Control of Flavobacterium Psychrophilum: Tests of Erythromycin, Streptomycin, Osmotic and Thermal Shocks, and Rapid Ph Change Randall W

This article was downloaded by: [Department Of Fisheries] On: 20 January 2013, At: 18:28 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Aquatic Animal Health Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/uahh20 Control of Flavobacterium psychrophilum: Tests of Erythromycin, Streptomycin, Osmotic and Thermal Shocks, and Rapid pH Change Randall W. Oplinger a & Eric J. Wagner a a Utah Division of Wildlife Resources, 1465 West 200 North, Logan, Utah, 84321, USA Version of record first published: 14 Dec 2012.

To cite this article: Randall W. Oplinger & Eric J. Wagner (2013): Control of Flavobacterium psychrophilum: Tests of Erythromycin, Streptomycin, Osmotic and Thermal Shocks, and Rapid pH Change, Journal of Aquatic Animal Health, 25:1, 1-8 To link to this article: http://dx.doi.org/10.1080/08997659.2012.720636

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material. Journal of Aquatic Animal Health 25:1–8, 2013 C American Fisheries Society 2013 ISSN: 0899-7659 print / 1548-8667 online DOI: 10.1080/08997659.2012.720636

ARTICLE

Control of Flavobacterium psychrophilum:Tests of Erythromycin, Streptomycin, Osmotic and Thermal Shocks, and Rapid pH Change

Randall W. Oplinger* and Eric J. Wagner Utah Division of Wildlife Resources, 1465 West 200 North, Logan, Utah 84321, USA

Abstract Flavobacterium psychrophilum, the etiological agent that causes bacterial coldwater disease, has been implicated in signiﬁcant reductions in the numbers of salmonids reared at ﬁsh hatcheries. In this study, we performed a series of in vitro experiments to test the effectiveness of rapid temperature, pH, and osmotic pressure changes in killing three strains of the bacterium. We also evaluated the effectiveness of antibiotics (erythromycin, streptomycin, and a penicillin–streptomycin mixture) against F. psychrophilum. The bacterium tolerated temperatures of 40–50◦Cfor up to 30 min (when acclimated to 15◦C). The bacterium can survive lower temperatures for >60 min. Although temperatures ≥55◦C appeared to kill F. psychrophilum on contact, we found that eyed eggs of Rainbow Trout Oncorhynchus mykiss were not able to survive short (<60-s) exposures at these temperatures. We found that rapid changes in pH (15-min exposures to pH 3.0, 4.0, 5.0, 7.0, 9.0, 10.0, and 11.0) and osmotic pressure (15-min exposures to 0, 6, 8, 10, 12% sodium chloride) were not effective at killing the bacterium. Erythromycin concentrations up to 2,000 mg/L for 15 min were also ineffective. However, streptomycin concentrations ≥5,000 mg/L killed the bacterium during a 15-min exposure. The combination of penicillin and streptomycin was also effective, killing the bacterium at doses as low as 2.5 × 106 IU penicillin + 2,500 mg/L streptomycin. Our trials demonstrate that elevated temperatures and the combination of penicillin and streptomycin can kill F. psychrophilum under in vitro conditions. Erythromycin and rapid changes in pH and osmotic pressure are not effective at killing the bacterium.

Disease outbreaks in fish hatcheries can significantly in- been placed on research on egg disinfection methods that can crease the average cost of producing fish as well as reduc- prevent the vertical transmission of the disease (e.g., Kumagai ing the numbers of fish that are reared and ultimately stocked et al. 2004; Wagner et al. 2008, 2010). Vaccination also has

Downloaded by [Department Of Fisheries] at 18:28 20 January 2013 into the wild. Bacterial coldwater disease, caused by the long, the potential to prevent the vertical transmission of the disease. thin gram-negative bacterium Flavobacterium psychrophilum, Vaccines, however, are still under development and may not be has been implicated in significant reductions in the numbers available for use for a number of years (LaFrentz et al. 2004; of fish stocked in western North America. For example, the Plant et al. 2009). Antibiotics can be effective at preventing State of Utah estimates that 25–30% of the Rainbow Trout On- the horizontal transmission of F. psychrophilum. Some strains corhynchus mykiss raised in the state’s hatchery system each are susceptible to florfenicol but resistant to oxytetracycline and year are lost to F. psychrophilum (C. Wilson, Utah Division of sulfadimethoxine (Bruun et al. 2000). Unfortunately, these three Wildlife Resources, personal communication). Bacterial cold- drugs are the only antibiotics approved for aquaculture use by water disease has significantly influenced hatcheries worldwide, the U.S. Food and Drug Administration (USFDA 2011). and improved methods for controlling the disease are needed. Despite our knowledge of methods for the control of the dis- Unfortunately, few methods of controlling F. psychrophilum ease, F. psychrophilum continues to be a major problem in fish are available to hatchery managers. Considerable emphasis has hatcheries. This suggests that the methods that are currently used

*Corresponding author: [email protected] Received February 9, 2012; accepted August 6, 2012 1 2 OPLINGER AND WAGNER

to control the disease are not effective and that further research of several antibiotics to the bacterium. Each of the described on methods for controlling the spread of F. psychrophilum is experiments was performed twice. The purpose of the first trial required. A number of studies of F.psychrophilum have reported was to establish the basic tolerance limits of F.psychrophilum to on the efficacy of several antibiotics, including oxytetracy- the treatments. During this first trial, we tested the treatments on cline, amoxicillin, ormetoprim-sulfadimethoxine, ampicillin, a single strain of the bacterium (ATCC 49510). The second set doxycycline, enrofloxacin, florfenicol, flumequine, gentamicin, of trials were performed at a later time and were designed to de- oxolinic acid, sarafloxacin, and sulfamethoxazole-trimethoprim termine whether the tolerances determined during the first trial (Rangdale et al. 1997; Bruun et al. 2000; Dalsgaard and Madsen are similar across strains. During this second trial, three strains 2000; Schmidt et al. 2000; Kum et al. 2008; Hesami et al. 2010). (ATCC 49510, Clear Springs Foods 259-93 and Utah Division Two other candidate antibiotics include erythromycin and strep- of Wildlife Resources strain 09-131) of F. psychrophilum were tomycin. Erythromycin is primarily used against gram-positive tested. The latter two strains were isolated from moribund Rain- species, but it has been used with success against some gram- bow Trout. An established PCR procedure (Wiklund et al. 2000) negative species (Lorian and Sabath 1970). Streptomycin is a was used to confirm that each strain was F. psychrophilum. Both broad-spectrum antibiotic that is effective against both gram- trials used identical methodology. On occasion, the treatments positive and gram-negative species (Schatz et al. 1944). Strepto- tested varied between trials. The treatments selected for the sec- mycin and erythromycin are not approved for aquaculture use in ond trial were based on the results from the first trial. Any differ- the United States. Although these drugs may be beneficial for the ences between the two trials are discussed in more detail below. industry, approval may take several years. Therefore, alternative Experiment 1: temperature.—Electric water baths (Precision control methods that rely on FDA-approved compounds or cre- Model 66802) were used to create temperatures of 25, 30, 35, ative treatment methods should be pursued further. For example, 40, 45, 50, 55, 65, 75, and 85◦C. As a control, a 15◦C bath sudden changes in temperature could be effective at killing F. was also created by placing a plastic tub filled with water on psychrophilum that is attached to the surface of objects such a laboratory benchtop. Ice or hot water was added to maintain as eggs or hatchery equipment (e.g., waders, nets). Flavobac- the temperature at 15◦C. A rack containing test tubes (16 × terium psychrophilum has been reported to grow at temperatures 125 mm) filled with 15 mL of sterilized, hatchery well water ranging from 4◦Cto23◦C (Nematollahi et al. 2003). The critical (pH = 7.6; hardness and alkalinity = 220 mg/L) were added thermal maximum for F. psychrophilum has not been reported. to each bath. Prior to the experiment, the tubes were allowed Sudden changes in pH or sodium chloride concentration to sit in the water baths for at least 15 min to reach the desired could present other methods for killing surface-attached treatment temperature. The culture containing F. psychrophilum F. psychrophilum. Madetoja et al. (2003) successfully cultured was held at 15◦C until the start of the experiment. F. psychrophilum for over 50 d at a salinity of 6%. However, Ninety-six hours prior to the start of the experiment, the bacteria used by Madetoja et al. (2003) were not exposed F. psychrophilum was added to an Erlenmeyer flask contain- to a rapid osmotic shock but instead were acclimated to the test ing 100 mL of a maltose-infused tryptone yeast extract broth conditions. In contrast, Suomalainen et al. (2005) found in vitro (MAT; 0.4% tryptone, 0.04% yeast extract, 0.05% CaCl2, 0.05% that 15-min exposures to a pH of 4.6 and 60-min exposures MgSO4, 1% maltose, and 0.02% C2H3NaO2; Crump et al. 2001). to sodium chloride concentrations of 4% or pH of 4.86–5.0 The bacteria were slowly stirred using a stir-plate and were cul- significantly reduced (by ∼99%) the numbers of the bacterium tured at 15◦C. To start the experiment, 200 ¯L of the inoculated F. columnare. It is possible that rapid changes in pH or MAT broth was added to the test tubes within the water baths. sodium chloride concentration are also effective at controlling The solution was mixed by drawing two 4-mL aliquots of the F. psychrophilum. solution from each test tube into a sterilized pipette. The so- Downloaded by [Department Of Fisheries] at 18:28 20 January 2013 Bacterial coldwater disease is a major source of mortality lution was immediately returned to the test tube. Immediately for hatchery-reared salmonids. New methods for the control of afterwards, 100 ¯L of solution was withdrawn and added to F. psychrophilum need to be investigated. In this study, we con- a petri dish (100 × 15 mm) containing tryptone yeast extract ducted a series of in vitro tests to assess the viability of several salt agar media (TYES; 0.4% tryptone, 0.04% yeast extract, new control methods for the disease. We determined whether 0.05% CaCl2, 0.05% MgSO4; Holt et al. 1993). The solution rapid changes in temperature, pH, or osmotic pressure have was distributed using a sterilized spreader. The petri dishes were the potential to kill F. psychrophilum. We also determined the wrapped with laboratory film and incubated at 15◦C. concentrations of erythromycin, streptomycin, and a penicillin– Every 10 min for 60 min after the addition of the bacteria, streptomycin mixture required to kill the bacterium during a the solution in each test tube was mixed and plated as described 15-min exposure. above. Bacteria were only added to each tube once. Therefore, repeated samples were taken from each test tube for the first 60 min after the addition of F. psychrophilum. For every five METHODS test tubes that were inoculated with F. psychrophilum,asixth We performed a series of in vitro tests to determine the tol- tube was established as a procedural control. This tube was not erance of F. psychrophilum to rapid changes in temperature, inoculated with bacteria but was otherwise handled in the same pH, and osmotic pressure. We also determined the toxicity manner as the other test tubes. CONTROL OF FLAVOBACTERIUM PSYCHROPHILUM 3

During the initial trial (in which only the ATCC strain of the After a 2-s rinse, the strip was rinsed in the other rinse tube bacterium was included), we tested temperatures of 15, 25, 35, for two additional seconds. After rinsing, the strip was streaked 40, 45, 55, 65, 75, and 85◦C. Samples were withdrawn from the twice (once per side of the strip) across the central axis of a test tubes 0, 10, 20, 30, 40, 50, and 60 min after the addition 100-mm-diameter petri dish containing TYES media. The petri of the bacteria (the 0-min sample was actually collected 5–10 s dish was then wrapped in laboratory film and incubated at 15◦C. after the addition of the bacteria). Five replicate test tubes of During the preliminary trial (in which only the ATCC strain was F. psychrophilum were exposed to each temperature. During a included), five pH treatments (4.0, 5.0, 7.0, 9.0, and 10.0) were follow-up trial (ATCC, Clear Springs Foods, and Utah strains tested. During the second trial (which included all three strains were compared), we tested temperatures of 15, 35, 40, 45, 50, of F. psychrophilum), we tested pH levels of 3.0, 4.0, 7.0, 10.0, and 55◦C and samples were removed from the tubes 0, 5, 10, 30, and 11.0. and 60 min after the addition of F. psychrophilum. Five replicate Experiment 3: sodium chloride.—The methods used to test tubes of each strain of F. psychrophilum were exposed to test the effect of rapid changes in osmotic pressure on each temperature. F. psychrophilum were identical to the procedures used to test In addition to testing the ability of thermal shocks to kill the effect of pH change. In the preliminary trial, we evaluated F. psychrophilum, we tested the ability of eyed Rainbow Trout 15-min exposures to sodium chloride concentrations of 0, 2, eggs to tolerate a temperature of 55◦C. In this trial, two water 4, 6, and 8% (pH = 7.2). The second trial (which determined baths were created using well water (pH = 7.4; total hardness = repeatability across strains) tested 15-min exposures to sodium 180 mg/L). One water bath contained water that was heated to chloride concentrations of either 0% or 12%. For both trials, 55◦C and the other water at 13◦C (control), the same temperature we exposed five test strips (replicates) from each strain to each the eggs experienced during incubation. A mesh bag was added sodium chloride concentration. to each water bath, and eggs (N = 10, drawn haphazardly from Experiment 4: streptomycin, erythromycin, and a bucket containing eggs from 10 females) were added. The streptomycin–penicillin mixture.—The in vitro methods eggs were exposed to the temperature for 10, 30, or 60 s. The were also used to test the toxicity of streptomycin (Sigma mesh bag was then removed and placed into a beaker containing Aldrich S9137), erythromycin (Sigma Aldrich E5389), and a ambient-temperature well water (∼13◦C) for 10 s and was then commercial penicillin–streptomycin mixture (Sigma Aldrich transferred to a vertical stack incubator tray. After 48 h in the P4333; 10,000 IU penicillin and 10 mg streptomycin/mL) incubation tray, the eggs were removed from the bag and the against F. psychrophilum. For the preliminary trial with number of live eggs or fry was counted. Three replicate groups of streptomycin, we tested concentrations of 0, 100, 500, 1,000, 10 eggs were exposed to each combination of exposure duration 5,000, and 10,000 mg/L. For the follow-up trial (with all (10, 30, or 60 s) and temperature (13◦Cor55◦C). three strains of F. psychrophilum), we tested 0, 100, 500, Experiment 2: pH.—In vitro methods were used to de- 1,000, and 2,500 mg/L. For erythromycin we tested 0, 10, termine whether rapid changes in pH can be used to kill 100, 500, 1,000, and 2,000 mg/L during the preliminary trial F.psychrophilum. For this experiment, several series of test tubes and 0 and 2,000 mg/L during the secondary trial. For the were placed into a rack, each series containing four test tubes. commercially prepared penicillin–streptomycin mixture, we The first tube (12 × 75 mm) was filled with 2.0 mL of sterilized tested concentrations of 0, 1, 5, 10, 25, 50, 75, and 100% of well water. Shortly (<30 min) prior to the start of the experiment, the prepared product (diluted with sterilized hatchery well 1.0 mL of MAT broth that was inoculated with F.psychrophilum water) during the preliminary trial and 0, 1, 5, 10, 25, and 50% was added to this test tube. The second test tube (16 × 125 mm) during the secondary trial. Depending on the concentration contained 15 mL of sterilized well water. The pH of the well and antibiotic, either five or ten replicate test strips were Downloaded by [Department Of Fisheries] at 18:28 20 January 2013 water in these test tubes was adjusted (using hydrochloric acid exposed. All dilutions were made into sterilized well water. or sodium hydroxide). The pH was measured with a freshly F. psychrophilum was exposed to the antibiotics for 15 min. calibrated Orion Model SA 720 pH meter. To ensure that the Data analysis.—For all experiments, the petri dishes were pH did not change during the experiment, readings were taken removed from the incubator once every 48 h over an 8-d obser- both before and after the experiment. The third and fourth test vation period. Bacterial growth was noted as presence/absence. tubes contained 15 mL of sterilized well water and served as The morphology of all the colonies cultured was consistent rinse tubes. Five test tube series (replicates) were prepared for with F. psychrophilum. Gram stains were collected from 5 to each pH tested. The bacteria used in this experiment were ac- 10 randomly selected petri dishes with bacterial growth from climated to a pH of 7.2. Each series of test tubes were held each experiment to confirm that the cultured bacteria was con- at 15◦C. sistent with F. psychrophilum. The gram stain results were con- To conduct the experiment, a sterilized plastic strip (50 × sistent with F. psychrophilum in all instances. The percentage 5 mm) was dipped into the first test tube containing of tubes from each replicate that had F. psychrophilum growth F. psychrophilum. After 10 s, the strip was transferred by ster- was determined. In addition, a probit analysis was performed ile forceps to the tube containing the pH-adjusted well water. using the PROC PROBIT procedure in SAS (SAS 1998) to esti- After 15 min, the strip was transferred to the first rinse tube. mate the 50%- and 99%-lethal concentrations (LC50 and LC99 4 OPLINGER AND WAGNER

values) for each antibiotic treatment. The LC50 and LC99 values RESULTS represent the antibiotic concentrations for which growth would only be expected in 1% (LC99) or 50% (LC50) of the replicates Experiment 1: Temperature during a 15-min exposure. For the thermal shock test, the pro- During the ﬁrst portion of the experiment, in which the basic bit analysis was used to estimate the temperature required for temperature tolerance of F.psychrophilum was determined using F. psychrophilum growth on either 1% (LC99) or 50% (LC50) just the ATCC strain of the bacterium, no bacteria were cultured ◦ of the replicates 0, 10, 30, or 60 min after the introduction of from any replicates treated at temperatures ≥55 C regardless the bacterium. of treatment duration (Table 1). In contrast, F. psychrophilum

TABLE 1. Percentages of plates of TYES media with culturable Flavobacterium psychrophilum (N = 4 or 8 plates, depending on the temperature and iteration). The bacteria were exposed to temperatures ranging from 15◦Cto85◦C for up to 60 min. Data for the ATCC 49510 strain were collected in two iterations. In the ﬁrst iteration (ATCC 1), the basic temperature tolerance of F. psychrophilum was determined; in the second iteration (ATCC 2), data were also collected for the Clear Springs Foods 259-93 (CSF 259-93) and Utah Division of Wildlife Resources 09-131 (09-131) strains.

