C at sunrise hadsignificantly higher mortality rates than pondswith higher morning dissolved oxygen con-centrations. No significant correlations betweenalkalinity, carbon dioxide, ammonia, nitrite, orpH on survival of redear sunfish were observed.The incidence of trematode infestation was stillevident on channel catfish. Approximately,31% of the catfish had trematodes on the gillsand in some cases within the flesh. The increaseobserved in the number of channel catfish withtrematodes visible on the gills or in the flesh islikely due to the decrease in redear sunfishnumbers in the ponds due to mortalities.

Objective 3. Develop and implement standardized methods for theisolation, culture, and antimicrobial susceptibility testing of strains ofcolumnaris-like bacteria isolated from diseased fish.

Louisiana State University. Various agarmedia were evaluated for optimum primary

isolation and maintenance of Flavobacteriumcolumnare. Media under investigation included





both selective and non-selective cytophaga agar(CA), Hsu-Shotts (HS), Shieh (S), tryptoneyeast extract (TYE), dilute Mueller Hinton(DMH) and Flavobacterium columnare growthmedium (FCGM). Media were made selectiveby the addition of 5 µg/mL neomycin and 200units/mL polymixin B. For primary isolation themedia were prepared as agar plates and formaintenance the media were prepared as 20-mLslants in 50-mL tubes with 1 mL of saline addedto preserve moisture. For the evaluation ofprimary isolation media, a standardized mixtureof F. columnare, Edwardsiella tarda, E. ictaluri,Aeromonas hydrophila, and Streptococcusdifficilis was prepared. This mixture wasdesigned to mimic the mixture of aquaticbacteria that might be present in contaminatedexternal sites such as the gills and skin ofdiseased fish. The mixture was inoculated ontothe various test media to evaluate their ability toproduce pure colonies of F. columnare whileinhibiting contaminating bacteria.

Selective cytophaga agar (SCA) performedthe best as a primary isolation medium forisolation of columnaris from a mixedinoculum of aquatic bacteria. The remainder

of the media were ranked as follows: (2) SS(3) SHS, (4) DMH and (5) FCGM. BothDMH and FCGM produced no isolatedF. columnare colonies.

For maintenance following isolation, TYEslants performed the best with some culturesmaintaining viability as long as 84 days. Theremaining media were ranked as follows:(2) CA, 52 days (3) DMH, 47days (4) HS,34 days (5) S, 32 days and (6) FCGM, 23 days.

Some of the above-mentioned media wereevaluated as broths for batch culture ofF. columnare. A 40-mL volume of media wasinoculated with 200 µL of a McFarland #5standard inoculum and growth performancemeasured by colony forming units (cfu) per mLand absorbance at 600 nm following 24 hours ofincubation at 28°C. Flavobacterium columnaregrowth medium (FCGM) out-performed otherformulations tested producing mean absorbancesof 0.2377 and colony counts of 2.2 × 10 /mL.9

The clumping of cells, which is a problem inother broth media, was avoided in FCGM. Theremaining broth media were ranked as follows:(2) Shieh (3) cytophaga and (4) DMH. For asummary of broth culture results see Table 1.

For disk-diffusion antimicrobial susceptibilitytesting, dilute Mueller Hinton (DMH) platesprepared with different levels of agar andnutrient were evaluated for clarity andconsistency of zone size. To insureuniformity, one lot of each of fiveanti-microbial agents, one lot of MuellerHinton medium, and one lot of equine serumwere used in all disk diffusion testevaluations. The anti-microbial disks (BBL)chosen were Sulfamethoxazole : trimethoprim(SXT 25 µg), Sulfadimethoxine : ormetoprim


Ú Selective cytophaga agar SCA has

performed the best as a primary

isolation medium in preliminary tests in

isolation of Flavobacterium columnare

from contaminated sites such as the gills

and skin. For maintenance following

isolation, tryptone yeast extract TYE

medium as a moist slant, held cultures

viable for as long as 84 days. For large

batch broth culture, FCGM outperforms

other formulations tested.