Time ATCC ATCC CSF Time ATCC ATCC CSF Temp (◦C) (min) 1 2 259-93 09-131 Temp (◦C) (min) 1 2 259-93 09-131 15 0 100 100 100 100 45 0 100 100 100 100 10 100 45 5 25 0 0 0 20 100 45 10 50 0 25 25 30 100 45 20 100 40 100 45 30 25 0 0 0 50 100 45 40 0 60 100 100 100 100 45 50 0 25 0 100 45 60 0 0 0 0 10 100 50 0 75 100 25 25 20 100 50 5 0 0 0 0 30 100 50 10 0 0 0 0 40 100 50 30 0 0 0 0 50 100 50 60 0 60 100 55 0 0 0 0 0 30 0 100 55 5 0 0 0 10 100 55 10 0 0 0 0 20 100 55 20 0 30 100 55 30 0 0 0 0 40 100 55 40 0 50 100 55 50 0 60 100 55 60 0 35 0 100 100 100 100 65 10 0 5 100 100 100 100 65 20 0

Downloaded by [Department Of Fisheries] at 18:28 20 January 2013 10 100 100 100 100 65 30 0 20 100 65 40 0 30 100 0 25 25 65 50 0 40 100 65 60 0 50 100 75 10 0 60 100 50 0 0 75 20 0 40 0 100 100 100 100 75 30 0 5 0 100 100 25 75 40 0 10 63 75 75 25 75 50 0 20 0 75 60 0 30 6 0 50 0 85 10 0 40 0 85 20 0 50 0 85 30 0 60 0 0 75 0 85 40 0 85 50 0 85 60 0 CONTROL OF FLAVOBACTERIUM PSYCHROPHILUM 5

TABLE 2. Estimated temperatures (◦C) required for 50% (LC50) and 99% TABLE 3. Percentages of TYES media plates with culturable Flavobacterium (LC99) reductions in the number of Flavobacterium psychrophilum after 0, psychrophilum (N = 5 plates for all treatments except the 0% and 25% concen- 10, 30, or 60 min of exposure at increased temperature. Estimates were derived trations of the penicillin–streptomycin mixture [N = 10]). The bacteria were using a probit analysis. The bacteria were acclimated to 15◦C prior to the thermal exposed to various concentrations of streptomycin, erythromycin, or a commer- shock. cial penicillin–streptomycin mixture for 15 min. The concentrations listed for the penicillin–streptomycin mixture refer to the percentage concentrations in Time (min) LC50 LC99 our test solution, which were obtained by dilution with sterilized well water. See Table 1 for additional information. 05152 10 43 54 Concentration ATCC ATCC CSF 30 38 49 Antibiotic (mg/L) 1 2 259-93 09-131 60 37 38 Streptomycin 0 100 100 100 100 100 100 100 100 60 500 60 90 80 50 ≤ ◦ was cultured at all temperatures 35 C. At temperatures of 1,000 0 70 40 50 40–50◦C, the bacterium did not survive the entire 60-min expo- ◦ ◦ 2,500 40 20 20 sure duration. At both 40 C and 45 C, we were able to culture 5,000 0 F. psychrophilum from plastic strips that were treated for 10,000 0 ≤30 min. At 50◦C, we recovered bacteria from strips that were Erythromycin 0 100 100 100 100 treated for ≤5min. 10 100 In the second trial, in which three strains of F. psychrophilum 100 100 were tested, the results were similar across strains. Regardless 500 100 of strain, bacteria were able to survive exposure to temperatures 1,000 100 ≤30◦C for at least 60 min (Table 1). All strains were killed on 2,000 100 100 100 100 contact at temperatures ≥55◦C. The CSF 259-93 strain survived ◦ + for at least 60 min at temperatures ≤40 C. We were only able to Penicillin 0% 100 100 100 100 recover the 09-131 strain after a 10-min exposure at 40◦C. This streptomycin 1% 100 100 90 80 strain survived for at least 30 min at temperatures ≤35◦C. Both 5% 80 90 20 0 the CSF 259-93 and the 09-131 strains survived for 10 min at 10% 80 60 0 0 45◦C. 25% 0 0 0 0 The probit analysis (Table 2) showed that the temperature re- 50% 0 0 0 0 quired to result in no bacterial growth decreased with exposure 75% 0 time. For example, in 10-min exposures, a temperature of 54◦C 100% 0 was estimated to remove bacteria from 99% of the replicates but that only 38◦C was required if the exposure period lasted 60 min (Table 2). Not surprisingly, higher temperatures were required to reduce growth to 1% of the isolates than to 50% of of the bacterium was killed after a 15-min exposure to concen- the isolates. In the egg treatments, no eggs that were exposed to ≥ ◦ trations 5,000 mg/L (Table 3). The probit analysis showed that a temperature of 55 C survived. Therefore, this temperature ex- a 15-min exposure to a streptomycin concentration of 505 mg/L ceeds the short-term (10–60-s) tolerance limit of eyed Rainbow would kill 50% of F. psychrophilum and that a concentration of Downloaded by [Department Of Fisheries] at 18:28 20 January 2013 Trout eggs. 567 mg/L would be required to kill 99% of the bacteria. Dur- Experiments 2 and 3: pH and Sodium Chloride ing the second trial, we found that all three strains were able Rapid changes in pH (3.0–11.0) were not effective in killing to survive treatment at concentrations of at least 2,500 mg/L. F. psychrophilum. Regardless of strain, we recovered bacteria Bacteria were cultured from a lower percentage (20% versus from every replicate. Similarly, rapid changes in sodium chlo- 40%) of the replicates from the CSF 259-93 and 09-131 strains ride concentration were not effective in killing the bacterium. than the ATCC strain, suggesting that these strains are more Since we were able to recover the bacterium from all of the strips susceptible to streptomycin treatment than the ATCC strain. Be- included in these tests, we were not able to calculate LD50 and cause we did not test any concentrations that killed 100% of the LD99 values for the effect of rapid changes in pH and osmotic F. psychrophilum, we were not able to perform a probit anal- pressure on the bacterium. ysis on the effects of streptomycin on the CSF 259-93 and 09-131 strains of the bacterium. We recovered F. psychrophilum Experiment 4: Streptomycin, Erythromycin, from every strip treated with erythromycin (regardless of strain and Penicillin–Streptomycin Mixture and concentration), indicating that concentrations >2,000 mg/L We found that streptomycin is effective at controlling F. psy- are required to kill F. psychrophilum with a 15-min exposure chrophilum. During the ﬁrst trial, we found that the ATCC strain (Table 3). 6 OPLINGER AND WAGNER

TABLE 4. Estimated concentrations of a penicillin–streptomycin mixture that elevated temperature cannot be used for the disinfection of required for 50% (LC50) and 99% (LC99) reductions in the number of Rainbow Trout eggs. Previous research has shown that stressed Flavobacterium psychrophilum after 15 min of exposure. Estimates were derived using a probit analysis. The listed concentrations are the percentages of a bacteria can enter a “viable but nonculturable” state (Barer et al. commercially prepared solution that contained 10,000 IU penicillin + 10 mg 1993). As a result, the conditions under which we were not able streptomycin/mL. to culture F. psychrophilum may not have killed the bacterium. Instead, the thermal shock may simply have caused it to enter a Strain LC50 LC99 state in which it was not culturable. ATCC 49510 11.2 47.7 We observed some differences among strains at some of the temperatures tested. In particular, the CSF 259-93 strain was the CSF 259-93 2.5 15.3 ◦ 09-131 1.1 1.4 only strain that was cultured after 60 min of exposure at 40 C. This may indicate that this strain tolerates higher temperatures than do the other strains. Alternatively, it is possible that temperatures in the 40–50◦C range represent a “gray area” in which The commercial penicillin–streptomycin mixture was effec- F. psychrophilum can either grow well or be killed outright. tive in killing the ATCC 49510 strain of F. psychrophilum dur- Bacteria from the other strains may also have survived 60 min ing 15-min exposures to concentrations of ≥25% of that of the of treatment at 40◦C, but we were not able to culture them. Re- prepared product (i.e., 2.5 × 106 IU penicillin + 2,500 mg gardless, we consistently were not able to culture bacteria from streptomycin / L were required; Table 3). The CSF 259-93 and samples that were treated at temperatures >55◦. We advise that 09-131 strains of the bacterium were more susceptible to treat- this temperature be used as a minimum for disinfecting surfaces ment with the penicillin–streptomycin mixture and were killed that may be contaminated with F. psychrophilum. at concentrations ≥10% of the commercial product (i.e., 1.0 × Antibiotic treatment could provide another avenue for killing 106 IU penicillin + 1,000 mg streptomycin / L were required). F. psychrophilum. Previous research has established that the in- The probit analysis showed that lower concentrations of the jection of steelhead (anadromous Rainbow Trout) brood with penicillin–streptomycin mixture would be required to kill the erythromycin helped reduce the prevalence of the disease among CSF 259-93 and 09-131 strains of the bacterium than to kill the the progeny (Brown et al. 1997). However, in our study, a 15- ATCC strain (Table 4). min exposure to concentrations up to 2,000 mg/L was insufficient to kill F.psychrophilum. To our knowledge, injections with DISCUSSION streptomycin have not been attempted. For streptomycin, we de- Our results demonstrate that sudden increases in temperature termined effective concentrations against F. psychrophilum and might be able to kill F. psychrophilum, as we were not able to recommend concentrations ≥5,000 mg/L to kill the bacterium. culture bacteria that were treated at temperatures ≥55◦C. The Currently, erythromycin and streptomycin are not approved for bacterium is able to survive for at least 60 min at tempera- aquaculture use by the FDA. Penicillin has been used for treating tures between 15◦C and 35◦C. Other research has shown that other species of bacteria in fish culture (DeCew 1972; Brown F. psychrophilum can reproduce when incubated at tempera- et al. 1990), but we are not aware of any literature on controlling tures between 4◦C and 23◦C, with the optimum growth rate F. psychrophilum infections with this antibiotic. Lumsden et al. being achieved at 15◦C (Holt 1987; Bernardet and Kerouault (1996) noted that F. psychrophilum was susceptible to peni- 1989). Clearly, F. psychrophilum can tolerate short exposures cillin in a disk test, an observation that has been corroborated to temperatures greater than optimal for incubation. The genus by our research (Oplinger, personal observation). Bustos et al. Flavobacterium contains species that can tolerate conditions (1995), however, noted that a Chilean isolate of F.psychrophilum Downloaded by [Department Of Fisheries] at 18:28 20 January 2013 that range from freezing (e.g., F. hibernum; McCammon et al. was not susceptible to penicillin, although the concentration of 1998) to ones with temperatures exceeding 70◦C (e.g., F. ther- the sensitivity disk was not reported. In this study, the combi- mophilum; Oshima and Yamakawa 1974). Because of its relation nation of penicillin and streptomycin was effective in killing to these species, it is likely that F. psychrophilum has similar at- F. psychrophilum at lower concentrations (2.5 × 106 IU peni- tributes and thus can tolerate suboptimal thermal conditions for cillin + 2,500 mg/L streptomycin) than streptomycin alone. short periods of time. In addition, most bacteria species possess Previous research has found that penicillin and streptomycin can heat shock proteins that help protect their cellular membranes work synergistically against bacteria (Watanakunakorn 1971). and essential nucleic acids and proteins during periods of ther- It is possible that this synergism also operates with respect to mal stress (Oshima and Yamakawa 1974; Lindquist 1986). Our the control of F. psychrophilum. F. psychrophilum culture was incubated at 15◦C. It is possible We observed variation among the three F. psychrophilum that bacteria acclimated to different temperatures would respond strains to the antibiotics tested. In general, the ATCC 49510 to the conditions tested in this study differently. Regardless, strain was the least susceptible to the antibiotics. Both the CSF our results suggest that thermal shock can be used to remove 259-93 and Utah 09-131 strains had similar susceptibility to surface-attached F.psychrophilum. These findings could be use- streptomycin. The Utah strain was the most susceptible to the ful in the disinfection of equipment. Our findings, however, show penicillin–streptomycin mixture. Our findings corroborate the CONTROL OF FLAVOBACTERIUM PSYCHROPHILUM 7

work of other researchers, who have also found differences in Bayer, M. E. 1967. Response of cell walls of Escherichia coli to a sudden antibiotic susceptibility among F. psychrophilum strains (Bruun reduction of the environmental osmotic pressure. Journal of Bacteriology et al. 2000). It is not known why these differences in suscep- 93:1104–1112. Bernardet, J. F., and B. Kerouault. 1989. Phenotypic and genomic studies tibility exist. It is likely, however, that the strains have differ- of Cytophaga psychrophila isolated from diseased Rainbow Trout (On- ent antibiotic exposure histories and that the more susceptible corhynchus mykiss) in France. Applied and Environmental Microbiology 55: strains have historically received less antibiotic exposure than 1796–1800. the less susceptible strains. Brown, L. L., L. J. Albright, and T. P. T. Evelyn. 1990. Control of vertical Osmotic shock and changes in pH have been used less transmission of Renibacterium salmoninarum by injection of antibiotics into maturing female Coho Salmon Oncorhynchus kisutch. Diseases of Aquatic frequently to kill bacteria. Working with Flavobacterium Organisms 9:127–131. columnare, Suomalainen et al. (2005) demonstrated that Brown, L. L., W. T. Cox, and R. P. Levine. 1997. Evidence that the causal agent 15-min exposures to a pH of 4.6 and 60-min exposures to sodium of bacterial coldwater disease Flavobacterium psychrophilum is transmitted chloride concentrations of 4% or pH of 4.86–5.00 significantly within salmonid eggs. Diseases of Aquatic Organisms 29:213–218. reduced bacteria numbers. In the present study, we found that Bruun, M. S., L. Madsen, and I. Dalsgaard. 2003. Efficiency of oxytetracycline treatment in Rainbow Trout experimentally infected with Flavobacterium these two processes were ineffective in killing F.psychrophilum. psychrophilum strains having different in vitro antibiotic susceptibilities. Unlike those in other studies, our bacteria were not conditioned Aquaculture 215:11–20. to the sodium chloride and pH conditions tested (e.g., Madetoja Bruun, M. S., A. S. Schmidt, L. Madsen, and I. Dalsgaard. 2000. Antimicro- et al. 2003), so our data represent a true “shock.” Bayer (1967) bial resistance patterns in Danish isolates of Flavobacterium psychrophilum. showed that Escherichia coli is more susceptible to sudden de- Aquaculture 187:201–212. Bustos, P. A., J. Calbuyahue, J. Montana,˜ B. Poazo, P. Entrala, and R. Soler- creases in osmotic pressure when growing logarithmically, so vicens. 1995. First isolation of Flexibacter psychrophilus, as causative agent culture age is a potential difference among the studies. Future re- of Rainbow Trout fry syndrome (RTFS), producing Rainbow Trout mortal- search should test other F. psychrophilum strains and life stages ity in Chile. Bulletin of the European Association of Fish Pathologists 15: to confirm the repeatability of the results observed in this study. 162–164. Using in vitro tests, we were able to establish guidelines for Crump, E. M., M. B. Perry, S. C. Clouthier, and W. W. Kay. 2001. Antigenic characterization of the fish pathogen Flavobacterium psychrophilum. Applied the elimination of F. psychrophilum through the use of thermal and Environmental Microbiology 67:750–759. shocks and antibiotics. In the current study, streptomycin (es- Dalsgaard, I., and L. Madsen. 2000. Bacterial pathogens in Rainbow Trout, pecially when used synergistically with penicillin) was shown Oncorhynchus mykiss (Walbuam), reared at Danish freshwater farms. Journal to be effective against F. psychrophilum. The treatment condi- of Fish Diseases 23:199–209. tions and concentrations required to kill bacteria in vitro are DeCew, M. G.1972. Antibiotic toxicity, efficacy, and teratogenicity in adult spring Chinook Salmon (Oncorhynchus tshawytscha). Journal of the Fisheries not necessarily the same as those required in vivo (Gee and Research Board of Canada 29:1513–1517. Sarles 1942;Bruun et al. 2003). In practice, poor water circu- Gee, L. L., and W. B. Sarles. 1942. The disinfection of trout eggs contaminated lation (if thermal shocks are utilized) or biofilms (Stewart and with Bacterium salmonicida. Journal of Bacteriology 44:111–126. Costerton 2001) reduce the effectiveness of treatments. Regard- Hesami, S., J. Parkman, J. I. MacInnes, J. T. Gray, C. L. Gyles, and J. S. Lums- less, in vitro studies are a necessary first step in testing new den. 2010. Antimicrobial susceptibility of Flavobacterium psychrophilum isolates from Ontario. Journal of Aquatic Animal Health 22:39–49. treatment methods against F. psychrophilum. Future research Holt, R. A. 1987. Cytophaga psychrophila, the causative agent of bacterial cold- should test thermal shocks, streptomycin, and the penicillin– water disease in salmonid fish. Doctoral dissertation. Oregon State University, streptomycin mixture in vivo to establish the effectiveness of Corvallis. these treatments in a fish culture setting. In addition, future re- Holt, R. A., J. S. Rohovec, and J. L. Fryer. 1993. Bacterial cold-water disease. search should determine the effects of these treatments on egg Pages 3–22 in V. Inglis, R., R. Roberts and N. R. Bromage, editors. Bacterial diseases of fish. Blackwell Scientific Publications, Oxford, UK. and larval development.

Downloaded by [Department Of Fisheries] at 18:28 20 January 2013 Kum, C., S. Kirkan, S. Sekkin, F. Akar, and M. Boyacioglu. 2008. Comparison of in vitro antimicrobial susceptibility in Flavobacterium psychrophilum isolated from Rainbow Trout fry. Journal of Aquatic Animal Health 20:245–251. Kumagai, A., C. Nakayasu, and N. Oseko. 2004. No evidence for the pres- ACKNOWLEDGMENTS ence of Flavobacterium psychrophilum within Ayu eggs. Fish Pathology 39: We would like to thank M. Bartley for his assistance in the 183–187. laboratory. Funding for this research was provided by the Fed- LaFrentz, B. R., S. E. LaPatra, G. R. Jones, and K. D. Cain. 2004. Protec- eral Aid in Sport Fish Restoration Program (project F-96-R) and tive immunity in Rainbow Trout Oncorhynchus mykiss following immuniza- tion with distinct molecular mass fractions isolated from Flavobacterium the Utah Division of Wildlife Resources. This manuscript was psychrophilum. Diseases of Aquatic Organisms 59:17–26. greatly improved by feedback from three anonymous reviewers. Lindquist, S. 1986. The heat-shock response. Annual Review of Biochemistry 55:1151–1191. Lorian, V., and L. D. Sabath. 1970. Effect of pH on the activity of erythromycin against 500 isolates of gram-negative bacilli. Applied and Environmental REFERENCES Microbiology 20:754–756. Barer, M. R., L. T. Gribbon, C. R. Harwood, and C. E. Nwoguh. 1993. The Lumsden, J. S., V. E. Ostland, and H. W. Ferguson. 1996. Necrotic myositis in viable but non-culturable hypothesis and medical bacteriology. Reviews in cage cultured Rainbow Trout, Oncorhynchus mykiss (Walbaum), caused by Medical Microbiology 4:183–191. Flexibacter psychrophilus. Journal of Fish Diseases 19:113–119. 8 OPLINGER AND WAGNER

Madetoja, J., P. Nyman, and T. Wiklund. 2003. Survival and virulence of Schmidt, A. S., M. S. Bruun, I. Dalsgaard, K. Pedersen, and J. L. Larsen. 2000. Flavobacterium psychrophilum in water microcosms. FEMS (Federation of Occurrence of antimicrobial resistance in fish-pathogenic and environmental European Microbiologial Societies) Microbiology Ecology 43:217–223. bacteria associated with four Danish Rainbow Trout farms. Applied and McCammon, S. A., B. H. Innes, J. P. Bowman, P. D. Franzmann, S. J. Dobson, Environmental Microbiology 66:4908–4915. P. E. Holloway, J. H. Skerratt, P. D. Nichols, and L. M. Rankin. 1998. Stewart, P. S., and J. W. Costerton. 2001. Antibiotic resistance of bacteria in Flavobacterium hibernum sp. nov., a lactose-utilizing bacterium from a biofilms. Lancet 358:135–138. freshwater Antarctic lake. International Journal of Systematic Bacteriology Suomalainen, L. R., M. Tiirola, and E. T. Valtonen. 2005. Treatment of colum- 48:1405–1412. naris disease of Rainbow Trout: low pH and salt as possible tools? Diseases Nematollahi, A., A. Decostere, F. Pasmans, and F. Haesebrouck. 2003. of Aquatic Organisms 65:115–120. Flavobacterium psychrophilum infections in salmonid fish. Journal of Fish USFDA (U.S. Food and Drug Administration). 2011. Approved drugs. Disease 26:563–574. USFDA, Silver Spring, Maryland. Available: http://www.fda.gov/Animal Oshima, M., and T. Yamakawa. 1974. Chemical structure of a novel glycolipid Veterinary/DevelopmentApprovalProcess/Aquaculture/ucm132954.htm. from an extreme thermophile, Flavobacterium thermophilum. Biochemistry (February 2012). 13:1140–1146. Wagner, E. J., R. E. Arndt, E. J. Billman, A. Forest, and W. Cavender. 2008. Plant, K. P., S. E. LaPatra, and K. D. Cain. 2009. Vaccination of Rainbow Comparison of the efficacy of iodine, formalin, salt, and hydrogen peroxide Trout, Oncorhynchus mykiss (Walbaum), with recombinant and DNA vac- for control of external bacteria on Rainbow Trout eggs. North American cines produced to Flavobacterium psychrophilum heat shock proteins 60 and Journal of Aquaculture 70:118–127. 70. Journal of Fish Diseases 32:521–534. Wagner, E. J., R. W. Oplinger, R. E. Arndt, A. M. Forest, and M. Bartley. Rangdale, R. E., R. H. J. Richards, and D. J. Alderman. 1997. Mini- 2010. The safety and effectiveness of various hydrogen peroxide and iodine mum inhibitory concentrations of selected antimicrobial compounds against treatment regimens for Rainbow Trout egg disinfection. North American Flavobacterium psychrophilum the causal agent of Rainbow Trout fry syn- Journal of Aquaculture 72:34–42. drome (RTFS). Aquaculture 158:193–201. Watanakunakorn, C. 1971. Penicillin combined with gentamicin or strepto- SAS Institute. 1998. SAS/STAT user’s guide, version 8.0, volume 1. SAS Insti- mycin: synergism against entercocci. Journal of Infectious Diseases 124: tute, Cary, North Carolina. 581–586. Schatz, A., E. Bugle, and S. A. Waksman. 1944. Streptomycin, a substance Wiklund, T., L. Madsen, M. S. Bruun, and I. Dalsgaard. 2000. Detection of exhibiting antibiotic activity against gram positive and gram negative bacteria. Flavobacterium psychrophilum from fish tissue and water samples by PCR Experimental Biology and Medicine 55:66–69. amplification. Journal of Applied Microbiology 88:299–307. Downloaded by [Department Of Fisheries] at 18:28 20 January 2013 This article was downloaded by: [Department Of Fisheries] On: 20 January 2013, At: 18:30 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Aquatic Animal Health Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/uahh20 Isolation of Yersinia ruckeri Strain H01 from Farm- Raised Amur Sturgeon Acipenser schrencki in China Li Shaowu a , Wang Di a , Liu Hongbai a & Lu Tongyan a a Department of Aquaculture, Heilongjiang River Fisheries Research Institute, Chinese Academy of Fishery Sciences, 43 Songfa Road, Daoli District, Harbin, 150070, China Version of record first published: 14 Dec 2012.