(PRI-MOR 25 Fg), Oxolinic acid (OA 2 µg),Oxytetracycline (T 30 µg), and Florfenicol(FFC 30 µg). The basal MH broth used was(Difco) lot # 3126187, and the degranulatedagar used was (Difco) lot #3265229. Theconcentrations of MH broth evaluated were 3 g,3.5 g, 4 g, and 5 g/L. The agar concentrationsthat were evaluated were 9 g, 12 g, 15 g, and17 g/L. Varying concentrations of MH brothwere used to determine the optimum amount ofnutrients for F. columnare growth. The varyingagar levels were evaluated for their effects onzone appearance by limiting the gliding motility

of the bacterium. Each medium was evaluatedfor thickness of growth, distinct zones, anduniformity of zone margins. Once the optimumconcentration of broth to agar was determined,5% equine serum lot #ANE18713 (Hyclone,Logan, Utah) was evaluated as a replacementfor the more expensive fetal calf serum as agrowth supplement for F. columnare cultures onMH agar plates.

The optimum media formulation for sus-ceptibility testing was determined to contain4 grams M-H broth and 17 grams of agar perliter with 5% equine serum. This medium

gave the highest bacterial growth and allowedfor better definition of zones of inhibition.Equine serum improved the growth ofF. columnare cultures on dilute M-H agar, butnot significantly different from fetal calfserum. Because of availability and cost ofequine serum compared to fetal calf serum itwas determined to be a suitable enrichmentfactor for the improved medium.

Objective 4. Characterize archived strains of columnaris-like bacteriabased on conventional and molecular techniques.

Morphologic and biochemical analysis ofcolumnaris-like bacteria

Louisiana State University. Forty-nine strains

of columnaris-like bacteria archived in theLADL and UAPB collections were analyzedusing conventional biochemical testing in testtubes, the API 20E system, the API NE system

Table 1. Growth of F. columnare invarious broth media at 28°C for 24 hours.Data is presented as absorbance at 600nmand by colony forming units (cfu) per mL.

Test Medium Mean MeanAbsorbance Colony(600nm) counts

(CFU/mL)

DMH* 0.0858 1.3 × 107

Cytophaga* 0.1646 6.3 × 107

Shieh* 0.2078 3.1 × 108

FCGM* 0.2377 2.2 × 109

* Indicates significance at P = 0.05


Ú An improved medium has been

developed for antimicrobial susceptibility

testing of Flavobacterium columnare.





and the API ZYM system. These strains wereobtained from a larger pool of yellow pigmentedcolumnaris-like isolates that were subjectedto screening procedures that involved aspecies-specific PCR (Bader et al. 2003), thefive point characteristics of Griffin (1992) andthe physiological characteristics of Bernardetand Grimont (1989). To conform to theGriffin screen, the bacterium must be shownto satisfy the following requirements:production of flat, spreading, yellow, andrhizoid colonies on cytophaga agar, growth inthe presence of neomycin and polymixin B,production of gelatin degrading enzymes,binding of congo red dye to the colony, andproduction of chondroitin sulfate A de-grading enzymes. Isolates were also tested forthe presence of flexirubin-type pigments usingthe potassium hydroxide (KOH) methodoutlined in Bergey's Manual of DeterminativeBacteriology 9th Edition (1994). The goal ofthis part of the project is to determine ifF. columnare can be identified using con-ventional and commercially available bio-chemical testing schemes by labs that may nothave molecular capabilities. Prior to the startof our study, 48 of 49 archived strains ofcolumnaris-like bacteria conformed to theGriffin screen and were confirmed by PCR asF. columnare.

Results indicate that F. columnare should beshown morphologically to be a long, thin,gram negative rod (3 to10 micrometers lengthand 0.3 to 0.5 micrometers in width) withgliding and flexing motility, and form rhizoid,yellow-pigmented colonies on agar plates.Morphology of the cells and colonyappearance is slightly variable among strainsand is influenced by growth conditions andage of the culture. The isolate should be a

strict aerobe, should not produce acid fromcarbohydrates, and should be cytochromeoxidase and catalase negative. Negativereactions were obtained in the GMD and TSIagar tests due to the lack of acid productionfrom carbohydrates. The API 20E, API NEand API ZYM systems (bioMérieux-Vitek)were examined for usefulness in theidentification of F. columnare. The API 20Eand API ZYM systems were determined to beinadequate due to the lack of positivereactions on the strips. The API NE was veryuseful producing an adequate number ofpositive reactions. Isolates of F. columnareuniformly gave reactions resulting in the APINE code number 0441455 at 24 hours. Thepositive reactions in the API NE strip were asfollows: esculin, D-glucose, L-arabinose,potassium gluconate, capric acid, malic acid,and citrate.