To cite this article: Li Shaowu , Wang Di , Liu Hongbai & Lu Tongyan (2013): Isolation of Yersinia ruckeri Strain H01 from Farm-Raised Amur Sturgeon Acipenser schrencki in China, Journal of Aquatic Animal Health, 25:1, 9-14 To link to this article: http://dx.doi.org/10.1080/08997659.2012.728169

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material. Journal of Aquatic Animal Health 25:9–14, 2013 C American Fisheries Society 2013 ISSN: 0899-7659 print / 1548-8667 online DOI: 10.1080/08997659.2012.728169

ARTICLE

Isolation of Yersinia ruckeri Strain H01 from Farm-Raised Amur Sturgeon Acipenser schrencki in China

Li Shaowu, Wang Di, Liu Hongbai, and Lu Tongyan* Department of Aquaculture, Heilongjiang River Fisheries Research Institute, Chinese Academy of Fishery Sciences, 43 Songfa Road, Daoli District, Harbin 150070, China

Abstract Yersinia ruckeri is the causative agent of enteric redmouth disease or yersiniosis, which affects salmonids and several other species of fish. However, there are no reports on the characteristics and pathogenicity of Y. ruckeri isolated from farm-raised Amur Sturgeon Acipenser schrencki. Here, we isolated and characterized Y. ruckeri strain H01 from the diseased Amur Sturgeon in China. The phenotypic and genotypic characteristics of Y. ruckeri were observed, and its virulence was tested by examining experimentally infected sturgeons. Examination of the flagellar morphology of Y. ruckeri by transmission electron microscopy showed five to eight peritrichous flagella located on the cell body. Actively dividing cells with an obvious cell membrane were approximately 0.64 µm in diameter and between 1.7 and 2.5 µm in length. The dose that was lethal to 50% of the test fish after intraperitoneal injection was determined to be 7.2 × 106 CFU, and Y. ruckeri could be reisolated from the liver and kidneys of infected sturgeon. Antimicrobial susceptibility tests showed that H01 was susceptible to 10 antimicrobial agents. Part of the 16S rRNA sequences (563 base pair) was amplified and sequenced to study the genotypic characterization in Y. ruckeri (GenBank accession number JQ657818). The phylogenetic tree revealed H01 was clustered together with Y. ruckeri strains. Together, this study describes the isolation, characterization, and phenotypic–genotypic analysis of a Y. ruckeri strain isolated from farm-raised Amur Sturgeon. The results discovered may provide some theoretical basis for the prevention and control of yersiniosis in Amur Sturgeon.

Yersinia ruckeri, a member of the family Enterobacteriaceae, (Vuillaume et al. 1987; Berc et al. 1999; Danley et al. 1999; is the causative agent of enteric redmouth disease or yersinio- Eissa et al. 2008). Infection with Y. ruckeri results in a bacterial sis, a general septicemia affecting mainly salmonids (Fernadez´ septicemia without disease-specific signs but is most commonly Downloaded by [Department Of Fisheries] at 18:30 20 January 2013 et al. 2007; Tobback et al. 2007). Since its first description in detected due to exophthalmos and hemorrhages in the eyes of the Hagerman Valley of Idaho in the 1950s, Y. ruckeri has been fish. The enteric redmouth disease exhibited obvious subcuta- found to be widely distributed in many salmonid species in North neous hemorrhage in the mouth and throat of Rainbow Trout America (Ross et al. 1966). More recently, the disease has been Oncorhynchus mykiss (Furones et al. 1993). To our knowledge, reported and the bacterium has been isolated from fish hosts few studies on Y. ruckeri in China have been performed. Xu et al. throughout the world, including in Europe, North and South (1991) isolated a Y. ruckeri strain from diseased Silver Carp Hy- America, Australia, New Zealand, and South Africa (Austin pophthalmichthys molitrix and Bighead Carp H. nobilis in 1991. and Austin 2007). Although salmonids are the main fish species Fan et al. (2010) reported the isolation and identification of susceptible to Y. ruckeri, susceptibility has also been reported Y. ruckeri FF003 from Channel Catfish Ictalurus punctatus.The in other fish species such as European Eels Anguilla anguilla, FF003 strain had apparent pathogenicity to Channel Catfish, Goldfish Carassius auratus, catfish Ictalurus spp., carp Cypri- with symptoms of punctate hemorrhaging on the body surface, nus spp., sturgeon Acipenser spp., and tilapia Oreochromis spp. especially on the submaxilla, abdominal wall, and the side of

*Corresponding author: [email protected] Received July 23, 2012; accepted September 5, 2012 9 10 LI ET AL.

the body (Fan et al. 2010). Accurate diagnosis of yersiniosis cillin (100 ¯g/disk), chloramphenicol (30 ¯g/disk), ciprofloxacin requires isolation and identification of Y. ruckeri from tissue (5 ¯g/disk), enrofloxacin (5 ¯g/disk), erythromycin (15 ¯g/disk), samples. More commonly it is based on its phenotypic profile furazolidone (300 ¯g/disk), gentamicin (10 ¯g/disk), kanamycin and can be readily differentiated from other taxa within the En- (30 ¯g/disk), norfloxacin (10 ¯g/disk), novobiocin (5 ¯g/disk), terobacteriaceae genera. In addition, more specific techniques ofloxacin (5 ¯g/disk), penicillineG (10 ¯g/disk), tetracycline such as the polymerase chain reaction (PCR) are available but (30 ¯g/disk), and sulfamethoxazole (23.75 ¯g/disk) was deter- are best used to confirm the identity of ambiguous isolates. mined on the Mueller Hinton agar by the disk diffusion method As the chonrdr-ganoid scale fish with some characteristics (NCCLS 1993). between Chondrichthyes and Osteichthyes, sturgeons have been Morphological characterization by negative-stain transmis- raised broadly in many countries for their considerable value sion electron microscopy.—The bacteria were streaked onto a in economy and ecology. There are eight indigenous species of TSA and incubated for 24 h at 25◦C. The single colony was Acipenseriformes in China, and the production of farm-cultured looped, suspended, and washed with phosphate-buffered saline. sturgeons in China appears to have become the largest in the One drop of the bacterial suspension was placed on parafilm for world since 2000 (Wei et al. 1997; Wei et al. 2004). However, 5 min, then onto a formvar-coated, 300-mesh grid and allowed with high-density culture, water pollution, and improper man- to dry. The grid was stained with 2% uranyl acetate (pH 6.5) agement, the development of this industry was severely limited for 30 sec, dried at room temperature, and then examined by by several bacterial and viral diseases, including septicemia, transmission electron microscope (Hitachi-7650, Japan). enteritis, rotted gills, Saprolegniasis and White Sturgeon iri- Virulence of Y. ruckeri H01 on Amur Sturgeon.—Healthy ju- dovirus. Antibiotics such as florfenicol and enrofloxacin are venile sturgeons (body weight of 40–60 g) obtained from the used to prevent and control these diseases nowadays. With re- fish farm (Hebei Province, China) were used for virulence as- spect to the clinical signs of the disease in other fish species, says. The health of the fish stock was checked upon arrival in Vuillaume et al. (1987) provided the first description of Y. ruck- the laboratory by collecting samples from the internal organs eri isolated from Siberian Sturgeon Acipenser baeri cultured in for microbiological analysis, and fish were held in tanks for 2 southwestern France. Hemorrhages were observed around the weeks prior to virulence tests. Virulence assays were done by mouth, at the base of the rostrum and pectoral fins, and around intraperitoneal injection with a dosage of 3 × 107 CFU, which the urogenital pore. The belly was swollen, and a bloody liquid was determined based on McFarland turbidimetric standards. exuded from the vent under finger pressure. However, there was Fish (n = 10) were maintained at 20◦Cin0.5-m3 tanks con- no report on the Amur Sturgeon until now. In this study, a Y. ruck- taining aerated, static fresh water and equipped with a filtering eri strain, H01, was isolated and identified from farm-cultured system. All of the fish were examined microbiologically at 2 d Amur Sturgeon according to morphological, physiological, and post inoculation. Experimental control groups were challenged biochemical characteristics and phylogenetic analysis. with sterile saline solution in a separate aquarium. Determination of lethal dose.— Yersinia ruckeri H01 was grown on TSA overnight at 25◦C, harvested, washed, and sus- METHODS pended in a 0.9% sodium chloride solution (pH 7.2). Serial Bacterial isolation.— Yersinia ruckeri strain H01 was iso- 1.fivefold dilution of a known number of bacteria/mL was made lated from outbreaks that occurred in 2010 in a fish farm located with normal saline, and 0.2 mL of each dilution was inoculated in Hebei Province, China. Farms consist of concrete tanks, and intraperitoneally into each group of fish (n = 10). Fish were ob- the freshwater is supplied from rivers, with water temperatures served for 8 d, and the number of surviving fish was recorded. ranging from 15◦Cto17◦C and an average pH of 7.0. After The dose that was lethal to 50% of the test fish (LD50) was de- Downloaded by [Department Of Fisheries] at 18:30 20 January 2013 the observation of clinical signs, we examined the gills and termined using the formula derived by Reed and Muench (Reed internal organs for parasitic and bacterial infections. For bacte- and Muench, 1938). rial isolation, samples were aseptically collected from the liver Sequencing and analysis of 16S rRNA.—Total bacterial and kidney and directly streaked onto trypticase soy-agar (TSA; DNA was extracted from pure bacterial cultures using a Bac- Difco) and incubated at 25◦C for 24–48 h. Pure cultures were teria Genomic DNA Extraction Kit (Tiangen, China). The kept frozen at −80◦C in tryptic soy broth (Difco) supplemented concentration of DNA was quantified spectrophotometrically with 15% glycerol. and adjusted to a concentration of 100 ng/¯L. The 16S rRNA Biochemical characterization.—Bacterial isolate was sub- gene from DNA templates was amplified using the primer jected to morphological, physiological, and biochemical tests Yer-F: 5 -CGAGGAGGAAGGGTTAAGT-3 and Yer-R: 5- using classical tube and plate procedures. In parallel, the iso- AAGGCACCAAGGCATCTCT -3 (Cunningham et al., 2010). late was identified using Biolog, the microbial identification The amplification mixture (50 ¯L) comprised 2 ¯L (10 pmol/¯L) system (Biolog, Hayward, California). The effect of tempera- each of the Yer-F and Yer-R primers, 0.5 ¯L (2 U/¯L) of Taq ture, salinity, and pH value on the growth of the bacteria was DNA polymerase, 5 ¯L of × 10 PCR reaction buffer, 1 ¯L of determined. In addition, the antimicrobial susceptibility of H01 dNTP mixture (10 mM each) and 0.5 ¯L of DNA template. The strain to amikacin (30 ¯g/disk), amoxicillin (20 ¯g/disk), carbeni- cycling program was as follows: initial denaturation at 95◦Cfor ISOLATION OF YERSINIA RUCKERI FROM AMUR STURGEON 11

TABLE 1. Characteristics and biochemical reactions for H01 isolate in this study. Note: “ + ” means positive reaction, “–” means negative reaction, “d” means variable reactions, and “n.d.” means the test was not done.

Yersinia ruckeri H01 Yersinia ruckeri Yersinia ruckeri FF003 Characteristics or Test (in this study) (Danley et al. 1999) (Fan et al. 2010) Gram stain – – – Oxidation–fermentation ++ + Oxidase – – – Motility + –n.d. Urease – n.d. – Lysine Decarboxylase + n.d. + Ornithine Decarboxylase ++ + Glutamyl transferase ++n.d. Simmon’s citrate + – + Gelatin + n.d. + Indole production – – – VP reaction – – + Gas production – d – D- xylose – – – D( + )-Cellobiose – n.d. – Melibiose – n.d. – Rafﬁnose – n.d. – Sucrose – n.d. – L-Rhamnose – – – L-arabinose – – – Trehalose – ++ D(–)-Salicin – n.d. – Sorbitolum ++ + Inositol – n.d. –

5 min, followed by 30 cycles of denaturation at 94◦C for 30 s, of the pectoral fins, the abdomen, and the urogenital pore as annealing at 55◦C for 30 s, extension at 72◦C for 1 min, and main external signals. Furthermore, they exhibited petechiae in a final extension at 72◦C for 10 min. Controls, without DNA, some internal organs, including the liver, hindgut, and swim were simultaneously included in the amplification process. The bladder. A dominant strain named H01 was isolated from the integrity of the PCR products was assayed by the development TSA plate after bacterial culture for 24 h at 25◦C. The colonies of single bands following electrophoresis for 30 min at 120 V in were 1–2 mm in diameter, smooth, round, raised, and with entire 1% (w/v) agarose gels in TAE buffer. The PCR products were se- edges. The optimum growth condition of Y. ruckeri H01 was pH quenced by Shanghai Shenggong Biotechnological (Shanghai, = 7.0, salinity = 0%, and temperature = 25◦C for 18 h, and Downloaded by [Department Of Fisheries] at 18:30 20 January 2013 China). The sequence obtained was compared with known se- their growth could be inhibited at the TSA medium when the quences by searching the National Center for Biotechnology In- pH value varied from 0 to 4 and 10 to 14 or the concentration formation databases with the BLAST program. A phylogenetic of NaCl was higher than 4%. The H01 strain was identified tree was constructed by comparing the 16S rRNA sequence of as Y. ruckeri with Biolog similarity of 0.561. The biochemical the strain H01 and other relative bacteria species in the Gen- characteristics of H01 strain are indicated in Table 1. Bank using MEGA 4.0 software based on the neighbor-joining Transmission electron microscopy was used to examine the method (Saitou and Nei, 1987). ultrastructure of flagella. As shown in Figure 1, there was an obvious cell membrane over the cell and actively growing cells were approximately 0.64 ¯m in diameter and between 1.7 and RESULTS 2.5 ¯m in length. The thickness of the membrane was about 0.12 ¯m. Each bacterial cell body contained between five and Bacterial Isolation and Biochemical Characteristics eight flagella in a peritrichous arrangement and was anchored Diseased Amur Sturgeon with variable size (from 15 to 30 g) firmly at the cell body. In addition, the cell membrane appeared showed hemorrhages around the mouth, the lower jaw, the base to be shrunken and rough in shape. 12 LI ET AL.

were collected from diseased fish, which exhibited obvious hemorrhaging. In the microbiological examination, Y. ruckeri was reisolated from those tissues, thus confirming infection. In addition, the number of deaths after intraperitoneal injection was observed in 72 h and found to have a LD50 value of 7.2 × 106 CFU (Table 3). The 563 base pair region of the 16S rRNA gene from the H01 isolate revealed a 98%–99% sequence identity to Y. ruckeri in a BLAST search. The sequence has been deposited with accession number JQ657818 in a genome database (i.e., GenBank) of the National Center for Biotechnology Information and compared with that of different Yersinia spp. and several other usual pathogens in aquaculture. A phylogenetic tree was constructed with a neighbor-joining method as shown in Figure 2. The phylogenetic tree analysis showed that the H01 strain was clustered with Y. ruckeri isolates (the black arrow) and was more distantly related to the genus Aeromonas and the genus Edwardsiella. FIGURE 1. Morphological characterization of Yersinia ruckeri H01 as shown by negative- staining transmission electron micrograph (25,000×). DISCUSSION Through testing antimicrobial susceptibility to 16 different After the first identification of Y. ruckeri as a pathogen of types of antibiotics, it was found that the H01 strain was suscep- Rainbow Trout, Y. ruckeri has appeared in various fish species tible to 10 antibiotics, including amikacin, amoxicillin, carbeni- throughout the world (Buller 2004). Outbreaks of yersiniosis cilli, ciprofloxacin, enrofloxacin, gentamicin, kanamycin, nor- are often associated with poor water quality, excessive stocking floxacin, ofloxacin, and tetracycline, while resistant to chloram- densities, and environmental stressors (Tobback et al. 2007). phenicol, erythromycin, furazolidone, novobiocin, penicillin G, Yersinia ruckeri can also persist in an asymptomatic carrier state, and sulfamethoxazole (Table 2). where infection through carrier fish is especially important under Artificial infection of Y. ruckeri was performed to determine stress conditions. However, before this report yersiniosis was the virulence of H01 strain. The results showed that 80% of ex- only found in Siberian Sturgeon but no other sturgeon species. perimental fish died in 2 days by intraperitoneal injection with In this study, we described the isolation of a Y. ruckeri strain a dosage of 3 × 107 CFU. The liver, kidney, and spleen tissues from cultured Amur Sturgeon and the characteristics of the Y. ruckeri isolate. The diseased fish showed obvious hemorrhages TABLE 2. Antibiotic susceptibility of Yersinia ruckeri strain H01. Sensitivity around the mouth, the lower jaw, the base of the pectoral fins, the is designated as S = sensitive and R = resistant. abdomen, and the urogenital pore as the main external signs, as well as petechiae in the liver, hindgut, and swim bladder in inter- Inhibition zone nal organs. Moreover, these clinical signs were very similar to Antibiotics diameter (N = 3; mm) Sensitivity diseased Siberian Sturgeon reported by Vuillaume et al. in 1987. Amikacin 18 ± 2.64 S The isolated strain H01 was mobile and flagellated, had a wrin- ± kled cell membrane (Figure 1), and exhibited a high pathogenic- Downloaded by [Department Of Fisheries] at 18:30 20 January 2013 Amoxicillin 23.3 2.30 S × 6 Carbenicillin 23 ± 1.73 S ity to this species, with an LD50 value of 7.2 10 CFU. This Chloramphenicol 0 R suggests that Y. ruckeri H01 has morphological characteristics Ciprofloxacin 28.7 ± 3.06 S similar to those in strains isolated from salmon species, and it Enrofloxacin 29.3 ± 3.06 S may be pathogenic or virulent to sturgeons (Davies and Frerichs Erythromycin 0 R 1989). Recently, the analysis of the 16S rRNA gene sequence has Furazolidone 10.7 ± 1.53 R been widely applied to the identification of pathogens isolated Gentamicin 16.3 ± 0.58 S from diseased aquatic animals (LeJeune and Rurangirwa 2000; Kanamycin 20 ± 2.64 S Eddy et al. 2007; Pourahmad et al. 2008). Here the 16S rRNA Norfloxacin 27.7 ± 3.79 S sequence of the H01 strain was used to construct a phylogenetic Novobiocin 0 R tree, and the results showed that H01 was close to Y. ruckeri, Ofloxacin 28 ± 4.00 S with a homology of 99%. A bacterial challenge study indicated Penicillin G 9.33 ± 0.58 R that H01 possessed a high virulence to Amur Sturgeon and Tetracycline 19 ± 1.73 S could induce easily observable clinical signs. In the antimicro- Sulfamethoxazole 0 R bial susceptibility tests, H01 was susceptible to 10 antimicrobial agents, indicating that the isolates might be unexposed or less ISOLATION OF YERSINIA RUCKERI FROM AMUR STURGEON 13

TABLE 3. The cumulative death number of ﬁsh injected with different concentrations of Yersinia ruckeri H01.

Bacterial dosage ( × 107 CFU) Time 3 2 1.33 0.89 0.59 0.39 0.26 0.17 0 The cumulative number of deaths 24 h 8/10a 4/10 3/10 2/10 1/10 0/10 0/10 0/10 0/10 48 h 9/10 7/10 5/10 4/10 3/10 1/10 0/10 0/10 0/10 72 h 10/10 9/10 7/10 6/10 4/10 3/10 1/10 0/10 0/10

aThe number of ﬁsh per group.

exposed to many antimicrobial agents. These findings proved (Fernadez´ et al. 2004; Coquet et al. 2005). However, informa- that Y. ruckeri was one of the pathogens of sturgeon bacterial tion about the pathogenicity or exact virulence of Y. ruckeri to the septicemia and could induce the obvious hemorrhages around Amur Sturgeon has not been available until now. Therefore, it the fish body. Antimicrobial susceptibility tests suggested 10 is of future interest to investigate the pathogenicity or virulence possible drugs that could be used to control this disease, which along with corresponding immune responses in Amur Stur- would be helpful to decrease the loss in the sturgeon aquaculture geon by using microarrays, silico experiments, 2D maps, and industry. so on. Although generally well controlled by means of vaccination In brief, we described a Y. ruckeri strain isolated from and antibiotic treatment, there have been continuous yersin- diseased Amur Sturgeon and investigated its basic character- iosis outbreaks, especially in endemic areas. In some cases, istics, which may be helpful in providing a theoretical ba- the losses due to this disease can be as high as 30–70% of sis for prevention and control of yersiniosis. Furthermore, it the stock (Horne et al. 1999). Over the past years, the use of would be important to isolate more bacteria strains from af- molecular techniques has opened new possibilities to probe the fected fish to understand the prevalence of Y. ruckeri in Amur mechanisms of the pathogenicity of Y. ruckeri in salmonids Sturgeon. Downloaded by [Department Of Fisheries] at 18:30 20 January 2013

FIGURE 2. Phylogenetic relationship among the Yersinia species based on 16S rRNA gene sequences. 14 LI ET AL.

ACKNOWLEDGMENTS Fernadez,´ L., I. Mequez,´ and J. A. Guijarro. 2007. Molecular virulence mecha- This work was supported by grants from the Central- nisms of the fish pathogen Yersinia ruckeri. Veterinary Microbiology 125:1– Level Nonprofit Scientific Research Institutes Special Funds 10. Furones, M. D., C. J. Rodgers, and C. B. Munn. 1993. Yersinia ruckeri, the (201003), National “Twelfth Five-Year” Plan for Science and causal agent of enteric redmouth disease (ERM) in fish. Annual Review of Technology (2012BAD25B10), and Natural Science Founda- Fish Diseases 3:105–125. tion of Heilongjiang Province of China (C201140). Horne, M. T., and A. C. Barnes. 1999. Enteric redmouth disease (Yersinia ruckeri). Pages 455–477 in P. T. K. Woo and D. W. Bruno, editors. Fish diseases and disorders, volume 3: viral, bacterial and fungal infections. CABI REFERENCES Publishing, Cambridge, Massachusetts. Austin, B., and D. A. Austin. 2007. Bacterial fish pathogens: diseases of farmed LeJeune, J. T., and F. R. Rurangirwa. 2000. Polymerase chain reaction for and wild fish, 4th edition. Springer-Praxis, Chichester, UK. definitive identification of Yersinia ruckeri. Journal of Veterinary Diagnostic Berc, A., Z. Petrinec, Z. Matasin, and Z. Kozaric.´ 1999. Yersinia ruckeri septi- Investigation 12:558–561. caemia in experimentally infected carp (Cyprinus carpio L.) fingerlings. Acta NCCLS (National Committee for Clinical Laboratory Standards). 1993. Veterinaria Hungarica 47:161–172. Performance standards for antimicrobial disc susceptibility tests. Clin- Buller, N. B. 2004. Bacteria from fish and other aquatic animals: a practical ical and Laboratory Standards Institute, Publication M2-A5, Wayne, identification manual. CABI Publishing, Cambridge, Massachusetts. Pennsylvania. Coquet, L., P. Cosette, E. De,´ L. Galas, H. Vaudry, C. Rihouey, P. Lerouge, G. A. Pourahmad, F., F. Cervellione, K. D. Thompson, J. B. Taggart, A. Adams, Junter, and T. Jouenne. 2005. Immobilization induces alterations in the outer and R. H. Richards. 2008. Mycobacterium stomatepiae sp. nov., a slowly membrane protein pattern of Yersinia ruckeri. Journal of Proteome Research growing, non-chromogenic species isolated from fish. International Journal 4:1988–1998. of Systematic and Evolutionary Microbiology 58:2821–2827. Cunningham, S. A., L. M. Sloan, L. M. Nyre, E. A. Vetter, J. Mandrekar, and R. Reed, L. J., and H. Muench. 1938. A simple method of estimating fifty per cent Patel. 2010. Three-hour molecular detection of Campylobacter, Salmonella, endpoints. American Journal of Hygiene 27:493–497. Yersinia, and Shigella species in feces with accuracy as high as that of culture. Ross, A. J., R. R. Rucker, and W. H. Ewing. 1966. Description of a bacterium Journal of Clinical Microbiology 48:2929–2933. associated with redmouth disease of Rainbow Trout (Salmo gairdneri). Cana- Danley, M. L., A. E. Goodwin, and H. S. Killian. 1999. Epizootics in farm-raised dian Journal of Microbiology 12:763–770. Channel Catfish, Ictalurus punctatus (Rafinesque), caused by the enteric Saitou, N., and M. Nei. 1987. The neighbor-joining method: a new method for redmouth bacterium Yersinia ruckeri. Journal of Fish Diseases 22:451–456. reconstructing phylogenetic trees. Molecular Biology and Evolution 4:406– Davies, R. L., and G. N. Frerichs. 1989. Morphological and biochemical dif- 425. ferences among isolates of Yersinia ruckeri obtained from wide geographical Tobback, E., A. Decostere, K. Hermans, F. Haesebrouck, and K. Chiers. 2007. areas. Journal of Fish Diseases 12:357–365. Yersinia ruckeri infections in salmonid fish. Journal of Fish Diseases 30:257– Eddy, F., A. Powell, S. Gregory, L. M. Nunan, D. V. Lightner, P. J. Dyson, A. 268. F. Rowley, and R. J. Shields. 2007. A novel bacterial disease of the Euro- Vuillaume, A., R. Brun, P. Chene, E. Sochon, and R. Lesel. 1987. First isolation pean shore crab, Carcinus maenas: molecular pathology and epidemiology. of Yersinia ruckeri from sturgeon, Acipenser baeri Brandt, in south west Microbiology 153:2839–2849. of France. Bulletin of the European Association of Fish Pathologists 7:18– Eissa, A. E., M. Moustafa, M. Abdelaziz, and N. A. Ezzeldeen. 2008. Yersinia 19. ruckeri infection in cultured Nile Tilapia, Oreochromis niloticus, at a semi- Wei, Q., J. He, D. Yang, W. Zheng, and L. Li. 2004. Status of sturgeon aqua- intensive fish farm in lower Egypt. African Journal of Aquatic Science culture and sturgeon trade in China: a review based on two recent nationwide 33:283–286. surveys. Journal of Applied Ichthyology 20:321–332. Fan, F. L., K. Y. Wang, Y. Geng, X. L. Huang, and D. F. Chen. 2010. Isolation, Wei, Q., F. Ke, J. Zhang, P. Zhuang, J. Luo, R. Zhou, and W. Yang. 1997. identification and phylogenetic analysis of Yersinia ruckeri in Channel Catfish Biology, fisheries, and conservation of sturgeons and paddlefish in China. Ictalures punctatus. Oceanologia et Limnologia Sinica 41:862–868. Environmental Biology of Fishes 48:241–255. Fernadez,´ L., I. Mequez,´ and J. A. Guijarro. 2004. Identification of specific in Xu, B. H., Z. Yin, Y. S. Chen, and Y. S. Wu. 1991. An outbreak of a new epi- vivo-induced (ivi) genes in Yersinia ruckeri and analysis of ruckerbactin, a zootic in Silver Carp and Bighead Carp: Yersinia ruckeri, a new pathogen catecholate siderophore iron acquisition system. Applied and Environmental of Silver Carp and Bighead Carp. Chinese Science Bulletin 36:1825– Microbiology 70:5199–5207. 1825. Downloaded by [Department Of Fisheries] at 18:30 20 January 2013 This article was downloaded by: [Department Of Fisheries] On: 20 January 2013, At: 18:31 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Aquatic Animal Health Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/uahh20 Mortality of American Bullfrog Tadpoles Lithobates catesbeianus Infected by Gyrodactylus jennyae and Experimentally Exposed to Batrachochytrium dendrobatidis Linda J. Paetow a , J. Daniel McLaughlin a , Bruce D. Pauli b & David J. Marcogliese c a Department of Biology, Concordia University, 7141 Sherbrooke Street West, Montreal, Quebec, H4B 1R6, Canada b National Wildlife Research Centre, Environment Canada, 1125 Colonel By Drive, Ottawa, Ontario, K1A 0H3, Canada c Aquatic Biodiversity Section, Watershed Hydrology and Ecology Research Division, Water Science and Technology Directorate, Science and Technology Branch, Environment Canada, 105 McGill Street, Montreal, Quebec, H2Y 2E7, Canada Version of record first published: 04 Jan 2013.

To cite this article: Linda J. Paetow , J. Daniel McLaughlin , Bruce D. Pauli & David J. Marcogliese (2013): Mortality of American Bullfrog Tadpoles Lithobates catesbeianus Infected by Gyrodactylus jennyae and Experimentally Exposed to Batrachochytrium dendrobatidis , Journal of Aquatic Animal Health, 25:1, 15-26 To link to this article: http://dx.doi.org/10.1080/08997659.2012.722170

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material. Journal of Aquatic Animal Health 25:15–26, 2013 C American Fisheries Society 2013 ISSN: 0899-7659 print / 1548-8667 online DOI: 10.1080/08997659.2012.722170

ARTICLE

Mortality of American Bullfrog Tadpoles Lithobates catesbeianus Infected by Gyrodactylus jennyae and Experimentally Exposed to Batrachochytrium dendrobatidis

Linda J. Paetow* and J. Daniel McLaughlin Department of Biology, Concordia University, 7141 Sherbrooke Street West, Montreal, Quebec H4B 1R6, Canada Bruce D. Pauli National Wildlife Research Centre, Environment Canada, 1125 Colonel By Drive, Ottawa, Ontario K1A 0H3, Canada David J. Marcogliese Aquatic Biodiversity Section, Watershed Hydrology and Ecology Research Division, Water Science and Technology Directorate, Science and Technology Branch, Environment Canada, 105 McGill Street, Montreal, Quebec H2Y 2E7, Canada

Abstract The fungal pathogen Batrachochytrium dendrobatidis (Bd), which causes the disease chytridiomycosis in postmetamorphic amphibians, has been linked to amphibian population declines. Different amphibian species, however, exhibit different susceptibility to Bd pathogenicity. At the same time, agricultural pesticides commonly found contaminating aquatic habitats have been reported to increase the susceptibility of amphibians to pathogens. To investigate whether certain pesticides are able to alter the pathogenicity of Bd to larval amphibians, we exposed larval American bullfrogs Lithobates catesbeianus to end-use formulations of the herbicides atrazine or glyphosate, and then exposed them to Bd. Following the experimental exposures, a preexisting infection of the tadpoles by the monogenean ectoparasite Gyrodactylus jennyae was detected in all experimental and control tadpoles. Gyrodactylus jennyae infection intensity varied, and individuals with heavy G. jennyae infections suffered more skin erosion due to grazing by the parasite. Tadpoles experimentally exposed to Bd,ortoBd and either herbicide, had signiﬁcantly reduced survival rates compared with untreated tadpoles that were only infected by G. jennyae. Increased mortality was also correlated with degree of skin erosion; survival of tadpoles with severe skin erosion was signiﬁcantly reduced compared with that of

Downloaded by [Department Of Fisheries] at 18:31 20 January 2013 tadpoles with no, or mild, skin erosion. While infected with G. jennyae, the group of tadpoles with the lowest survival rate (exposed only to Bd) included signiﬁcantly more individuals exhibiting severe skin erosion and signiﬁcantly fewer individuals without skin erosion, compared with the control group. These results emphasize the potential pathogenicity of gyrodactylid infections in larval amphibian hosts and suggest that concomitant exposures to Bd may enhance infections and effects of G. jennyae in bullfrog tadpoles.

Amphibians constitute the largest fraction of vertebrate amphibian species are experiencing population declines, with biomass in some ecosystems and are often important drivers several species faced with extinction (Stuart et al. 2004). In cer- of trophic dynamics (Blaustein et al. 1994). Therefore, it is tain amphibian species, the fungal pathogen Batrachochytrium a major conservation concern that almost half of the world’s dendrobatidis (Bd) can cause mass die-offs and Bd is deemed

*Corresponding author: [email protected] Received February 3, 2012; accepted August 10, 2012 15 16 PAETOW ET AL.

a proximate cause of the observed population declines (Fisher also ingest dermal cells, mucous cells, and melanocytes, and et al. 2009). It remains unclear whether Bd has become an im- they are well known for their exponential population growth on portant pathogen of amphibians because it is spreading through hosts living in stressful conditions (Bakke et al. 2007). Further, na¨ıve populations, or whether it has become more virulent or the feeding and attachment wounds suffered by ﬁsh carrying a host susceptibility has increased, due to changing environmental heavy gyrodactylid infection can destroy the osmotic integrity of factors (Fisher et al. 2009). In addition, the widespread occur- the epidermis and encourage secondary infections (Bakke et al. rence of pesticides in the environment (Relyea and Hoverman 2007). In the present study, the ubiquitous presence of the gyro- 2006), and evidence that some pesticides suppress amphibian dactylid parasite among all treatments and replicates presented immunity, making these species more vulnerable to pathogenic a unique opportunity to examine the combined effects of Bd in infections (e.g., Kiesecker 2002; Christin et al. 2004; Rohr et al. conjunction with another potential pathogen on the survival of 2008a; Gibble and Baer 2011), represents a further concern. bullfrog tadpoles exposed to pesticides. Studies of the combined This highlights a need to investigate whether pesticide exposure effects of Bd and another parasite on amphibians are rare (e.g., increases the pathogenicity of Bd in amphibians (Daszak et al. Romansic et al. 2011), and to our knowledge, none have con- 1999). sidered the additional effect of exposure of hosts to pesticides. Batrachochytrium dendrobatidis is a keratinophilic chytrid fungus that infects cutaneous epithelial cells of adult amphib- METHODS ians and the keratinized mouthparts of larval stages. In heavily Animal Husbandry infected postmetamorphic amphibians, it can cause the disease In February 2007, 102 American bullfrog tadpoles (Gosner chytridiomycosis (Berger et al. 1998). Although the patholog- stages 26–37; Gosner 1960) were obtained from an Environ- ical changes resulting from chytridiomycosis are more likely ment Canada laboratory located in Moncton, New Brunswick. to kill postmetamorphic animals (Berger et al. 1998; Longcore The tadpoles had originally been purchased from a commercial et al. 1999), oral Bd infections may threaten the survival of lar- source in the United States. After the tadpoles were received in vae as well, possibly indirectly (Blaustein et al. 2005; Garner our laboratory, and for 2 weeks prior to the start of the experi- et al. 2009; Venesky et al. 2010). ment, the tadpoles were acclimatized to a 16-h light : 8-h dark This study aimed to investigate whether larval American cycle and a water temperature of 21 ± 1.3◦C, and maintained bullfrogs Lithobates catesbeianus (formerly Rana catesbeiana) on a diet of boiled lettuce, trout chow, and Spirulina tablets. exposed to the formulated agricultural herbicides Aatrex Liq- uid 480 (atrazine) and Roundup Original (glyphosate) suffered Maintenance of Fungal Cultures increased mortality following subsequent exposure to Bd. Bull- The strain of Bd used (JEL423) was provided by J.E. Long- frogs were used because they are frequent carriers of nonlethal core (University of Maine, Orono). It was isolated from amphib- infections by Bd at both life stages, and they often demon- ians dying during a Bd outbreak in Panama and is considered strate a natural tolerance or resistance to the fungus (Mazzoni a virulent strain (J.E. Longcore, personal communication). To et al. 2003; Daszak et al. 2004; Blaustein et al. 2005; Garner maintain Bd in culture, every 3 months 1 mL of culture was et al. 2006; Peterson et al. 2007), yet their susceptibility to Bd transferred to 50 mL of fresh sterilized 1% T broth (10 g of may change in the presence of other stressors (Mazzoni et al. tryptone per 1 L of distilled water, autoclaved, and cooled). 2003). Aatrex Liquid 480 and Roundup Original were used be- Penicillin-streptomycin (Sigma, St. Louis, Missouri) was added cause atrazine and glyphosate are two of the most widely used (10 mL per 1 L of culture) during each passage to prevent bacte- herbicides in the world (Ackerman 2007; Struger et al. 2008). rial contamination. New cultures were incubated at 23◦Cfor7d Although information regarding the immunomodulatory effects ◦

Downloaded by [Department Of Fisheries] at 18:31 20 January 2013 before storage at 4 C (J.E. Longcore, personal communication). of atrazine on amphibians is inconsistent (Du Preez et al. 2008; LaFiandra et al. 2008; Oka et al. 2008), evidence suggesting Experimental Design that it can suppress amphibian immune function is mounting All randomizations (e.g., the assignment of tadpoles to treat- (Larson et al. 1998; Rohr and Palmer 2005; Forson and Stor- ments and replicates following their haphazard capture from fer 2006; Rohr et al. 2006, 2008b; Brodkin et al. 2007). The their holding tanks, and the selection of aquariums and petri potential immunomodulatory effects of glyphosate and its for- plates to measure concentrations of pesticides and zoospores, mulated products remain poorly understood and merit further respectively) were performed using a random number generator. study (Rohr et al. 2008a). Any tadpoles that, unknowingly to us, had already been infected Upon completion of the chemical and fungal exposures as- by Bd should therefore have been equally distributed among the sociated with the present experiment, an ectoparasite was dis- treatment and control groups at the start of the experiment, due covered to be infecting the experimental tadpoles. This para- to the random distribution process employed. Hence, any risk of site, a new species of Gyrodactylus (Monogenea) subsequently mortality associated with a preinfection by Bd would have been described as Gyrodactylus jennyae, belongs to a group of evenly distributed among the treatment and control groups. helminths that normally infects ﬁsh (Bakke et al. 2007; Pae- For the experimental exposures to Bd, test animals were ex- tow et al. 2009). All gyrodactylids are epidermal browsers that posed to zoospores harvested from the ongoing culture (see GYRODACTYLUS AND CHYTRID FUNGUS POTENTIALLY LETHAL TO BULLFROG TADPOLES 17

below). Four experimental treatment and control groups were Prior to each exposure, a stored culture of Bd was incubated established: (1) a group of tadpoles acutely exposed to two at 23◦C to stimulate growth. After 10 d of incubation, 1 mL doses of Bd zoospores (called ‘Bd’), (2) a group exposed for aliquots of culture were used to inoculate 1% T agar plates 14 d (see below) to atrazine, followed by two acute doses of (10 g of tryptone and 10 g of agar per 1 L of distilled water zoospores (called ‘atrazine + Bd’), (3) a group exposed for 14 d that had been autoclaved and cooled). Aliquots of plain 1% T to glyphosate, followed by two acute doses of zoospores (called broth (lacking culture but supplemented with 10 mL penicillin- ‘glyphosate + Bd’), and (4) a group that received no experimen- streptomycin per l L of broth) were pipetted onto control plates. tal exposures (called ‘control’). There were four replicates for All plates were then sealed with Parafilm and incubated at 23◦C each group, except for the control group, which had five; each for an additional 10 d, at which point visual inspection revealed replicate consisted of six tadpoles. Nominal chemical exposure that all Bd-inoculated plates exhibited uniform growth. concentrations were 21 μg of atrazine, and 1.5 mg acid equiv- First exposure.—At day 14 of the experiment (14 d after her- alents (a.e.) of glyphosate, per liter of water. In the herbicide- bicide exposures were initiated), three plates inoculated with Bd exposed animals, the two acute exposures to Bd occurred at 14 were randomly selected from the incubator to obtain an estimate and 16 d after the start of the pesticide exposures, and each Bd of the potential fungal zoospore harvest per plate and to confirm exposure lasted 24 h (see below). During all exposures, separate zoospore motility. Each plate was flooded with 3 mL of sterile gloves and nets were used for each tank to avoid the risk of distilled water to stimulate the release of zoospores from thalli cross-contamination. As the goal was to determine whether the (Longcore et al. 1999). After 30 min, the zoospore suspensions pesticide exposures enhanced susceptibility to infection by Bd, were decanted separately and zoospore harvest estimates were no pesticide-only exposures were included in the design. obtained using a hemocytometer. These harvest estimates were Each replicate group was maintained in a 33-L aquarium con- 11.20 × 106, 11.52 × 106, and 10.08 × 106 zoospores/mL taining 8 L of aerated, dechlorinated water. Three-quarters of from the three plates, respectively. Given the reasonably simi- the water in each tank was changed every 3 d. During the pesti- lar zoospore harvest per plate, we used a method of exposure cide exposure phase, the herbicides were replenished with each similar to that of Blaustein et al. (2005): two inoculated plates water change. Water volume was maintained at 1.3 L per tadpole were deposited directly into each treatment aquarium for 24 h. and reduced by 1 L with each tadpole’s death, to a minimum of Simultaneously, two control plates were deposited into each 4 L. At day 45, all surviving tadpoles were killed by immersion control aquarium. Based on this method we estimated that the in a 0.8% solution of MS-222 (tricaine methanesulfonate). Tad- Bd-exposed tadpole groups were exposed to a mean ± SD dose poles that died or were killed were immediately fixed in Bouin’s of at least 2,733 ± 189 zoospores/mL in their aquariums. Addi- solution, in a 70% solution of ethanol, or in a 10% solution of tional zoospores would have been released into the water during buffered formalin (two tadpoles in each solution per aquarium). the 24-h period that the plates were in the aquaria, making these estimates slightly conservative. Pesticide Exposures Second exposure.—At day 16, all 102 tadpoles were sepa- At day 0, a stock solution of each herbicide was prepared. rated into individual 1.4-L plastic containers containing 1 L of Stock solutions of the end-use formulations of the herbicides dechlorinated, aerated water. Four plates inoculated with Bd were made in distilled water at 1,000 times the nominal expo- were flooded with water as above, but all four plates were sure concentration and were vortexed (3,000 RPM for 5 min) decanted into the same container. This provided a harvest of to ensure the herbicides went into solution. Stock solutions 4.28 × 107 zoospores in 12 mL of solution, based on an esti- were stored at 4◦C in amber flasks sealed with Parafilm dur- mate using a hemocytometer. To this solution, 60 mL of sterile ing the 14-d exposure period. During the pesticide exposure distilled water was added, and 1 mL aliquots of the diluted so- Downloaded by [Department Of Fisheries] at 18:31 20 January 2013 phase of the study, each liter of water that was siphoned from lution were pipetted into each of the 72 containers holding tad- the exposure tanks every third day was replaced with 1 mL poles receiving the Bd exposure. The estimated exposure dose of pesticide stock solution that was vortexed once again and per tadpole was 594 zoospores/mL for this second 24-h pe- combined with 999 mL of dechlorinated filtered tap water. At riod. The 30 containers holding control tadpoles received 1 mL day 14, water changes were made without herbicide addition, in aliquots of sterile distilled water decanted from control 1% T preparation for Bd challenges. At two time points (at 12 h after agar plates. After 24 h, all tadpoles were regrouped in their orig- herbicide addition and at day 14), water samples were drawn inal 17 aquariums, which contained only dechlorinated water. directly from two randomly selected pesticide-containing aquar- The tadpoles were left in their aquariums until the experiment iums (one containing atrazine and one containing glyphosate) was terminated at day 45. to compare the actual and nominal concentrations. Pesticide analyses were performed at the Centre d’Expertise en Analyze Confirmation of Infection by Batrachochytrium Environnementale du Quebec´ (Quebec City, Canada). dendrobatidis Batrachochytrium dendrobatidis Exposures To confirm infection by Bd, the oral disk of a tadpole that Batrachochytrium dendrobatidis exposures were based on a died 30 d post-Bd inoculation was immediately excised and a protocol provided by J.E. Longcore (personal communication). fresh tissue fragment was examined under a light microscope, at 18 PAETOW ET AL.