Adhesion to plastic, cultured cells, andisolated gills

Auburn University. Six types of plasticmultiwell plates (BD Biosciences, FranklinLakes, NJ) were compared for use in a bacterialadhesion assay. Two hours after washedF. columnare cells were added to the wells,there were significant differences among theplates. The same results were obtained with twoisolates. Adhesiveness of F. columnare wasgreater for bacteria grown in Hsu-Shotts brothrather than in Shieh broth. The addition ofcalcium and magnesium to water used in theadhesion assay increased the adhesiveness ofone isolate of F. columnare (PL-04-02) but hadno effect on another isolate (PB-02-110). Otherwaters tested, which had high concentrations ofNaCl, tended to reduce the adhesiveness of theisolates tested.





Isolates of F. columnare tested with themultiwell plate assay had a wide range ofadhesiveness to plastic. As additional isolatesare obtained for testing, these results will becompared to other types of results obtained byinvestigators at other institutions to determine ifthere is a relationship between adhesiveness andother characteristics. Attempts to quantify theadhesion of F. columnare to cultured cells washindered by the adhesion of the bacteria to theglass or plastic substrate used for cell cultureand by problems with accurate counting ofbacteria stained by conventional methods. Toovercome these problems, antibodies againstF. columnare were made in rabbits as a firststep in development of antibody-based methodsfor bacterial quantification.

Three types of assays for adhesiveness ofF. columnare (isolated fish gills, larval fish,and cultured cells) were developed and thencompared with a multi-well plastic plateassay. The assay with plastic plates waspreviously found to be useful for quanti-fication of F. columnare adhesiveness.

For the gill assay, gills were dissected fromchannel catfish, bluegill, and common carp. Foreach fish, one section of gill was used as acontrol and two sections were exposed toF. columnare. After a 10-minute exposure tobacteria, gills were rinsed twice andhomogenized. Plate counts of serial dilutions ofthe gill homogenate were used to quantify theF. columnare adhering to gills. There was asignificant difference among F. columnareisolates adhering to for gills from bluegill(8 isolates) and common carp (3 isolates). Onlytwo isolates were tested with normal channelcatfish gills; however, there was a significant

increase in the number of bacteria adherent togills of channel catfish with proliferative gilldisease or with Aeromonas infection.

A larval zebrafish assay for adhesiveness ofF. columnare was developed. Whole fish wereexposed to F. columnare for 1 hour, rinsed for2 minutes, and then homogenized. The adherentF. columnare were enumerated by plate counts.There were significant differences in adhesive-ness of the 11 isolates of F. columnare evaluatedwith this assay. The CFU/mg of fish varied overa 150-fold range for the isolates tested.

A cultured-cell assay was used to examineadhesiveness to EPC cells. The cell culturemedium was removed from the cells andreplaced by a suspension of F. columnare ineither phosphate-buffered saline (PBS) or wellwater. After incubation at 30°C for 10 minutes,plate counts were used to determine the numberof adherent bacteria. Seven isolates ofF. columnare were tested for adhesiveness inhypotonic conditions (well water), and there wasno significant differences among isolates. Fourisolates were tested in PBS, and one isolate hadsignificantly reduced adhesiveness. The assaywith EPC cells was not satisfactory because thehigh concentration of sodium chloride in PBSreduces the adhesiveness of F. columnare. Theuse of fresh water during the incubation of EPCcells with F. columnare resulted in swelling ofthe EPC cells because of the hypotonicconditions. The variation in adhesiveness amongthe F. columnare isolates tested was small forthe cultured-cell and gill assays.

Based on adhesion to plastic, the mean adhesionof 11 isolates of F. columnare isolated fromexternal organs was higher than the mean for





11 isolates from internal organs. The virulenceof 13 isolates (as reported by Thomas-Jinu andGoodwin in 2004) was not correlated withadhesion to plastic. Adhesion of 11 isolates ofF. columnare to larval zebrafish was also notcorrelated with the virulence reported byThomas-Jinu and Goodwin (2004). The lack ofcorrelation between virulence and adhesioncould be the result of additional passages inculture between the determination of virulenceand the adhesion evaluations. There could alsobe differences in virulence or adhesion related tospecies of fish. Additional experiments willevaluate adhesion and virulence in simultaneousexperiments.