FIGURE 1. (A) Larval American bullfrog exhibiting a moderate amount of skin erosion caused by dermal foraging of the monogenean Gyrodactylus jennyae. Some tadpoles sustained more severe skin damage with skin erosion also occurring elsewhere on the body; (B) numerous monogeneans, visible as white, elongated flecks, attached to the surface of the eye and surrounding skin of an infected bullfrog tadpole. [Figure available online in color.] 200× and 400×. In addition, one tadpole from each of the three or were killed, an attempt was made to count the gyrodactylids treatment groups that had been exposed to Bd was also ran- on their body surface in order to relate gyrodactylid infection domly selected at the end of the experiment to confirm infection levels to tadpole mortality. This was not possible, however, due by histological analysis (Berger et al. 2000). When killed, these to the high numbers of gyrodactylids, their small size, and their three tadpoles were immediately fixed in buffered 10% formalin continuous movement on the hosts; one tadpole that died had and later transferred to 70% ethanol. They were then decapitated at least 104 gyrodactylids around the oral region and inside the and their heads were processed through graded concentrations buccal cavity alone. Therefore, the degree of skin erosion (none, of ethanol and xylene, embedded in paraffin, and serially sec- mild, moderate, severe) exhibited by each tadpole was recorded tioned along the anteroposterial axis at a thickness of 8 μmsoas and used as a surrogate measure of intensity of gyrodactylid to make all tooth rows and both jaw sheaths visible for analysis. infection. The term “mild” was defined as patches of erosion The sections were mounted on slides, stained with hematoxylin on the face that were smaller than noneroded zones, with little and eosin, and examined under a light microscope at 400×. contrast of color between zones; “moderate” referred to patches of erosion on the face that were large compared with noneroded Gyrodactylus jennyae Infections zones, with a high contrast of color between zones; “severe” de- By day 30 of the experiment, a consistent pattern of dermal scribed complete erosion of skin on the face and at the tail base, depigmentation was observed around the eyes and at the base of with a high contrast in color between eroded and noneroded skin. the tail of tadpoles in every group, including the control group Downloaded by [Department Of Fisheries] at 18:31 20 January 2013 (Figure 1A). At days 30 and 31, two tadpoles that displayed Histology severe depigmentation (each from the group exposed to both The skin of one control tadpole killed at day 45, exhibiting glyphosate and Bd) were removed from the experiment for ex- a mild infection by G. jennyae, was examined histologically to amination. Depigmentation was found to be the result of skin observe pathological changes associated with the infection. The erosion (see below) caused by the tadpoles’ infection by the tadpole was fixed in a 10% solution of buffered formalin and monogenean ectoparasite G. jennyae (Figure 1B; Figure 2; Pae- later transferred to a 70% solution of ethanol. The front third of tow et al. 2009). Further examination revealed that G. jennyae its head was removed, dehydrated, cleared through a series of was present in each aquarium. ethanol and xylene solutions, embedded in paraffin, sectioned at 8 μm, mounted on slides, and stained with hematoxylin and Endpoint Measurements eosin. The total length (TL), body weight (BW), and Gosner developmental stage (Gosner 1960) of all tadpoles were recorded at Statistics day 0. During the experiment, survivorship was recorded daily. All statistical analyses were performed using SAS version The TL, BW, and Gosner stages were recorded again when the 9.1 (SAS Institute, Cary, North Carolina). Mixed-model nested tadpoles either died or were killed at day 45. When tadpoles died ANOVAs (PROC MIXED) were used to compare the mean GYRODACTYLUS AND CHYTRID FUNGUS POTENTIALLY LETHAL TO BULLFROG TADPOLES 19

of individual growth and development over the course of the experiment was not possible because tadpoles within the replicates were not individually marked. The survival analysis was performed using PROC LIFETEST and PROC TPHREG in SAS. First, Kaplan–Meier plots of survival (S) versus time (T) and log(–log(S)) versus log(T)produced by PROC LIFETEST were examined to confirm the assumption of proportional hazards. Subsequently, Cox proportional hazards modeling (PROC TPHREG) and the likelihood ratio test were used to test the null hypothesis that there was no difference in survival among the treatment groups. Cox regression modeling was then repeated with three additional covariates in the model to evaluate potential prognostic factors for mortality: (1) Gosner stage at time of death (because tadpoles can become infected by Bd in regions beyond the mouth when their skin becomes keratinized as they approach metamorpho- sis; Marantelli et al. 2004), (2) the degree of skin erosion caused by G. jennyae (to evaluate the effect of gyrodactylid infections on mortality), and (3) the presence or absence of scoliosis (an abnormal lateral curvature of the spine, which developed among some tadpoles, and can indicate nutritional deficiencies and other health problems). From the full model, a backward stepwise analysis was employed to identify the model with the best fit.

RESULTS Concentrations of the Pesticides The concentrations of glyphosate measured at 12 h and at 14 d postaddition of the pesticides to the aquariums were both 2.1 mg/L a.e., which was 1.4 times higher than the nominal concentration of 1.5 mg/L a.e. The concentrations of atrazine measured on the same days were 140 μg/L and 77 μg/L, respectively, which were 6.7 times and 3.7 times higher than the FIGURE 2. Monogenean Gyrodactylus jennyae found on larval American nominal concentration of 21 μg/L. bullfrogs. Gyrodactylids were observed foraging on the surface of bullfrog tadpoles during the experiment. An accumulation of host pigment granules is visible in the intestinal cecae of this worm indicating that it was feeding on the Conﬁrmation of Infection by Batrachochytrium dermal tissues of the host tadpole. The hooks with which gyrodactylids attach dendrobatidis themselves to host skin are visible at the posterior end (arrows). Scale bar = Infection by Bd was conﬁrmed in the Bd-exposed tadpole Downloaded by [Department Of Fisheries] at 18:31 20 January 2013 100 μm. With kind permission from Springer Science + Business Media: Syst. that died 30 d after inoculation. Sporocysts were observed en- Parasitol., Morphology and molecular taxonomy of Gyrodactylus jennyae n.sp. cysted in unkeratinized cells of its oral tissues when viewed by (Monogenea) from tadpoles of captive Rana catesbeiana (Anura) with a review × of species of Gyrodactylus parastizing amphibians, 73, 2009, 219–227, Paetow, light microscopy at 200 . One sporocyst was observed shed- L., D.K. Cone, T. Huyse, J.D. McLaughlin, and D.J. Marcogliese, Figure 1A. ding active zoospores when examined more closely at 400×, while the others contained live zoospores that were not being shed. Histological examination of the oral tissues of the three TL and BW of tadpoles among the treatment and control tadpoles killed at day 45, however, yielded no further evidence groups at the start of the experiment. Nonparametric nested of infection. ANOVAs (PROC MIXED on ranked data) were used to compare the initial Gosner developmental stages because the residuals of Experimental Results the data were not normally distributed. Similar analyses were At day 0, the mean ± SD TL, BW, and Gosner stage of used to compare the TL, BW, and Gosner stages of the tadpoles development of all tadpoles (N = 102) were 95.85 ± 6.66 mm, that survived until the end of the experiment. For all ANOVAs, 8.10 ± 1.55 g, and 33.5 ± 2.8, respectively. None of these the Satterthwaite approximation corrected the degrees of free- measurements differed among the treatments (initial TL: nested dom to account for unbalanced distributions of data. Assessment ANOVA F3, 13 = 0.35, P = 0.788; initial BW: nested ANOVA 20 PAETOW ET AL.

FIGURE 4. Percent survival of 100 larval American bullfrogs at the end of FIGURE 3. Kaplan–Meier survival curves of four experimental treatment a 45-d experiment. Control tadpoles experienced signiﬁcantly greater survival groups of larval American bullfrogs monitored over 45 d. Tadpoles were either (*) than those in the three treatment groups (planned statistical contrasts, P < exposed or not to an herbicide (atrazine or glyphosate) for 14 d, and then ex- 0.050). Bullfrog tadpoles were either exposed or not to an herbicide (atrazine or posed to Batrachochytrium dendrobatidis (Bd) on days 14 and 16. All groups, glyphosate) for 14 d, and then exposed to the fungal pathogen Batrachochytrium including controls, had preexisting infections of the monogenean parasite Gyro- dendrobatidis (Bd) on days 14 and 16. All groups of tadpoles, including the con- dactylus jennyae. Survival differed among the treatment groups (likelihood ratio trol group, harbored preexisting infections by the monogenean parasite Gyro- test: P = 0.003): control tadpoles had a signiﬁcantly greater chance of survival dactylus jennyae. Two tadpoles exposed to glyphosate and Bd that were removed compared with tadpoles in the three treatment groups (statistical contrasts: P < from the experiment to diagnose infection by G. jennyae were excluded from 0.050). the analysis of survival.

F3, 13 = 1.14, P = 0.371; initial Gosner stage: nonparametric difference among the treatment groups over the course of the 2 nested ANOVA: F3, 13 = 0.63, P = 0.606). experiment (likelihood ratio test: df = 3; X = 13.80; P = 0.003; The final Gosner stages of two tadpoles went unrecorded. effect of treatment, type 3 tests: df = 3; X2 = 9.04; P = 0.029). An additional tadpole that reached Gosner stage 43 (beginning Planned statistical contrasts performed with the Cox regression of tail resorption; Gosner 1960) was excluded from the analysis analysis revealed that survival in the control group was greater comparing final TL due to obvious tail resorption. The data from over the course of the experiment than in the three treatment another tadpole were excluded from the analyses comparing all groups (control versus Bd-exposed: df = 1; X2 = 8.99; P = final measurements because it was the only surviving tadpole in 0.003; control versus glyphosate and Bd-exposed: df = 1; X2 = its aquarium, although its inclusion did not change the interpre- 6.33; P = 0.012; control versus atrazine and Bd-exposed: df = tation of the results. Excluding these data, the mean ± SD TL, 1; X2 = 5.79; P = 0.016) (Figure 4). BW, and Gosner stage of development of the surviving tadpoles Inclusion of the three potential prognostic factors in the Cox at day 45 were 92.61 ± 6.92 mm, 6.17 ± 1.96 g, and 35.2 ± regression still revealed a difference in survival among the treat- 1.5, respectively. None of these measurements differed among ment groups (likelihood ratio test: df = 17; X2 = 41.24; P = the treatments (final TL: nested ANOVA N = 62, F3, 10.6 = 0.45, 0.0009). At the same time, the P-value associated with the like- P = 0.720; final BW: N = 63, nested ANOVA F3, 12.1 = 0.06, lihood ratio test decreased (i.e., from P = 0.003), indicating Downloaded by [Department Of Fisheries] at 18:31 20 January 2013 P = 0.978; final Gosner stage: nonparametric nested ANOVA: that including Gosner stage, degree of skin erosion, and scolio- N = 61, F3, 57 = 1.13, P = 0.345). sis in the full model provided a better fit to the data than just Survival.—The Kaplan–Meier survival curves of the treat- the experimental treatment. The fit improved further with the ment groups are shown in Figure 3. No mortality occurred during removal of Gosner stage (likelihood ratio test: df = 7; X2 = the experimental exposures to the herbicides or Bd. However, 33.09; P < 0.0001) and scoliosis (likelihood ratio test: df = 6; following exposures to the herbicides and Bd (i.e., between day X2 = 31.68; P < 0.0001) during the backward step analysis. 17 and the end of the experiment at day 45), 36 (35.3%) of Hence, the Cox model that included experimental treatment and 102 tadpoles died (including 3 of 30 control tadpoles, 13 of degree of skin erosion as covariates provided the best overall fit 24 tadpoles exposed only to Bd, 10 of 22 tadpoles exposed to to the survival data. However, while this model indicated that glyphosate and Bd, and 10 of 24 tadpoles exposed to atrazine the effect of skin erosion on survival was significant, the effect and Bd). Two tadpoles exposed to glyphosate and Bd were re- of treatment no longer was (type 3 tests: effect of skin erosion: moved to diagnose the gyrodactylid infection (they were the df = 3; X2 = 17.89; P = 0.0005; effect of treatment: df = 3; X2 = first tadpoles to exhibit severe degrees of skin erosion), while 2.04; P = 0.564). This suggests that the degree of skin erosion the remaining 64 tadpoles (62.7%) survived (Figure 4). The was either a significant prognostic factor for survival, or that results of the Cox regression on survival data demonstrated a its effect was confounded with that of experimental treatment. GYRODACTYLUS AND CHYTRID FUNGUS POTENTIALLY LETHAL TO BULLFROG TADPOLES 21

TABLE 1. Cox proprotionate hazards regression model demonstrating the effects of experimental treatment and degree of skin erosion on the survival of larval American bullfrogs, all preinfected with Gyrodactylus jennyae and experimentally exposed to either atrazine (nominal concentration of 21 μg/L), glyphosate (nominal concentration of 1.5 mg a.e./L), or no herbicide for 14 d, followed or not by exposure to the chytrid fungus Batrachochytrium dendrobatidis (Bd) (signiﬁcant when P < 0.050).

95% CI for hazard ratio Regression Standard Hazard Variable coefﬁcient error P-value ratio Upper Lower Degree of skin erosiona Mild 0.838 0.845 0.321 2.313 0.441 12.122 Moderate 1.495 0.830 0.072 4.458 0.877 22.673 Severe 2.534 0.784 0.001 12.602 2.711 58.568 Experimental treatmentb Bd 0.940 0.685 0.170 2.560 0.669 9.802 Atrazine + Bd 0.701 0.696 0.314 2.016 0.515 7.885 Glyphosate + Bd 0.899 0.706 0.203 2.458 0.616 9.813

aEach degree of skin erosion is compared with no skin erosion. bEach experimental treatment is compared with no treatment; Bd = only exposed to Bd; Aatrazine + Bd = exposed to atrazine followed by Bd; glyphosate + Bd = exposed to glyphosate followed by Bd.

The TPHREG Procedure improved the precision with which was marginally nonsignificant (α = 0.05/6; df = 1, X2 = 6.64, we could estimate the treatment effect by modeling the hazard P = 0.010) following the Bonferroni adjustment (Table 2). function on the variables “experimental treatment” and “degree Histology.—Histological examination of a control tadpole of skin erosion” (Table 1). This procedure showed that, after mildly infected by G. jennyae suggested that the monogeneans adjustment for the prognostic factor “degree of skin erosion”, fed on the epidermis and on the stratum spongiosum of the none of the experimental treatments had an effect on survival compared with the controls (i.e., compared with tadpoles that to our knowledge were only preinfected with G. jennyae). That is, the P-values associated with each treatment were greater than 0.050 and each of the 95% confidence intervals (CI) associated with the treatment hazard ratios included 1 (Table 1). The TPHREG Procedure also showed that the prognostic factor “degree of skin erosion” had an increasing effect on mortality as the erosion worsened (Table 1). In particular, it revealed that a “severe” degree of skin erosion had a significant effect on survival compared to “no skin erosion” even after adjustment for the experimental treatment variable, because the P-value associated with the comparison was less than 0.050 and the 95% CI Downloaded by [Department Of Fisheries] at 18:31 20 January 2013 associated with the hazard ratios did not include 1. Posthoc statistical contrasts, corrected for multiple comparisons, revealed that tadpoles with severe skin erosion experienced reduced sur- FIGURE 5. Percent survival of 95 larval American bullfrogs that exhibited vival compared with tadpoles with no or only mild skin erosion various degrees of skin erosion at the time of death. Tadpoles exhibiting severe 2 (α = 0.05/6; no versus severe erosion: df = 1, X = 10.4, skin erosion experienced reduced survival compared with tadpoles exhibiting P = 0.001; mild versus severe erosion: df = 1, X2 = 10.3, P = no or mild skin erosion (posthoc statistical contrasts corrected for multiple 0.001) (Figure 5). Additional posthoc tests (statistical contrasts comparisons: α = 0.050/6). Different superscripts above the columns indicate < performed with logistic regression, using PROC GENMOD and significant differences (P 0.0083). Skin erosion was caused by dermal foraging by the monogenean parasite Gyrodactylus jennyae. Five out of 102 experimental also corrected for multiple comparisons) showed that the con- tadpoles (including one control tadpole that survived, two tadpoles exposed only trol group had a lower proportion of tadpoles with severe skin to Bd that died, one tadpole exposed to glyphosate and Bd that died and one erosion compared with the group only exposed to Bd (α = tadpole exposed to atrazine and Bd that died) are not included in the figure 0.05/6; df = 1, X2 = 7.09, P = 0.0078) (Table 2). The two other because the degrees of skin erosion that they exhibited were not recorded. groups of tadpoles exposed to Bd and pesticides had an inter- Two additional tadpoles that were exposed to glyphosate and Bd and that each developed severe degrees of skin erosion are not accounted for in the figure mediate proportion of animals with severe skin erosion, and in because they were removed from the experiment and killed to diagnose infection the case of tadpoles exposed to atrazine and Bd, the difference by G. jennyae. 22 PAETOW ET AL.

TABLE 2. Proportion of larval American bullfrogs exhibiting various degrees of skin erosion caused by the monogenean parasite Gyrodactylus jennyae, in each group. Data were recorded at the time of death (natural or induced). Shown are percentages with actual proportions in brackets. Signiﬁcant differences within the columns are demonstrated with different letters (posthoc statistical contrasts corrected for multiple comparisons, P < 0.008).