Molecular identification of columnaris-like bacteria using rapid sequence analysisof a portion of the 16S ribosomal gene andthe 16S-23S intergenic spacer region

Mississippi State University. Isolates ofcolumnaris-like bacteria obtained from LSUand MSU were cultured, and DNA isolatedusing Purgene DNA isolation Kit (GentraSystems, Inc., Minneapolis, Minnesota). Aportion of the 16S and the entire 16S-23Sintergenic spacer of one isolate wasPCR-amplified using primers to regions of the16S and 23S ribosomal sequences that areconserved among the gram negative bacteria.One predominant product was obtained andcloned into pPCR4 TOPO cloning vector(Invitrogen) and sequenced. This was anintergenic sequence containing the tRNA foralanine and the tRNA for isoleucine. Severalproducts were expected, representing differentribosomal operons, but as of yet only this ITSproduct was found. Alignment of thesesequences with the tRNA sequences from

related organisms were used to identify con-served sequences, and primers were developedto allow direct PCR of the specific ITS anddirect sequencing of the products. These PCRproducts have been produced and both strandsof both products sequenced for all isolates.

The fragment of DNA between the 16S and23S ribosomal RNA encoding (ITS) of a totalof 50 Flavobacterium columnare case isolateswere amplified by polymerase chain reactionusing the common 16S-tRNA and tRNA-23Sprimer sets reported in year one of the project.The products were cloned and sequenced. Thesequences consist of a total of 748 base pairs(bp) and include a 100 bp portion of the 16Sfragment and 200 bp overlapping region. Insequence comparisons the 16S region wasuseful in identifying isolates that were notactually F. columnare. The ITS demonstratedsubstantial variation, however, at least 3distinct clusters of similar sequences wereidentified. These clusters demonstrated 5-10%sequence differences in the ITS region but lessthan 2% divergence within a cluster. Thissuggests that the isolates represent at least 3different strains. We are evaluating anadditional 20 isolates and comparing sourcesto see if sequence data correlates with host orseason. Also, the conserved sequences will beevaluated for differentiating diagnostic PCR.All sequencing data will be submitted toGenBank so that other diagnostic and researchlabs can use this information.

Ribotyping techniques to differentiateisotypes of columnaris-like bacteria

University of Tennessee. A number ofFlavobacterium columnare isolates from fish





disease cases were acquired, as well as severalFlavobacterium columnare-like bacteria,which share a close taxonomic relationship tothe target organism, including Flavobacteriumhydatis, F. succinicans, and F. psychrophilum.Those isolates are currently being typed usingribotyping methodologies, and assay com-ponents and procedural variations that providethe greatest fingerprint definition between thevarious isolates are being determined. Onceestablished, the optimal methodology will beused to generate a fingerprint database of theabove control isolates, which will then formthe basis for comparison of wild type isolatesobtained from other investigators involvedwith this project.

We have replicated and validated ribotypingmethods developed to differentiate variousisolates of Flavobacterium columnare (ATCCstrains, wild type strains isolated frominfected fish), F. hydatis, F. resiovorum,F. aquatile, F. flevense, and suspectedFlavobacterium spp. obtained from varioussources and other diverse species of bacteriathat might inhabit aquatic environments(Citrobacter freundii, Brochothrixthermosphacta, Aeromona veronii,Sphingomonas capsulate, Vibrio Cholerae,Pseudomonas stutzeri, Micrococcus luteus,Glaciecolia pallidula). While the Qualicondatabase is quite extensive across manyspecies of bacteria, the database forFlavobacterium spp. is rather limited, andthus as a part of this process, we are buildinga riboprint database, which will becomeavailable to other users of that system. In thatanalysis, we have noted good homology andyet acceptable separation of riboprintsbetween various Flavobacterium species andgood separation from and between most of the

diverse isolates, including Flavobacterium-like bacteria.

We have also progressed in studies todetermine the efficacy of pulsed-field gelelectrophoresis (PFGE) analysis for identi-fication and differentiation of a number of theabove species, and including various strains ofpathogenic Flavobacterium columnare. Thefirst phase of this work has involved thedevelopment of culture methods and PFGEprotocols, based on restriction digests, toobtain optimal fingerprints for identifying andseparating various bacterial isolates. BecauseF. colmnare colonies cannot be readilyseparated from solid growth media, obtainingan appropriate mass of cells for PFGE assaysvia picks or loops, as is done with otherbacterial species, was found to be difficult.Thus, isolates were grown in liquid culture,followed by centrifugation to obtain a cellpellet. Varying cell densities/concen-trationswere used for DNA isolation procedures,following by restriction enzyme digests priorto running digested DNA through PFGE gelsin conjunction with CHEF-mapper system(Bio-Rad Laboratories, Hercules, California).Test growth media have included ATCCNutrient Broth 3, 1839 Harpo’s HTYE, 1750Anacker and Ordal Medium. Separate trialswere conducted to test various restrictionenzymes, including SpeI, XbaI, BamHI, SmaI,ApaI. Additionally, varying densities ofSeaKem Gold Agarose (Fisher Scientific, FarLawn, New Jersey) were used for PFGEseparation of bands.

For the PFGE analysis, SpeI, XbaI, BamHIrestriction enzymes produced small andnumerous bands from DNA extracted fromFlavobacterium spp. and some Flavobacterium-





like bacteria, and those patterns were found to beunsuitable for PFGE type analysis; whereassuitable fingerprints were generated for othergram-negative species, including Salmonellacholeraesuis Braenderup control (ATCCBAA-664). We believe that restriction sites inFlavobacterium genome are too numerous forthe three restriction enzymes above. SmaI andApaI digests produced fewer but less distinctbands in Flavobacterium spp. and someFlavobacterium-like bacteria, whereas thoseenzymes again produced suitable fingerprintsfor other bacteria including control strains.Varying densities of PFGE gels did notprovide better resolution of banding patterns.

Currently, we are implementing additionalprotocols and alternate restriction enzymes todevelop more definitive PFGE typingpatterns. Such protocols will likely requiremore lengthy DNA extraction procedures,which will in turn require longer turn-aroundtimes for sample analyses.

To date, our studies indicate that ribotypingmay offer a more efficient and timesavingoption for identification and differentiation ofFlavobacterium, Flavobacterium-like, and

non-Flavobacterium isolates.

Determine the presence of unique outermembrane proteins of various strains ofcolumnaris-like bacteria

Clemson University. Several outermembrane proteins (OMP) from Flexibactercolumnare have been isolated and areconsistently found in all F. columnare. Overthe last two years we reported that a 30 kDaOMP isolated from F. columnare is a potentinducer of type II nitric oxide synthase(iNOS) and inducible prostaglandin H2synthase (cyclooxygenase-2; COX-2) inisolated catfish phagocytes, and that theseactivities can be blocked using specificantibodies against the OMP.

The proteomics facility within the ClemsonUniversity Genomics Institute (CUGI) helpedus to identify F. columnare OMPs, with aninitial focus on the 30 kDa, 40 kDa, and largerOMPs. Protein fragments between 12 and 15amino acid residues were isolated from eachband, yet were not identified because of a lackof homology with any known proteins incurrent data banks. These fragments, how-ever,may be useful for generating PCR primer sets toidentify larger coding regions of the OMPgenes. Very recently, we constructed a cDNAlibrary from a virulent Clemson Universitystrain of F. columnare using a commerciallyavailable kit (Ambion) to remove 18S and 20SRNA, thereby enriching mRNA that wassubsequently used to generate a long-PCR-based library (Clonetech). Our in-handantibodies against OMPs recognizedrecombinant proteins expressed in the library,and those positive plasmids are beingsequenced at the moment. Over the course of


Ú Molecular (PCR) and conventional (API

system) methods may be used for the

identification of F. columnare. Molecular

methods such as ribotyping, RAPD

analysis, and sequencing of the

intergenic spacer region between the

16S-23S ribosomal RNA genes can

allow for discrimination between

different genotypic strains.





the next 5 months recombinant OMP proteinswill be expressed and purified, then screenedfor both in vitro and in vivo activity in channelcatfish and tilapia phagocytes, and finally willbe used as potential vaccines against our

virulent strain of F. columnare The mostimportant product of this study will be theavailability of OMP-specific antibodies andrecombinant OMPs for general use by thecooperators in this project.

Objective 5. Develop challenge models for columnaris-like bacteriaisolated from major warmwater aquaculture species in thesoutheastern United States.

Internal genetic labeling of columnaris-likebacteria for use in the development of aneffective challenge model

Mississippi State University and AuburnUniversity. Our objectives for this project areto 1) ligate a Bacteroides consensus promotersequence upstream of gfp mut3a to allowexpression of green fluorescent protein inFlavobacterium columnare, 2) ligate the gfpgene and promoter into shuttle vector pCP11,3) transfer the pCP11-gfp plasmid intoFlavobacterium columnare for expression ofgreen fluorescent protein, and 4) use thefluorescent-labeled F. columnare to developan effective challenge model (in cooperationwith Joe Newton at Auburn University).