Proportion with degree of skin erosionb Exposure treatment None Mild Moderate Severe Control 51.7 (15/29) z 31.0 (9/29) 10.3 (3/29) z 6.9 (2/29) z Bd a 13.6 (3/22) y 27.2 (6/22) 22.7 (5/22) zy 36.4 (8/22) y Glyphosate + Bd 13.0 (3/23) y 17.4 (4/23) 47.8 (11/23) y 21.7 (5/23) zy Atrazine + Bd 26.1 (6/23) zy 30.4 (7/23) 8.7 (2/23) z 34.8 (8/23) zy

aBd = Batrachochytrium dendrobatidis. bThe degrees of skin erosion exhibited by one control tadpole, two tadpoles exposed only to Bd, one tadpole exposed to glyphosate and Bd, and one tadpole exposed to atrazine and Bd were not recorded. dermis, which included chromatophores (Figure 6). This pos- depigmentation in highly infected zones of their skin. The de- sibility was further supported by knowledge that, while gyro- gree of skin erosion that the tadpoles exhibited was found to be dactylid parasites of fish feed mainly on host epidermal tissues, a significant prognostic factor of tadpole survival indicating a dermal tissues of the host can occasionally be ingested as well main pathogen effect of G. jennyae on host survival. This study (Bakke et al. 2007). demonstrates unequivocally the potential for gyrodactylids to be important pathogens of larval amphibian hosts under conditions that are favorable to the parasite. DISCUSSION Other studies have demonstrated that the pathogenicity of Exposure to Bd, even without manifestation of infection or Gyrodactylus spp. can increase with concomitant exposure of disease, can prove detrimental to the health and survival of their hosts to other stressors (e.g., Moles and Wade 2001; larval amphibians (Blaustein et al. 2005; Blaustein et al. 2009; Gheorgiu et al. 2006; Xu et al. 2007). This is greatly due to Garner et al. 2009). However, amphibian larvae exhibit interspe- a hyperviviparous reproductive strategy and direct transmis- cific variability in susceptibility to Bd and bullfrog tadpoles can sion between hosts, which allow gyrodactylids to expand their display resistance to lethal infections (Blaustein et al. 2005). In populations exponentially when unimpeded by a host response the present study, bullfrog tadpoles that were experimentally ex- (Bakke et al. 2007). We provide circumstantial evidence that posed to Bd while harboring preexisting infections of G. jennyae concomitant exposure to Bd can intensify the pathogenicity of experienced significantly greater mortality than gyrodactylid- G. jennyae infecting bullfrog tadpoles. Albeit, when symptoms infected tadpoles that were not exposed to Bd. Cox regression of disease, including skin erosion, appeared in the tadpoles, they modeling indicated that infection by G. jennyae was directly arose in all groups. This indicates that the overall experimen- linked to elevated bullfrog tadpole mortality while experimental tal conditions were somehow favorable for the proliferation of exposure to Bd was not. The evidence of an interactive effect G. jennyae, perhaps because G. jennyae benefitted from in- between G. jennyae and Bd was limited, but the possibility that creased host stress associated with conditions in the experimen- the experimental exposures to Bd facilitated the proliferation of tal aquariums compared with the holding tanks. However, the G. jennyae to highly pathogenic levels, thereby affecting tadpole group of tadpoles that was exposed only to Bd had a significantly survival indirectly, could not be ruled out. higher proportion of animals presenting severe skin erosion than Downloaded by [Department Of Fisheries] at 18:31 20 January 2013 Epizootics of Gyrodactylus spp. cause extensive mortality in the group that was not exposed to Bd. In addition, tadpoles in the wild and captive fish populations (Buchmann and Lindenstrøm two groups exposed to pesticides and Bd had an intermediate 2002). While few species of Gyrodactylus have thus far been proportion of tadpoles exhibiting severe skin erosion. Interest- described as parasites of amphibians (Paetow et al. 2009), high ingly, our results resemble those of another experiment, in which mortality rates among captive bullfrog tadpoles infected with concomitant exposures to Bd and the larval trematode Ribeiroia Gyrodactylus spp. have been reported (Wootton et al. 1993; spp. caused increased mortality of larval Pacific treefrogs Pseu- Crawshaw 1997). Gyrodactylids are epidermal-browsing par- dacris regilla compared with exposures to either pathogen alone asites that can destroy the osmotic integrity of epidermal tis- (Romansic et al. 2011). As in this study, the statistical evidence sues and encourage secondary infections in hosts via feeding that an interaction between the trematode and Bd occurred was and attachment wounds (Buchmann and Lindenstrøm 2002; limited (Romansic et al. 2011). Bakke et al. 2007). The pathology that we observed in histo- Two potential issues should be addressed with respect to the logical preparations of the tadpoles was consistent with feeding possible interaction between Bd and G. jennyae. First, follow- and tissue damage typically caused by gyrodactylids (Buch- ing the experimental exposures to Bd, we only found evidence mann and Lindenstrøm 2002). Gyrodactylid-infected tadpoles of infection by Bd in one of the four tadpoles that we exam- in each of the four experimental groups suffered skin erosion and ined for diagnostic purposes (i.e., the tadpole that died 30 d GYRODACTYLUS AND CHYTRID FUNGUS POTENTIALLY LETHAL TO BULLFROG TADPOLES 23

postinoculation). It is therefore unknown whether the majority of tadpoles failed to become infected by Bd during the experimental fungal exposures or whether they acquired infections that they then cleared. The former possibility could raise questions regarding whether opportunities actually arose for within-host interactions to occur between Bd and G. jennyae. However, exposure to Bd may adversely affect larval amphibian health with or without establishment of infection (Blaustein et al. 2009). In one instance, Blaustein et al. (2005) reported higher rates of mortality in larval western toads Bufo boreas exposed to Bd than in controls, yet none of the Bd-exposed larval toads that died provided histologic evidence of chytrid infection. There- fore, the possibility that Bd produces lethal toxins was raised (Blaustein et al. 2005). In another instance, Garner et al. (2009) found that exposure of larval amphibians to Bd always incurs a growth cost and can lead to larval mortality before meta- morphosis, even when individuals do not exhibit infection at time of death. Among the hypotheses that they put forward to explain this outcome was the possibility that Bd-exposed tadpoles invested an overwhelming amount of energy to prevent infection, perhaps via mechanisms that inhibit zoospore attachment to host cells (Garner et al. 2009). It follows that a decline in physiological condition can make an organism more susceptible to other types of pathogenic infection (Beldomenico and Begon 2010). At the same time, when larval amphibians do acquire oral Bd infections, their foraging efficiency may be compromised (Venesky et al. 2010), possibly leading to nutritional deficiencies that can in turn diminish both host condition and immuno- competency (Beldomenico and Begon 2010). Alternatively, any stress associated with exposure to Bd could diminish host im- munocompetence via the action of stress hormones (Sapolsky 2002). The second issue to be addressed is the fact that the tadpoles we used were not tested prior to the exposures to determine if they had already acquired infections by Bd from other sources. Therefore, it cannot be confirmed that our “control” group was uninfected by Bd prior to the exposures (while we still call them controls in the figures and tables). However, while bullfrogs are frequent carriers of Bd infections, they also demonstrate a toler- Downloaded by [Department Of Fisheries] at 18:31 20 January 2013 ance to those infections, which generally remain asymptomatic (Daszak et al. 2004; Blaustein et al. 2005; Garner et al. 2006; Peterson et al. 2007). In the present study, previous infections by Bd or of any other kind were benign in the experimental tadpoles during the several months prior to experimentation (i.e., the total amount of time that the tadpoles were held captive in FIGURE 6. Hematoxylin and eosin-stained sections through the skin of a two laboratory facilities) and until all of the experimental ex- larval American bullfrog infected by the monogenean parasite Gyrodactylus posures were completed. During this time, all of the tadpoles jennyae. (A) A cross-section of G. jennyae (Gj) observed on epidermal tissues were healthy looking, vigorously active and with good appetite, (E) of tadpoles; (B) the epidermis (E), where stratum spongiosum (SS)ofthe dermis and chromatophores (in the SS) were present in a region unfavored by indicating that they were good candidates to undergo the exper- the parasites (i.e., immediately behind the head) but (C) absent (arrow) in the iment that we conducted. In addition, we exposed the tadpoles zone heavily infected by the parasites (i.e., the face). (PG = poison gland; toastrainofBd that was isolated in Panama (JEL423), which SC = stratum compactum of the dermis). [Figure available online in color.] the tadpoles very likely had never been exposed to. As the virulence of Bd varies significantly with the strain (Fisher et al. 2009), the results of this study should be viewed in light of the 24 PAETOW ET AL.

possibility that any preinfections by Bd that the bullfrog tad- nia recently were reported to be coinfected by Bd, G. aurorae, poles may have harbored were strains that the tadpoles were and other parasites (Nieto et al. 2007). Interestingly, while in- tolerant to, while the experimental exposures to strain JEL423 fection by G. aurorae was widespread among sampling sites, of Bd represented a new immunological challenge to the tad- it was also associated with a reduced risk of infection by Bd; poles, which altered their physiological condition. In this sense, the degree of parasite diversity and total parasite abundance in we argue that the tadpoles that were not experimentally exposed individual tadpoles infected with Bd were significantly lower to strain JEL423 of Bd during the course of the study could than in Bd-uninfected tadpoles (Nieto et al. 2007). The authors be considered as experimental controls. Nevertheless, with the proposed that Bd may outcompete other parasitic species on random method that we used to assign the tadpoles to their re- tadpole hosts, or alternatively, that tadpoles infected by both Bd spective experimental groups, we should have evenly distributed and other parasites may not be surviving. Our findings appear tadpoles with preinfections by any type of pathogen among the to support the latter hypothesis. groups, such that the average state of health of all tadpoles upon Our results show that populations of G. jennyae survived initiation of the experiment was equal. Thus, while we cannot direct exposure to the pesticides used in this experiment. This is state unequivocally that none of our control tadpoles harboured consistent with the findings of Dusekˇ et al. (1998), who noted Bd infections prior to the experimental exposures, we are con- that gyrodactylid communities inhabiting the skin and fins of fident that our results demonstrate effects of the experimental European Chub Leuciscus cephalus appeared to be resistant to exposures to Bd and the pesticides on the initial state of tadpole environmental pollution relative to other monogeneans, and of health, as all other conditions were equal (see Blaustein et al. Gheorgiu et al. (2006), who found that G. turnbulli was more 2009 for a discussion on well-designed disease studies). resistant to waterborne zinc than their host Guppies Poecilia Despite the limited statistical evidence of an interaction be- reticulata. Other studies have reported similar findings (e.g., tween the pathogens, this study highlights the importance of Moles and Wade 2001, and references therein). considering net effects of concomitant pathogenic exposures There also was no significant difference in mortality between when studying amphibian disease dynamics (Busch et al. 2003; tadpoles exposed only to Bd and tadpoles additionally exposed Pederson and Fenton 2007; Graham 2008; Beldomenico and to atrazine or glyphosate, which suggests that the pesticide expo- Begon 2010; Telfer et al. 2010; Martins et al. 2011; Ottesen and sures did not alter the immune system or enhance the virulence Amin 2011; Romansic et al. 2011). Recent declines of amphib- of the pathogens in this experiment. However, the combined ian populations worldwide, even in remote and pristine areas, effects of contaminants and pathogens on hosts are context de- have prompted investigations into the causative factors involved pendent, and there are numerous examples whereby exposure (Stuart et al. 2004). Infectious diseases including chytridiomy- to pesticides increases susceptibility to pathogens (Marcogliese cosis and ranaviral disease, along with the individual pathogens and Pietrock 2011). Few studies have explored the potential that cause them, have received much attention as important con- influence of pesticide exposure on the virulence of Bd (Char- tributors of these declines (Daszak et al. 1999; Romansic et al. bonneau 2006; Davidson et al. 2007; Gahl et al. 2011; Buck 2011). Yet like most organisms, amphibians encounter multiple et al. 2012). Among them, Davidson et al. (2007) demonstrated species of pathogens simultaneously or sequentially, resulting that the insecticide carbaryl could increase the susceptibility in coinfections that form ecological communities inside the host of metamorphosed amphibians to chytridiomycosis by inhibit- (Duellman and Trueb 1994; Pedersen and Fenton 2007; Telfer ing certain innate immune defenses. Given the immunosup- et al. 2010; Romansic et al. 2011). Within hosts, parasite species pressive potential of pesticides (Voccia et al. 1999), the effects may interact additively, antagonistically, or synergistically in of other pesticides on susceptibility to infection with Bd and manners that include mechanical facilitation (providing portals other pathogens individually and in combination warrant further Downloaded by [Department Of Fisheries] at 18:31 20 January 2013 of entry through injury to the host), interference competition investigation. (e.g., competing for space), food resource competition, and immunosuppression (Cusack and Cone 1986; Pedersen and Fenton 2007; Graham 2008; Beldomenico and Begon 2010; Telfer et al. CONCLUSION 2010). The result may be sudden and unpredictable changes in Amphibians are hosts to a variety of parasites, and in na- infection and disease dynamics (Pedersen and Fenton 2007; ture, they probably encounter multiple species of pathogens Telfer et al. 2010). Taking a more comprehensive approach to simultaneously (Duellman and Trueb 1994). In this study, we studies of pathogenic infections and disease in amphibians by found strong evidence that gyrodactylids have a high potential evaluating the net effects of coinfections may be fundamental to to be important pathogens of larval amphibians. Moreover, we species conservation (Pedersen and Fenton 2007; Beldomenico found a slight indication that experimental exposures of bull- and Begon 2010; Telfer et al. 2010; Romansic et al. 2011). frog tadpoles to Bd had a significant impact on survival among Batrachochytrium dendrobatidis and gyrodactylids have tadpoles that were already infected by G. jennyae, irrespective been observed to co-occur naturally within amphibian commu- of pesticide exposure. Exposure to Bd may have unanticipated nities (Green and Dodd 2007; Nieto et al. 2007). For instance, consequences for larval amphibians in the presence of other larval northern red-legged frogs Rana aurora aurora in Califor- pathogens, potentially putting their populations further at risk. GYRODACTYLUS AND CHYTRID FUNGUS POTENTIALLY LETHAL TO BULLFROG TADPOLES 25

ACKNOWLEDGMENTS the immune system of Xenopus laevis and Rana pipiens. Aquatic Toxicology This work was partially funded by Natural Sciences and En- 67:33–43. gineering Research Council of Canada Discovery Grant A6979 Crawshaw, G. J. 1997. Diseases in Canadian amphibian populations. Pages 258–270 in D. M. Green, editor. Amphibians in decline: Canadian studies awarded to J.D.M., by a Natural Sciences and Engineering Re- of a global problem. Society for the Study of Amphibians and Reptiles, St. search Council of Canada Postgraduate Scholarship awarded to Louis, Missouri. L.J.P., and by funds from the Pesticide Science Fund of Environ- Cusack, R., and D. K. Cone. 1986. A review of parasites as vectors of viral and ment Canada (to D.J.M. and B.D.P.). The authors thank Roger bacterial diseases of fish. Journal of Fish Diseases 9:169–171. Cue (McGill University, Montreal) for providing advice on sta- Daszak, P., L. Berger, A. A. Cunningham, A. D. Hyatt, D. E. Green, and R. Speare. 1999. Emerging infectious diseases and amphibian population tistical methods. We thank Ken Doe and Paula Jackman (En- declines. Emerging Infectious Diseases 5:735–748. vironment Canada) for providing the tadpoles, Joyce E. Long- Daszak, P., A. Strieby, A. A. Cunningham, J. E. Longcore, C. C. Brown, and core (University of Maine, Orono) for providing the fungal cul- D. Porter. 2004. Experimental evidence that the bullfrog (Rana catesbeiana) ture, along with Andree´ Gendron, Jean-François Lafond, and is a potential carrier of chytridiomycosis, an emerging fungal disease of Stephanie´ Arseneault for their assistance with the experiment. amphibians. Herpetological Journal 14:201–207. Davidson, C., M. F. Benard, H. B. Shaffer, J. M. Parker, C. O’Leary, J. M. Con- lon, and L. A. Rollins-Smith. 2007. Effects of chytrid and carbaryl exposure on survival, growth and skin peptide defenses in foothill yellow-legged frogs. REFERENCES Environmental Science and Technology 41:1771–1776. Ackerman, F. 2007. The economics of atrazine. International Journal of Occu- Duellman, W. E., and L. Trueb. 1994. Enemies and defense. Pages 241–260 in pational and Environmental Health 13:441–449. Biology of amphibians. Johns Hopkins University Press, Baltimore, Mary- Bakke, T. A., J. Cable, and P. D. Harris. 2007. The biology of gyrodactylid land. monogeneans: the “Russian-doll killers.” Advances in Parasitology 64:161– Du Preez, L. H., N. Kunene, G. J. Everson, J. A. Carr, J. P. Giesy, T. S. Gross, 376. A. J. Hosmer, R. J. Kendall, E. E. Smith, K. R. Solomon, and G. J. Van Der Beldomenico, P. M., and M. Begon. 2010. Disease spread, susceptibility and Kraak. 2008. Reproduction, larval growth, and reproductive development in infection intensity: vicious circles? Trends in Ecology and Evolution 25:21– African clawed frogs (Xenopus laevis) exposed to atrazine. Chemosphere 27. 71:546–552. Berger, L., R. Speare, P. Daszak, D. E. Green, A. A. Cunningham, C. L. Goggin, Dusek,ˇ L., M. Gelnar, and S.ˇ Sebelovˇ a´ 1998. Biodiversity of parasites in a R. Slocombe, M. A. Ragan, A. D. Hyatt, K. R. McDonald, H. B. Hines, K. R. freshwater environment with respect to pollution: metazoan parasites of chub Lips, G. Marantelli, and H. Parkes. 1998. Chytridiomycosis causes amphibian (Leuciscus cephalus L.) as a model for statistical evaluation. International mortality associated with population declines in the rain forests of Australia Journal for Parasitology 28:1555–1571. and Central America. Proceedings of the National Academy of Sciences of Fisher, M. C., T. W. J. Garner, and S. F. Walker. 2009. Global emergence of the USA 95:9031–9036. Batrachochytrium dendrobatidis and amphibian chytridiomycosis in space, Berger, L., R. Speare, and A. Kent. 2000. Diagnosis of chytridiomycosis in time, and host. Annual Review of Microbiology 63:291–310. amphibians by histologic examination. Zoos’ Print Journal 15:184–190. Forson, D. D., and A. Storfer. 2006. Atrazine increases ranavirus susceptibility in Blaustein, A. R., R. A. Alford, and R. N. Harris. 2009. The value of well- the tiger salamander, Ambystoma tigrinum. Ecological Applications 16:2325– designed experiments in studying diseases with special reference to amphib- 2332. ians. EcoHealth 6:373–377. Gahl, M. K., B. D. Pauli, and J. E. Houlahan. 2011. Effects of chytrid fungus Blaustein, A. R., J. M. Romansic, E. A. Scheessele, B. A. Han, A. P. Pessier, and and a glyphosate-based herbicide on survival and growth of wood frogs J. E. Longcore. 2005. Interspecific variation in susceptibility of frog tadpoles (Lithobates sylvaticus). Ecological Applications 21:2521–2529. to the pathogenic fungus Batrachochytrium dendrobatidis. Conservation Bi- Garner, T. W. J., M. W. Perkins, P. Govindarajulu, D. Seglie, S. Walker, A. A. ology 19:1460–1468. Cunningham, and M. C. Fisher. 2006. The emerging amphibian pathogen Blaustein, A. R., D. B. Wake, and W. P. Sousa. 1994. Amphibian declines: Batrachochytrium dendrobatidis globally infects introduced populations of judging stability, persistence, and susceptibility of populations to local and the North American bullfrog, Rana catesbeiana. Biology Letters 2:455–459. global extinctions. Conservation Biology 8:60–71. Garner, T. W. J., S. Walker, J. Bosch, S. Leech, J. M. Rowcliffe, A. A. Cun- Brodkin, M. A., H. Madhoun, M. Rameswaran, and I. Vatnick. 2007. Atrazine ningham, and M. C. Fisher. 2009. Life history tradeoffs influence mortality

Downloaded by [Department Of Fisheries] at 18:31 20 January 2013 is an immune disruptor in adult northern leopard frogs (Rana pipiens). Envi- associated with the amphibian pathogen Batrachochytrium dendrobatidis. ronmental Toxicology and Chemistry 26:80–84. Oikos 118:783–791. Buchmann, K., and T. Lindenstrøm. 2002. Interactions between monogenean Gheorgiu, C., D. J. Marcogliese, and M. Scott. 2006. Concentration-dependent parasites and their fish hosts. International Journal for Parasitology 32:309– effects of waterborne zinc on population dynamics of Gyrodactylus turn- 319. bulli (Monogenea) on isolated Guppies (Poecilia reticulata). Parasitology Buck, J. C., E. A. Scheessele, R. A. Relyea, and A. R. Blaustein. 2012. The 132:225–232. effects of multiple stressors on wetland communities: pesticides, pathogens Gibble, R. E., and K. N. Baer. 2011. Effects of atrazine, agricultural runoff, and competing amphibians. Freshwater Biology 57:61–73. and selected effluents on antimicrobial activity of skin peptides in Xenopus Busch, S., I. Dalsgaard, and K. Buchmann. 2003. Concomitant exposure of Rain- laevis. Ecotoxicology and Environmental Safety 74:593–599. bow Trout fry to Gyrodactylus derjavini and Flavobacterium psychrophilum: Gosner, K. L. 1960. A simplified table for staging anuran embryos and larvae effects on infection and mortality of host. Veterinary Parasitology 117:117– with notes on identification. Herpetologica 16:183–190. 122. Graham, A. L. 2008. Ecological rules governing helminth–microparasite coin- Charbonneau, M. 2006. Amphibian diseases: pesticide immunotoxicity and fection. Proceedings of the National Academy of Sciences of the USA chytridiomycosis in larval Rana catesbeiana and ranaviral disease in Rana 105:566–570. sylvatica tadpoles of central Ontario. Master’s thesis. Trent University, Peter- Green, D. E., and C. K. Dodd Jr. 2007. Presence of amphibian chytrid fungus Ba- borough, Ontario. trachochytrium dendrobatidis and other amphibian pathogens at warm-water Christin, M. S., L. Menard,´ A. D. Gendron, S. Ruby, D. Cyr, D. J. Marcogliese, fish hatcheries in southeastern North America. Herpetological Conservation L. Rollins-Smith, and M. Fournier. 2004. Effects of agricultural pesticides on and Biology 2:43–47. 26 PAETOW ET AL.

Kiesecker, J. M. 2002. Synergism between trematode infection and pesticide Peterson, J. D., M. B. Wood, W. A. Hopkins, J. M. Unrine, and M. T. Mendonça. exposure: a link to amphibian limb deformities in nature? Proceedings of the 2007. Prevalence of Batrachochytrium dendrobatidis in American bullfrog National Academy of Sciences of the USA 99:9900–9904. and southern leopard frog larvae from wetlands on the Savannah River site, LaFiandra, E. M., K. J. Babbitt, and S. A. Sower. 2008. Effects of atrazine South Carolina. Journal of Wildlife Diseases 43:450–460. on anuran development are altered by the presence of a nonlethal predator. Relyea, R., and J. Hoverman. 2006. Assessing the ecology in ecotoxicology: a Journal of Toxicology and Environmental Health 71:505–511. review and synthesis in freshwater systems. Ecology Letters 9:1157–1171. Larson, D. L., S. McDonald, A. J. Fivizzani, W. E. Newton, and S. J. Hamilton. Rohr, J. R., and B. D. Palmer. 2005. Aquatic herbicide exposure increases 1998. Effects of the herbicide atrazine on Ambystoma tigrinum metamorpho- salamander dessication risk eight months later in a terrestrial environment. sis: duration, larval growth, and hormonal response. Physiological Zoology Environmental Toxicology and Chemistry 24:1253–1258. 71:671–679. Rohr, J. R., T. R. Raffel, S. K. Sessions, and P. J. Hudson. 2008a. Understanding Longcore, J. E., A. P. Pessier, and D. K. Nichols. 1999. Batrachochytrium the net effects of pesticides on amphibian trematode infections. Ecological dendrobatidis gen. et sp. nov., a chytrid pathogenic to amphibians. Mycologia Applications 18:1743–1753. 91:219–227. Rohr, J. R., T. Sager, T. M. Sesterhenn, and B. D. Palmer. 2006. Exposure, pos- Marantelli, G., L. Berger, R. Speare, and L. Keegan. 2004. Distribution of texposure, and density-mediated effects of atrazine on amphibians: breaking the amphibian chytrid Batrachochytrium dendrobatidis and keratin during down net effects into their parts. Environmental Health Perspectives 114:46– tadpole development. Pacific Conservation Biology 10:173–179. 50. Marcogliese, D. J., and M. Pietrock. 2011. Combined effects of parasites and Rohr, J. R., A. M. Schotthoefer, T. R. Raffel, H. J. Carrick, N. Halstead, J. contaminants on animal health: parasites do matter. Trends in Parasitology T. Hoverman, C. M. Johnson, L. B. Johnson, C. Lieske, M. D. Piwoni, P. 27:123–130. K. Schoff, and V. R. Beasley. 2008b. Agrochemicals increase trematode Martins, M. L., C. A. Shoemaker, D. Xu, and P. H. Klesius. 2011. Effect of infections in a declining amphibian species. Nature 455:1235–1240. parasitism on vaccine efficacy against Streptococcus iniae in Nile Tilapia. Romansic, J. M., P. T. J. Johnson, C. L. Searle, J. E. Johnson, T. S. Tunstall, Aquaculture 314:18–23. B. A. Han, J. R. Rohr, and A. R. Blaustein. 2011. Individual and combined Mazzoni, R., A. A. Cunningham, P. Daszak, A. Apolo, E. Perdomo, and G. effects of multiple pathogens on Pacific treefrogs. Oecologia 166:1029–1041. Speranza. 2003. Emerging pathogen of wild amphibians in frogs (Rana Sapolsky, R. M. 2002. Neuroendocrinology of the stress-response. Pages 409– catesbeiana) farmed for international trade. Emerging Infectious Diseases 450 in J. B. Becker, S. M. Breedlove, D. Crews, and M. M. McCarthy, 9:995–998. editors. Behavioral endocrinology. Massachusetts Institute of Technology Moles, A., and T. L. Wade. 2001. Parasitism and phagocytic function among Press, Cambridge. Sand Lance Ammodytes hexapterus Pallas exposed to crude oil-laden Struger, J., D. Thompson, B. Staznik, P. Martin, T. McDaniel, and C. Mar- sediments. Bulletin of Environmental Contamination and Toxicology 66: vin. 2008. Occurrence of glyphosate in surface waters of southern Ontario. 528–535. Bulletin of Environmental Contamination and Toxicology 80:378–384. Nieto, N. C., M. A. Camann, J. E. Foley, and J. O. Reiss. 2007. Disease associated Stuart, S. N., J. S. Chanson, N. A. Cox, B. E. Young, A. S. L. Rodrigues, D. L. with integumentary and cloacal parasites in tadpoles of northern red-legged Fischman, and R. W. Waller. 2004. Status and trends of amphibian declines frog Rana aurora aurora. Diseases of Aquatic Organisms 78:61–71. and extinctions worldwide. Science 306:1783–1786. Oka, T., O. Tooi, N. Mitsui, M. Miyahara, Y. Ohnishi, M. Takase, A. Kashiwagi, Telfer, S., X. Lambin, R. Birtles, P. Beldomenico, S. Burthe, S. Paterson, and T. Shinkai, N. Santo, and T. Iguchi. 2008. Effect of atrazine on metamorpho- M. Begon. 2010. Species interactions in a parasite community drive infection sis and sexual differentiation in Xenopus laevis. Aquatic Toxicology 87: risk in a wildlife population. Science 330:243–246. 215–226. Venesky, M. D., R. J. Wassersug, and M. J. Parris. 2010. Fungal pathogen Ottesen, O. H., and A. B. Amin. 2011. Mortality and cellular response in the changes the feeding kinematics of larval anurans. Journal of Parasitology skin and gills of plaice (Pleuronectes platessa L.) to parasite and bacteria 96:552–557. infection. Bulletin of the European Association of Fish Pathologists 31:16– Voccia, I., B. Blakley, P. Brousseau, and M. Fournier. 1999. Immunotoxicity of 22. pesticides: a review. Toxicology and Industrial Health 15:119–132. Paetow, L., D. K. Cone, T. Huyse, J. D. McLaughlin, and D. J. Marcogliese. Wootton, D. M., K. A. Ryan, R. S. Demaree, and R. L. Critchfield. 1993. A 2009. Morphology and molecular taxonomy of Gyrodactylus jennyae n. sp. new species of Gyrodactylus (Monogenea: Monopisthocotylea) on tadpoles (Monogenea) from tadpoles of captive Rana catesbeiana Shaw (Anura), with of Rana catesbeiana from California, U.S.A. Transactions of the American a review of the species of Gyrodactylus Nordmann, 1832 parasitising am- Microscopical Society 112:230–233. phibians. Systematic Parasitology 73:219–227. Xu, D. H., C. A. Shoemaker, and P. H. Klesius. 2007. Evaluation of the link

Downloaded by [Department Of Fisheries] at 18:31 20 January 2013 Pedersen, A. B., and A. Fenton. 2007. Emphasizing the ecology in parasite between gyrodactylosis and streptococcosis of Nile Tilapia, Oreochromis community ecology. Trends in Ecology and Evolution 22:133–139. niloticus (L.). Journal of Fish Diseases 30:233–238. This article was downloaded by: [Department Of Fisheries] On: 20 January 2013, At: 18:31 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Aquatic Animal Health Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/uahh20 Viral Surveillance of Cultured Rainbow Trout in the Eastern Black Sea, Turkey H. Ogut a , C. Altuntas a & R. Parlak a a Department of Fisheries Technology Engineering, Faculty of Marine Sciences, Karadeniz Technical University, 61530 Camburnu-Surmene, Trabzon, Turkey Version of record first published: 04 Jan 2013.

To cite this article: H. Ogut , C. Altuntas & R. Parlak (2013): Viral Surveillance of Cultured Rainbow Trout in the Eastern Black Sea, Turkey, Journal of Aquatic Animal Health, 25:1, 27-35 To link to this article: http://dx.doi.org/10.1080/08997659.2012.732652

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material. Journal of Aquatic Animal Health 25:27–35, 2013 C American Fisheries Society 2013 ISSN: 0899-7659 print / 1548-8667 online DOI: 10.1080/08997659.2012.732652

ARTICLE

Viral Surveillance of Cultured Rainbow Trout in the Eastern Black Sea, Turkey

H. Ogut,* C. Altuntas, and R. Parlak Department of Fisheries Technology Engineering, Faculty of Marine Sciences, Karadeniz Technical University, 61530 Camburnu-Surmene, Trabzon, Turkey

Abstract To study the presence and spread of viral fish pathogens in Rainbow Trout Oncorhynchus mykiss in the eastern Black Sea region of Turkey, 172 pooled samples (seven fish per pool) of Rainbow Trout fry from 28 hatcheries were examined from December 2006 to July 2007. Forty-three pools (seven broodfish per pool) of seminal and ovarian fluids from 182 female and 119 male brood Rainbow Trout were also sampled during spawning. Moreover, reproductive fluids (22 pools) of wild trout (Salmo trutta labrax, S. t. caspius, S. t. abanticus,andS. t. macrostigma), captured by electroshocking in the rivers in and around the region, were sampled. Triplicate groups of 40 or 80 Rainbow Trout fry was also challenged with two similar isolates to determine their virulence on trout fry. Enzyme-linked immunosorbent assay was performed on the samples producing cytopathic effect on CHSE-214 cells. The positive results were confirmed with a reverse transcriptase polymerase chain reaction assay. Neither infectious hematopoietic necrosis virus (IHNV) nor viral hemorrhagic septicemia virus (VHSV) was detected during the survey. Of the 28 hatcheries sampled in the Black Sea region, 15 from six provinces tested positive for infectious pancreatic necrosis virus (IPNV) in trout fry. Unexpectedly, all reproductive fluids from both male and female cultured and wild broodstock tested negative for IPNV. Nucleotide sequences of the VP2/NS region of IPNV showed that all isolates collected (n = 38) in the region and surrounding areas belonged to the genogroup III. The findings strongly suggest that IPNV is endemic in the fry of farmed Rainbow Trout within the region. Virus prevention measures should be taken to prevent in-farm spread of these highly contagious, low-virulence isolates.

Viral diseases of ﬁsh—speciﬁcally infectious pancreatic have a shorter aquatic life (Pietsch et al. 1977; Ahne 1982; necrosis virus (IPNV), infectious hematopoietic necrosis virus Hawley and Garver 2008), are primarily transferred horizon- (IHNV), and viral hemorrhagic septicemia virus (VHSV)—have tally, and can cause mortality up to 100% in Rainbow Trout

Downloaded by [Department Of Fisheries] at 18:31 20 January 2013 major impacts on both salmonids and nonsalmonids. Rainbow (Wolf 1988). Owing to huge losses caused by these two rhab- Trout Oncorhynchus mykiss is a highly susceptible host to all doviruses (Hill 1992), eradication programs are put into action three of these viruses (Jørgensen 1982; Wolf 1988; Bootland whenever they are detected, not only in Turkey but throughout and Leong 1999; Reno 1999; Smail 1999; Ogut and Reno 2004, Europe (OIE 2006). 2005). One of these three, IPNV, which belongs to the aquatic Viral diseases of cultured and wild fish have largely been family Birnaviridae, can be transmitted vertically and horizon- ignored in Turkey. Candan (2002) reported the first incidence tally, and survives in the aquatic environment for a long time of IPNV in Rainbow Trout fry. Later, the transfer of IPNV (Toranzo and Metricic 1982; Smail et al. 1993), making it an ex- from hatcheries to the Rainbow Trout smolts and its fate in the tremely contagious and persistent aquatic virus. Therefore, once Black Sea were investigated (Ogut and Altuntas 2012). Viral established in an ecosystem, IPNV spreads in an extremely ef- hemorrhagic septicemia virus, in Turkey, was first reported in ficient manner and its eradication is difficult. However, both Turbot Psetta maxima (Nishizawa et al. 2006) and later in Euro- IHNV and VHSV, which belong to the family Rhabdoviridae, pean Whiting Merlangius merlangus (Altuntas and Ogut 2010).

*Corresponding author: [email protected] Received May 8, 2012; accepted September 16, 2012 27 28 OGUT ET AL.

Infectious hematopoietic necrosis virus has not been reported 16◦C. Brood reproductive fluid samples were collected at 6–7◦C in Turkey yet. However, there are no epidemiological data eval- during the major spawning period in the region. uating the extent of viral fish pathogen spread and the factors Juvenile trout 1–11 months old (preferably 1–2 months old) affecting this spread in the region where Rainbow Trout culture and alive were sampled randomly as a single pool of seven fish is widespread. Therefore, new introductions of viral pathogens per pond, from each of five ponds in each farm. The ponds by uncontrolled fish transfers in and out of the region go largely having problems were especially preferred to include into the unnoticed. Such an uncontrolled setting would lead to uncon- sample to promote the chance of IPNV detection. Similarly, trolled and unintentional spread of viral fish pathogens into reproductive fluid samples of broodfish from the four biggest cultured and wild fish stocks. regional hatcheries were also pooled as reproductive fluids of In this study, we report the occurrence and spread of IPNV, five to seven broodfish per sample. Ovarian and seminal fluids of VHSV, and IHNV among the trout hatcheries in the eastern various wild trout species from various rivers close to the region Black Sea region. Though the occurrence and prevalence of were also tested for the presence of any viral fish pathogen IPNV were assessed only in the Black Sea region, it also pro- (Table 1; Figure 1). vides insight for other regions because fish transfers between Wild trout were collected opportunistically in and around the regions were and are highly common. Genogroup diversity and region by electroshocking and were subjected for viral testing spread throughout Turkey is reported first in this report. (Figure 1; Table 1). Each species was collected from a single and separate river. Only targeted species, listed in Table 1, were collected. Brood wild trout species were brought to the laboratory METHODS and kept there for up to 20 d to collect eggs or milt. Rainbow Samples of cultured Rainbow Trout fry were collected from Trout was not caught in the samplings from these rivers. 28 trout hatcheries in six provinces in the Black Sea region— One hatchery, the biggest hatchery in the region, containing Artvin, Rize, Trabzon, Giresun, Ordu, and Samsun—to deter- infected Rainbow Trout larvae in the Maçka River system in mine the presence of IPNV, VHSV, and IHNV (Figure 1). Re- Trabzon, was visited twice a month to assess the persistence productive fluids from broodfish belonging to the four biggest of the infection during larval-rearing season from December hatcheries that are more dynamic about fish transfers inside or 2006 to June 2007. All the ponds (n = 13) after first feed- outside the region were collected from December 2006 to June ing were monitored for viral pathogens. A remote hatchery in 2007, except when water temperatures were below 7◦C or above the Gumushane province, which is assumed to be free of any Downloaded by [Department Of Fisheries] at 18:31 20 January 2013

FIGURE 1. Map of the provinces surveyed. Samples of Rainbow Trout fry (ﬁlled squares) were collected from Samsun, Ordu, Giresun, Trabzon, Rize, Artvin, Gumushane, Sivas, and Mugla.˘ Wild brood trout samples (asterisks) were collected from Bolu, Kars, Gumushane, and Trabzon. VIRAL PATHOGEN SURVEY IN TURKEY 29

TABLE 1. Screening of seminal and ovarian ﬂuids of cultured Rainbow Trout broodstock (hatcheries A, B, C, and D) in the hatcheries in the Black Sea region of Turkey and wild trout species collected by electroshocking. Wild ﬁsh broods were brought to the laboratory alive and kept there until spawning (15–20 d).

Number of pools Trout species Provinces Temperature (◦C) Male Female Age (years) Brood ﬁsh per pool IPNV Rainbow Trout Hatchery A Trabzon 7.1 6 6 2–3 7 No Hatchery B Trabzon 7.4 5 9 2–4 7 No Hatchery C Gumushane 7.1 3 3 1–5 7 No Hatchery D Rize 6.1 3 8 2–3 7 No S. t. labrax Gumushane 11.2a 23 2 5 No S. t. caspius Kars 11.2a 21 2 7 No S. t. abanticus Bolu 11.2a 4 4 1–2 7 No S. t. macrostigma Trabzon 11.2a 1 2 1–2 6 No

aThe temperature of spring water used in the laboratory.

important pathogens including viruses, was also sampled in the O. tshawytscha embryo cell line CHSE-214 (Lannan et al. 1984) spawning season of 2007. This hatchery is important in that it and of epithelioma papulosum cyprini (EPC) cell lines (Fijan is often used as a broodstock source for the hatcheries in the re- et al. 1983) on 24-well plates at dilutions of 1:10 and 1:100 in gion. Moreover, an IPNV isolate from a sample submitted by a duplicates. After 1.5-h incubation at 15◦C, the inoculum was province, Sivas, in the Eastern Anatolia region, and another iso- decanted and 1 mL EMEM supplemented with 2% fetal bovine late obtained from a sample submitted by the biggest Rainbow serum and antibiotics (100 IU penicillin/mL, 100 µg strepto- Trout hatchery in the Mugla˘ province in the western part of mycin/mL, and 0.25 µg amphotericin/mL) including HEPES Turkey were sequenced to use in the comparison of sequences buffer (25 mM) was added to each well. The cultures were then from different regions in Turkey. incubated at 15◦C for 14 d. Regardless of the observation of Each fish sampled was examined for external disease signs cytopathic effect (CPE) in the first inoculation onto cell culture, or other abnormalities. The total lengths of the fish and their the inoculum was reinoculated as described above to (1) con- ages (in months) were also recorded. The samples were assayed firm that the observed CPE was not the result of toxicity, and on the same day, or the day after they were collected, at the (2) increase the probability of virus isolation. Fish Diseases Laboratory of the Faculty of Marine Sciences, All samples causing CPE on the cell lines were subsequently Trabzon, Turkey. tested by enzyme-linked immunosorbent assay (ELISA) for the The largest Rainbow Trout hatchery in the Maçka River sys- presence of IPNV or VHSV using a commercial kit according to tem was also sampled twice a month during the spawning season the manufacturer’s instructions (Test-Line Clinical Diagnostics, from January to June 2007. In the farm, bacterial and parasit- Brno, Czech Republic). Aliquots (50 µL) from CPE-positive ical examinations were also carried out to prevent parasite- or wells, centrifuged at 10,000 × g at 4◦C for 20 min, were bacteria-related mortality. Twelve spawning lots were monitored added to microtiter wells precoated with purified polyclonal bimonthly for viral pathogens. anti-VHSV or anti-IPNV antibody with positive and negative Some physical characteristics of the water (temperature, oxy- controls supplied with the kit. The wells were first incubated Downloaded by [Department Of Fisheries] at 18:31 20 January 2013 gen, pH, salinity, and total suspended solids) were measured us- with 50 µL of rabbit anti-VHSV antibody at 37◦C for 60 min, ing a handheld multiparameter probe (YSI 556; Yellow Springs followed by incubation with peroxidase-labeled antibody to rab- Instruments, Yellow Springs, Ohio) during sample collection bit immunoglobulin at 37◦C for 60 min. The final incubation trips. with a chromogenic substrate, tetramethylbenzidine, took place Virus isolation with fish cell cultures.—The whole fish was at 20◦C for 10 min. The results of the assay were recorded as homogenized if younger than 2–3 months old. The viscera (all positive or negative depending on the presence of strong color interior organs) and head were also used if the fish was 3– formation. 5 months old. If the fish was older than 5 months, pieces of Nucleotide analysis.—Of 38 IPNV isolates collected during liver, kidney, and spleen were used. The pooled tissue sam- the survey, 27 were subjected to nucleotide sequencing. Total ples and reproductive fluid samples from broodfish collected in RNA, isolated from infected cells using the Mini RNeasy Kit minimum essential medium with Earle’s salts (EMEM), and an- (Qiagen) according to the manufacturer’s instructions, was used tibiotics (100 IU penicillin/mL, 100 µg streptomycin/mL, and as a template for reverse transcriptase polymerase chain reac- 0.25 µg amphotericin/mL) were homogenized and centrifuged tion (RT-PCR) to amplify 210 bp covering the VP2/NS junction for 20 min at 4,000 × g at 4◦C. Then, 100 µL of aliquots were coding region of IPNV (1,403–1,761 bp), based on the genome inoculated onto 90% confluent monolayers of Chinook Salmon (1,403–1,761 bp) of the Jasper strain, GenBank accession 30 OGUT ET AL.