Two 65 base-pair DNA oligonucleotides con-taining a consensus promoter sequence fromBacteroides fragilis were synthesized andhybridized. The double stranded DNA con-taining the promoter was then digested withEcoRI and SacI and ligated upstream of gfpmut3a in plasmid pFPV25. Expression of thegfp gene from this plasmid (designatedpFCgfp) in E. coli was confirmed using afluorescence plate reader.

A SpeI-EcoRV fragment from pFCgfp was

ligated into pCP11, and the resulting plasmidwas designated pMWFCgfp. Expression ofgfp from this plasmid in E. coli was alsoconfirmed using a fluorescence plate reader.Another plasmid was also constructed byamplifying the ermF (erythromycin resis-tance) gene from pCP11 by PCR and cloningit into the EcoRI site of the broad host rangeplasmid pBBR1MCS4. The resulting plas-mid, pBBRermF, will allow selection inF. columnare based on erythromycinresistance. We then transferred the gfp genewith the Bacteroides promoter frompMWFCgfp into pBBRermF on a SmaI/SpeIfragment to construct pBBRFCgfp.

Objective 3 has not been successfully com-pleted despite numerous attempts to transferpMWFCgfp and pBBRFCgfp into multipleF. columnare isolates. We have been utilizinga conjugation technique using E. coli SM10lpir as a donor strain to attempt transfer of theplasmids into F. columnare. We have beenusing 25 columnaris strains that werecollected from John Hawke (LSU-SVM) aswell as an additional 20 isolates received fromJoe Newton. This year, we have pheno-typically characterized all the isolates toenable optimal growth conditions forF. columnare during the conjugation experi-





ments. We have also conducted experimentsthat resulted in adjustment of the erythromycinconcentration used on the selection plates. Inaddition, we have obtained two additionalplasmids, pCP23 and pCP29, from MarkMcBride at University of Wisconsin,Milwaukee that will enable us to utilize otherantibiotic selection markers. pCP23 carries atetracycline resistance gene that is expressedin flavobacteria, and pCP29 carries a cefoxitingene. Experiments are ongoing using ouroptimized growth/antibiotic conditions totransfer pMWFCgfp, pBBRFCgfp, pCP23,and pCP29 into the F. columnare isolatesfrom our panel.

A nalidixic acid resistant F. columnare mutant(spontaneous–not recombinant) has been iso-lated that can be used as a tagged organism (ifit is still virulent) until the gfp gene inF. columnare is successfully expressed.Several F. columnare isolates (including thenalidixic acid isolate) have been used inchallenge experiments following the pro-cedures of Andy Goodwin and S. Thomas-Jinuat the University of Arkansas at Pine Bluff. Todate it has not been possible to cause columnarisdisease in these experiments using theirprocedures.

Challenge models for channel catfish andgolden shiners

University of Arkansas at Pine Bluff. In thefirst part of our study we compared the bio-chemistry, DNA sequence (by RAPD) andpathogenicity of a large group of columnarisstrains. Channel catfish and golden shinerswere subjected to temperature shock and thenimmersed in a bath of columnaris bacteria at

a concentration sufficient to cause 60 to 70%mortality in 2 days using the more pathogenicof archived columnaris strains for therespective host. Each experiment wasperformed in triplicate with 20 fish per tank.Moribund fish were necropsied and the causeof death verified. Columnaris bacteria werere-isolated and identified by biochemical (tubetests) and molecular (randomly amplifiedpolymorphic DNA, RAPD, Promega)techniques to verify that the fish died frominfection by the challenge bacteria. In thiswork, we found that the catfish and cyprinidfish isolates fell in different clades, but thatthere was no correlation between these geneticand biochemical results, and any othermeasure including fish species of origin orpathogenicity to catfish and golden shiners.

Another way to look at differences betweencolumnaris isolates is to challenge fish andthen look for differences in response of theinfections to practical disease treatments.Columnaris disease was produced in channelcatfish, Ictalurus punctatus (Rafinesque) bybath exposure to 4 highly virulent isolates ofFlavo-bacterium columnare. In untreatedcontrols, mortality began 20 hours afterexposure and was 100% by 48 hours afterexposure. Mortality in channel catfish givenantibiotic treatments with oxytetracycline(OTC) or a combination of sulfadimethoxineand ormetoprim (SOR) in feed prior tobacterial challenge was 0% with all four

strains of F. columnare. Diquat was the mosteffective bath treatment; mortality with allfour strains was 0%. With potassiumpermanganate, chloramine-T, hydrogen per-oxide, and copper sulfate bath treatments,efficacy varied significantly among bacterial





isolates tested, pathogenicity in channelcatfish and shiners was similar. There was nocorrelation between biochemical character-istics or RAPD genogroup and pathogenicity.

Hybrid striped bass

Strains identified by the RAPD groupingsysytem of Farmer 2004 were used inchallenge studies. Representative isolatesfrom each RAPD group and host species wereobtained from our collection of archivedstrains at LSU. Strains used in the challengeswere: LADL 04-046 isolated from channelcatfish (RAPD Group I), LADL 04-066

isolated from large-mouth bass (RAPD GroupI), PB-02-12 isolated from fathead minnow(RAPD Group II), and LADL 94-141 isolatedfrom channel catfish (RAPD Group III).Strains LADL 04-066 largemouth bass(Group I) and PB-02-12 fathead minnow(Group II) were virulent in hybrid striped basscausing 100% mortality in 96 hours. StrainsLADL 94-141 channel catfish (Group III) andLADL 04-046 channel catfish (Group I) werenon-virulent in hybrid striped bass. Thus farvirulence appears to be correlated more withhost source than with RAPD group.Additional strains are currently being tested.

PUBLICATIONS, MANUSCRIPTS, OR PAPERS PRESENTED

Publications in print

Doffitt, C., L. M. Pote, and T. King. 2005. Bolbophorus damnificus infections of piscivorous birds in the

Mississippi Delta. Southeastern Biology 52:3.

Flowers J. R., M. F. Poore, L. M. Pote, R. W. Litaker and M. G. Levy. 2005. Cercariae of Bolbophorus

damnificus and Bolbophorus sp. with notes on North American bolbophorids. Comparative

Parasitology 72(2):220-226.

Labrie, L., C. Komar, J. Terhune, A. Camus, and D. Wise. 2004. Effect of sublethal exposure to the

trematode Bolbophorus spp. on the severity of enteric septicemia of catfish in channel catfish

fingerlings. Journal of Aquatic Animal Health 16(4):231-237.

Ledford, J.J. and A.M. Kelly. In press. Comparison of black carp, redear sunfish, and blue catfish as

biological controls of snail populations. North American Journal of Aquaculture.

Mischke, C.C., D.J. Wise, and L.M. Pote. 2005. Acute toxicity of chemicals to the ram’s horn snail

Planorbella trivolvis. Journal of the World Aquaculture Society 36:559-562.

Thomas-Jinu, S. and A. E. Goodwin. 2004. Morphologic and genetic characteristics of Flavobacterium

Columnare isolates: Correlations with virulence in fish. Journal of Fish Diseases 27:29-35.

Thomas-Jinu, S. and A. E. Goodwin. 2004. Acute columnaris infection in channel catfish: Efficacy of

treatments practical in warmwater aquaculture ponds. Journal of Fish Diseases 27:23-28.

Theses

Zhang, Y. 2004. Adhesiveness of Flavobacterium columnare: comparison of different methods and

isolates. M.S. thesis. Auburn University, Auburn, Alabama.





strains and among treatments. Bath treat-ments with chloramine-T and potassiumpermanganate reduced mortality from 100% to75% and 69%, respectively, but copper sulfateand hydrogen peroxide treatments were noteffective.

Based on our results, oral antibiotics pre-vented columnaris disease but, of the bathtreatments, only Diquat produced a dramaticreduction in the mortality of acutely infectedfish. Diquat is labeled for aquatic use as anherbicide in the United States but in largeponds it is prohibitively expensive.

Challenge models for hybrid striped bass

Louisiana State University. Strains ofFlavobacterium columnare, archived in theLSU Aquatic Diagnostic Laboratory reposi-tory, are being used in virulence studies inhybrid striped bass. Methods that produceuniform mortality rates of 75% or greater

following exposure are classified as virulentstrains of F. columnare. These criteria will beadopted for use to compare virulence ofarchived strains from various locations andvarious species outlined in Objective 6.Hybrid striped bass (20 g mean weight) wereacclimated and held in the AquaticPathobiology Building at the LSU School ofVeterinary Medicine. The strains (isolates)were evaluated for virulence in a standardizedchallenge procedure where scales were re-moved and the skin scarified in a 1 cm area.2

Cultures of the F. columnare bacteria wereswabbed on the scarified area rather thanusing the immersion method. This was donedue to difficulties with clumping of thebacteria and difficulty enumerating bacteria inthe challenge bath. Virulent strains ofF. columnare colonized the scarified skinreadily and caused infection and diseasewhereas avirulent isolates did not causeinfection. Mortality was evident 96 hours afterinfection with virulent strains.

Objective 6. Use challenge models for each fish species to correlatevirulence with biotype and or genotype of columnaris-like bacteria.

Channel catfish and golden shiners

University of Arkansas at Pine Bluff. Vari-ability in Flavobacterium columnare patho-genicity makes disease treatment difficultbecause there is currently no way to easilyrecognize those strains that warrant aggressivetreatments. In order to identify suitable markers,17 isolates of F. columnare were culturedfrom six different fish species. The DNA fromall isolates was analyzed using randomlyamplified polymorphic DNA (RAPD).

Bootstrap analysis of the RAPD data produceda tree with three major groups supported byscores of 80% to 100% similarity.

The remaining objective was to completechallenge assays to see if there were any staindifferences in the virulence of columnarisisolates using the golden shiner as a host. Ofthe strains tested, two of catfish origin thatwere highly virulent in channel catfish (100%mortality) produced much lower mortality(30% to 40%) in golden shiners. For the other

Identification, Characterization, and Evaluation of Mechanisms of Control of Bolbophorus-like Trematodes and Flavobacterium



Farmer, B. 2004. Improved methods for the isolation and characterization of Flavobacterium columnare.

M.S. thesis. Department of Pathobiological Sciences, School of Veterinary Medicine, Louisiana

State University, Baton Rouge, Louisiana.

Manuscripts in preparation

Mitchell, A. J. and S. Snyder. Optimizing slurried-hydrated lime pond-shoreline treatments for the control

of rams-horn snails

Farmer, B. and J. P. Hawke. Improved methods for culture and antibiotic susceptibility testing of

F. columnare.

Kelly, A. M. Prey preference of redear sunfish and blue catfish trained on various conditioning diets.

Aquaculture.

Presentations

Farmer, B. and J. Hawke. 2004. Improved methods for the diagnosis and characterization of

Flavobacterium columnare. Annual Meeting of the Fish Health Section, American Fisheries

Society, 25-29 July, Shepherdstown, West Virginia.

Hanson, L. A., L. M. Ford, B. Scheffler and J. P. Hawke. 2004. Sequence analysis of the 16S-23S

intergenic spacer of Flavobacterium columnare to identify potential strains. Annual Meeting of

the South Central Branch of the American Society for Microbiology, 5-6 November, Mississippi

State, Mississippi.

Zhang, Y. and J.M. Grizzle. 2004. Evaluation of an assay for adhesiveness of Flavobacterium

columnare. Annual Meeting of the Fish Health Section, American Fisheries Society, 25-29 July,

Shepherdstown, West Virginia.

Goodwin, A. E. 2004. Efficacy of treatments for columnaris disease. Arkansas Aquaculture 2004, 16-17

January, Hot Springs, Arkansas.

Goodwin, A. E. 2004. Efficacy of treatments for columnaris disease. Ninth Biennial Fish Disease

Diagnosticians Meeting. 8-10 February, Biloxi, Mississippi.

Goodwin, A. E. 2004. Farm Bureau Fish Disease Research and Extension. Arkansas Farm Bureau

Aquaculture Meeting, June, Little Rock, Arkansas.

Doffitt, C., D. T. King, and L. M. Pote. 2005. Infections of piscivorous birds in the Mississippi delta by

Bolbophorus damnificus. Association of Southeastern Biologists, 13-16 April, Florence,

Alabama.

Doffitt, C., D. T. King, and L. M. Pote. 2005. Trematode infections of piscivorous birds in the

Mississippi delta The Wildlife Society-Mississippi Chapter, 13-14 October, Choctaw,

Mississippi.

Pote, L. M., M. C. Yost, C. Doffitt, B. S. Dorr, D. T. King, A. Camus, and D. Wise. 2005. Further

elucidation of the life cycle and pathology of the digenetic trematode, Bolbophorus damnificus.

Thirtieth Eastern Fish Health Workshop. Kearneysville, West Virginia.

Wise D., C. Mischke, T. Byars, and A. Camus. 2005. Evaluation of copper sulfate to control snail

numbers in catfish ponds affected by Bolbophorus trematodes. Thirtieth Eastern Fish Health

Workshop. Kearneysville, West Virginia.

� � �

Documents

IDENTIFICATION, CHARACTERIZATION, AND EVALUATION …srac.msstate.edu/Project PDFs/Trematodes.pdf · Identification, Characterization, and Evaluation of Mechanisms of Control of Bolbophorus-like