number M18049 (Nishizawa et al. 2005), using the primers TABLE 2. Results of viral survey of cultured Rainbow Trout fry and juveniles ABV-P1(5-AGAGATCACTGACTTCACAAGTGAC-3) and in the hatcheries in the Black Sea region. ABV-P2 (5 -TGTGCACCACAGGAAAGATGACTC-3 ), in a Total number Number of Positive single-step RT-PCR system (One Tube RT-PCR, Roche). The of fish fish per sample Farm following RT-PCR program was used on a thermocycler (MJ Re- Province sampled pool prevalencea prevalenceb search, BioRad): for reverse transcription, reactions were held at 50◦C for 30 min and at 70◦C for 15 min; denaturation took Artvin 105 7 1/15 1/3 place at 94◦C for 10 s, annealing at 54◦C for 30 s, and extension Rize 109 5–7 4/17 2/7 at 62◦C for 2 min for 35 cycles. The PCR products were vi- Trabzon 798 5–7 32/114 9/13 sualized in agarose gel electrophoresis using ethidium bromide Gumushane 37 5–7 1/6 1/1 and UV light. The PCR products were purified and sequenced Giresun 105 7 2/15 1/3 in duplicate by ABI 3730XL DNA sequencer (DNA Service; Ordu 35 7 1/5 1/1 Macrogen). Total 1,189 41/172 15/28 Sequence analysis.—Database analyses of the sequences aPositive sample prevalence = number of IPNV positive pools/number of total pools were accomplished using BLASTN procedures. For sequence examined. comparisons of the VP2/NS junction coding region of IPNV bFarm prevalence = number of IPNV-positive Rainbow Trout farms/total number of isolates to determine the degree of homology, the multiple- Rainbow Trout farms sampled. alignment algorithm, ClustalW, in MEGA5 (Thompson et al. 1994; Tamura et al. 2011) was employed. Phylogenetic anal- ELISA and RT-PCR, and IPNV was isolated from all provinces ysis was performed on our 11 IPNV isolates and the results surveyed (Table 2). More than half the hatcheries (15 out of 28) compared with the 55 reference strains retrieved from GenBank sampled from six provinces tested positive for IPNV. Clinical (Figure 2). Phylogenetic trees were constructed by the neighbor- infectious pancreatic necrosis (IPN) with symptoms of disease joining method (Saitou and Nei 1987) using MEGA version 2.1 was not observed in Rainbow Trout fry, brood, or wild fish (Kumar et al. 2001). The neighbor-joining tree and its topol- broods during the survey, and infected fish were asymptomatic ogy were evaluated by bootstrap analyses (Felsenstein 1985) carriers of IPNV bearing less than 102.3 TCID50 virus/mL. The based on a thousand resamplings. The evolutionary distances in-hatchery prevalence of IPNV in some of the hatcheries that were then computed using the maximum composite likelihood were surveyed was also high. In three cases, all five ponds sam- method of Tamura et al. (2004) and are in the units of the number pled tested for IPNV and usually at least one pond tested positive of base substitutions per site. for IPNV. The number of IPNV-positive ponds increased when Challenge experiment.—Rainbow Trout fry produced from the temperature rose from 7◦Cto15◦C, then sharply decreased. sperm and eggs obtained from four IPNV-negative males and In the determination of persistence of IPNV, the bimonthly females were used in the challenge experiments right after first survey of the largest farm revealed that, unexpectedly, only one feeding at the Fish Diseases Laboratory, Karadeniz Technical pond of trout fry from the first broodstock spawners tested pos- University. The fry were virologically examined for viruses (20 itive for IPNV. The same pond stayed persistently positive for fish, 5 fish per pool) prior to infection. Triplicate groups of IPNV up to 6 months without any mortality. During the study pe- 40 or 80 trout fry were experimentally challenged with IPNV riod, the temperature of the hatchery water ranged from 10.4◦C by immersion with a dose 104.8 TCID50 of IPNV (Uzungol to 12.6◦C. and Cos3 isolates) for 5 h in 1-L aerated water. At the end of The age of the larvae, the time between hatching and sam- exposure, water flow was reestablished and volume was restored pling, was also recorded. Infectious pancreatic necrosis virus Downloaded by [Department Of Fisheries] at 18:31 20 January 2013 to 3.5 L. Water temperature for the duration of the experiment was isolated from Rainbow Trout larvae up to the age of 8 was maintained at 15◦C with aeration. Fish in four control tanks months (Figure 3). The sampling was mainly targeted at 2- were challenged with culture medium only. Each fish that died month-old fish since this is the age at which IPN disease is most was recorded and individually processed for viruses every day likely to be observed (Wolf and Quimby 1971). In this age- for 16 d after virus exposure. At the end of the experiment, nine group, one out of four pools tested positive for IPNV. Higher fish from each tank were individually sampled to determine ratios of positive pools were obtained from 2.5-month-old trout in-tank prevalence of IPNV. fry (Figure 3). Ovarian and seminal fluids of brood Rainbow Trout (female = 182, male = 119) and wild trout (n = 19) broods RESULTS were also examined during the spawning period, December One hundred and seventy-two pooled samples (five to seven 2006 to June 2007 (Table 1). All Rainbow Trout broods (male fish per pool) of Rainbow Trout fry were examined for viral and females) and wild fish broods were free of all three viruses pathogens (Table 2). Neither IHNV nor VHSV was detected in surveyed. Spawning adult fish samples collected below 8◦C cultured Rainbow Trout during the survey. All CPE (n = 38) tested negative for IPNV. Moreover, seminal and ovarian fluids observed in CHSE-214 cells were IPNV related, confirmed with collected from various wild trout species—Salmo trutta labrax VIRAL PATHOGEN SURVEY IN TURKEY 31 Downloaded by [Department Of Fisheries] at 18:31 20 January 2013

FIGURE 2. A phylogenetic tree based on the VP2/NS junction coding region sequences of 55 reference strans and of our 11 isolates (denoted with the starting letter “T”) sequenced in this study. The evolutionary history was inferred using the neighbor-joining method (Saitou and Nei 1987). The evolutionary distances were computed using the maximum composite likelihood method (Tamura et al. 2004) and are in the units of the number of base substitutions per site. The bar indicates 5% estimated sequence divergence. 32 OGUT ET AL.

also tested positive for IPNV. Repeated samples were not taken in these provinces after isolation of IPNV in the most probable season.

Sequence Analysis We amplified viral RNA by PCR and sequenced 210 bp fragments of cDNA corresponding to the VP2/NS junction coding region of IPNV isolates from provinces (A1Dere Trabzon; Cos3 Trabzon; Uzungol Trabzon; Ser2007 Trabzon, Ser0608 Trabzon, BC Trabzon, AltH3 Trabzon, Sivas Sivas, DM5 Rize, DE Rize, K20 Mugla,˘ and Gumus Gumushane). To verify that the errors were not responsible for the observed nucleotide differences during Taq DNA polymerase amplification, we sequenced two independent amplifications from each virus isolate. The phylogenetic analyses showed that all IPNV isolates collected during the survey belonged to genogroup III and were FIGURE 3. Frequency of IPNV isolations from different age-groups of similar to each other, even though they were isolated from dif- Rainbow Trout larvae–fry. Each data point comes from test results of a sample ferent provinces (Figures 2, 4). The GenBank comparisons using (pool of five to seven fish); N = negative, P = positive. BLASTN showed that IPNV isolates were highly similar (≥98) to the Sp-type genogroup III isolates from various countries, (n = 35), S. t. caspius (n = 21), S. t. abanticus (n = 56), and especially Croatia (Figure 5). S. t. macrostigma (n = 18)—also tested negative for viral fish pathogens. Challenge Experiment The fry in the remote hatchery in the Gumushane province To determine the virulence of two common isolates (Uzungol used for broodstock supply for the hatcheries in the region. All and Cos3), a challenge experiment with Rainbow Trout fry that broodfish in this hatchery tested negative for viruses. Two out of just started active feeding after yolk sac absorption was con- five pools (from five ponds) of fry from the hatchery in Mugla˘ ducted. Infectious pancreatic necrosis virus was not detected province, and three pools out of five pools from Sivas province either in mortal fish during the experiment or survivors in the Downloaded by [Department Of Fisheries] at 18:31 20 January 2013

FIGURE 4. Cumulative IPNV-positive mortality and virus titers. Rainbow Trout fry were challenged with 104.8 TCID50 mL of Uzungol or Cos3 isolate for 5 h. Mortalities during the experiment were processed individually and tested for IPNV on cell culture. Each data point for virus titers was obtained from the mean of virus titers from infected mortalities at a given day. VIRAL PATHOGEN SURVEY IN TURKEY 33

FIGURE 5. The amino acid sequences of the VP2/NS junction coding region of IPNV aligned with 11 Turkish IPNV isolates: T K20, T AltH3, T BC, T Cos3, T DM5, T Gumus, T Ser2007, T Ser0806, T Sivas, and T Uzungol. The accession numbers of the earlier published sequences used in the comparison are as follows: Croatian/06 (EU219618.1), a Spanish isolate, 31–75 (AJ622822), and a French isolate (AJ489222).

control tanks. Both isolates caused low levels of virulence sup- 1990. Intensive fish transfer between the two regions could be ported by high virus titers from the seventh to the 13th day the main reason for the phenomenon. postexposure at two densities (Figure 4). Mortality levels were A high rate of IPNV prevalence in and among Rainbow Trout similar except for one low-density group with Cos3 isolate. The hatcheries could happen for many different reasons considering level of mortality in this low-density group was significantly the highly contagious nature of IPNV (Reno 1998). The rate of higher than that observed in other groups (P < 0.005). Mean in-farm IPNV prevalence (i.e., the rate of IPNV-positive juvenile virus titers were the highest on the ninth day postexposure. ponds) in IPNV-positive hatcheries was 38.59% (± 36.74%; ratio of positive pools to ponds in positive hatcheries), suggesting a high rate of in-hatchery vertical and horizontal transmission DISCUSSION without virus awareness and virus spread prevention protocols. Our results revealed that IPNV is endemic in Rainbow Trout Jarp (1999) reported that only 9% of egg yolk fry tested posi- cultured in the eastern Black Sea region of Turkey. More than tive for IPNV and later prevalence in juvenile and smolt stage half of the hatcheries in the region had IPNV belonging to increased. Furthermore, the largest hatchery was monitored for genogroup III. In the absence of biosecurity protocols designed viral fish pathogens during the spawning season (December– to prevent virus introduction and spread, IPNV is probably en- June). The Rainbow Trout fry in these hatcheries were contam- demic in the farmed Rainbow Trout in the rest of Turkey as inated with IPNV in the past 3 years, and quarantine procedures well. Neither IHNV nor VHSV was detected in the survey of were implemented by the Agricultural Ministry in 2005 as a cultured Rainbow Trout. Therefore, IPNV will be the focus of result. Interestingly, in our study only one pond, containing fish Downloaded by [Department Of Fisheries] at 18:31 20 January 2013 the remaining discussion herein. from the first spawners, tested positive for IPNV and remained A comparison of the dendogram of the VP2/NS junction positive for 6 months without any mortalities. The same pond coding region of IPNV clearly shows the close relationship be- was specifically screened twice a month to assess other infec- tween the IPNV isolates from provinces, and they all belonged tions (parasitical or bacterial) that could cause mortalities. The to the genogroup III. Comparative analysis suggested that the farmer kept the IPNV-positive pond isolated from the others isolates Ser2007, Ser0806, A1Dere, Sivas, Alth3, BC, Uzungol, to prevent pond-to-pond contamination. Unexpectedly, all the and Cos3 were identical. Ser0806 and Ser2007, isolated from other ponds tested negative for IPNV for the rest of the larval- the same farm in 2006 and 2007, were also identical; K20 and rearing season. This observation strongly suggests that the virus Gumus, which are identical to each other, are different from had been spread to other ponds by handling practices in the the aforementioned group by only one to three substitutions previous years. (Figure 5). The IPNV isolate in Gumushane, the brood provider Hatchery-to-hatchery transmission of IPNV usually takes in the Black Sea region, was identical to the isolate in Mugla˘ place through the exchange of infected eggs, fish, contaminated province. The IPNV isolates from these two provinces were water, or other material used in a hatchery (Bullock et al. 1976; expected to be the most different because the hatchery in Wolf et al. 1968; McAllister and Bebak 1997). The disinfection Gumushane province had kept its original broodstock since of the eggs does not totally control IPNV since the virus can 34 OGUT ET AL.

be transferred vertically by residing inside eggs (Bullock et al. first spawning lot of broods had not been screened for IPNV. 1976). The hatcheries on the Maçka River, which are the four Lack of IPNV-positive fry from early spawners for 6 months largest hatcheries in the Trabzon province and are all within a after spawning further suggests that the early spawners were kilometer of one another, have obtained their broodstock from a the source of IPNV. According to the statements made by the hatchery in the Gumushane province that had never been tested farmer, the early spawners were imported from South Africa as for any viruses. This hatchery has never allowed in fish or eggs eggs in 1996. These fish could be more susceptible to the virus but has continued its operations using the same broodstock ac- than late spawners, or higher temperatures during early spawn- quired in late 1980s. Apparently, IPNV isolates obtained in the ing could enhance the occurrence or replication of IPNV. More Trabzon province were genetically the same, genogroup III, as research is needed to corroborate these findings before using that in Gumushane province. Discharged water from each hatch- them in management practices. ery is used by the hatcheries downstream in the Macka River Clinical IPNV was not observed during the survey even in system, just as is in the other regions producing Rainbow Trout the hatcheries where all five ponds tested positive for IPNV. seed (e.g., Mugla). This is also an efficient method of spreading Infectious pancreatic necrosis virus-related mortality by a viru- IPNV in the system since IPNV is viable for a long time in lent isolate would occur in the first 4 months (Frantsi and Savan freshwater (McAllister and Bebak 1997). 1971; Wolf and Quimby 1971) or during the first 1,500 degree- Fish transfers in and out of the region are not screened for days (Dorson and Touchy 1981). No mortality was anticipated viral fish pathogens nor have they been screened in the past. An during monitoring of infected larvae for 6 months. However, individual fish may stay at many places during its life, owing the challenge experiments showed that there were low IPN- to the type and nature of Rainbow Trout culture in and around related mortality in infected fry on the ninth day of challenge the region. For example, eggs are obtained from one place (a accompanied with high IPNV titers. Moreover, on the 16th day hatchery), the larvae grows into a juvenile at another place (a of exposure, prevalence among survivors ranged from 89% to hatchery), and the period from juvenile to 100–200-g growth 100% with low titers (104 TCID50/mL). It is important to note takes place in cages at a lake or a dam. The final grow-out that from the 16th day, any fish dying of different diseases tested phase—to 1,500 g—is carried out in cages in the Black Sea. positive for IPNV just as in the 6-month survey of the biggest Therefore, infected fish could potentially spread IPNV to three hatchery. Candan (2002) reported massive IPN-related mortal- other locations. To control the spread of IPNV, in-hatchery virus ity (up to 50%) in Rainbow Trout between 1990 and 2002. No surveillance and disease prevention protocols should be applied specific data about prevalence or titer of infected fish was pro- and all transferred fish should be checked for IPNV. Our se- vided to support the idea that the observed mortality in the farms quence analysis also indicates that the isolates obtained in the is indeed IPN related. There may have been a mixed infection biggest hatcheries located in the region (A1Dere, Uzungol, Cos) with other pathogens (e.g., pathogenic myxobacteria to which and in trout egg and juvenile providers in other parts of Turkey most mortalities in larval rearing are attributed; H. Ogut, un- (Gumus, K20) are identical, further pinpointing endemicity and published data) or parasitical diseases endemic to Turkey (Ogut uncontrolled and highly dynamic fish transfers. et al. 2005; Ogut and Akyol 2005, 2007). Infectious pancreatic necrosis virus replication is limited be- In conclusion, IPNV is endemic in the cultured Rainbow low 8◦C and suppressed around 16◦C (Dorson and Touchy 1981; Trout in the Black Sea region. The phylogenetic analysis re- Bergan and Robertsen 2004; Ogut and Altuntas 2012). Most vealed that the collected IPNV isolates in the region belong spawning in the eastern Black Sea region usually takes place to the genogroup III, and they are substantially homologous at water temperatures around 8◦C. One could argue that IPNV throughout Turkey. There is an urgent need to epidemiologically was present in the broodfish but would have been suppressed determine the factors leading to such an efficient spreading of Downloaded by [Department Of Fisheries] at 18:31 20 January 2013 below 8◦C to such an extent that it could not be detected by our the virus in the hatcheries. Although the isolates spread in the screening procedure. It would be expected that fry from brood- region have low virulence for Rainbow Trout fry, the virulence stock that tested negative for IPNV be positive for IPNV when on other cultured and wild trout species and marine species water temperatures were optimal. However, negative fry stayed should be determined. negative for about 6 months after spawning. The source of IPNV deserves more in-depth evaluation. The broodfish that tested negative (late spawners) could be resistant ACKNOWLEDGMENTS to IPNV (Okamoto et al. 1993), or more probably virus replica- This project was supported with the funds from the State tion was suppressed due to lower water temperatures (Dorson Planning Organization (2003K120750). We also thank an and Touchy 1981). The unexpected finding that broodfish in anonymous reviewer for reviewing the manuscript. the biggest spawning lot were negative for IPNV directed us to follow the seeds of these fish in the spring water to determine whether these fish will turn out positive later. However, again REFERENCES unexpectedly, all fry and juveniles stayed negative for the first Ahne, W. 1982. Untersuchungen zur Tenazitat¨ der Fischviren. [Investigations 6 months. A single pond of Rainbow Trout juveniles from the on the tenacity of fish viruses.] Fortschritte der Veterinarmedizin¨ 35:305–309. VIRAL PATHOGEN SURVEY IN TURKEY 35

Altuntas, C., and H. Ogut. 2010. Monthly occurrence and prevalence of viral Ogut, H., and A. Akyol. 2007. Prevalence and intensity of ectoparasites in haemorrhagic septicaemia virus (VHSV) in Whiting Merlangius merlangus. Rainbow Trout (Oncorhynchus mykiss) from larvae stage to market size in Diseases of Aquatic Organisms 88:107–113. Turkey. Israeli Journal of Aquaculture Bamidgeh 59:23–31. Bergan, V., and B. Robertsen. 2004. Characterization of Atlantic Halibut (Hip- Ogut, H., A. Akyol, and M. Z. Alkan. 2005. Seasonality of Ichthyophthirius poglossus hippoglossus) Mx protein expression. Developmental and Com- multifiliis in the trout (Onchorhynchus mykiss) farms of the eastern Black Sea parative Immunology 28:1037–1047. region of Turkey. Turkish Journal of Fisheries and Aquatic Sciences 5:23– Bootland, L. M., and J. A. C. Leong. 1999. Infectious haematopoietic necrosis 27. virus. Pages 57–121 in P. T. K. Woo and D. W. Bruno, editors. Fish diseases Ogut, H., and C. Altuntas. 2012. Occurrence and prevalence of infectious pan- and disorders. CAB International, Oxfordshire, UK. creatic necrosis virus in Rainbow Trout (Oncorhynchus mykiss) cultured in Bullock, G. L., R. R. Rucker, D. Amend, K. Wold, and H. M. Stuckey. 1976. In- cages in the Black Sea. Aquaculture Research 43:1550–1556. fectious pancreatic necrosis: transmission with iodine-treated and nontreated Ogut, H., and P. W. Reno. 2004. Early kinetics of infectious hematopoietic eggs of Brook Trout (Salvelinus fontinalis). Journal of the Fisheries Research necrosis virus (IHNV) infection in Rainbow Trout. Journal of Aquatic Animal Board of Canada 33:1197–1198. Health 16:152–160. Candan, A. 2002. First report on the diagnosis of infectious pancreatic necrosis Ogut, H., and P. W. Reno. 2005. In vitro host range of aquatic birnaviruses. (IPN) based on reverse transcription polymerase chain reaction (RT-PCR) Bulletin of the European Association of Fish Pathologists 25:53–63. in Turkey. Bulletin of the European Association of Fish Pathologists 22: OIE (World Organisation for Animal Health). 2006. Diagnostic manual for 45–48. aquatic animal diseases, 5th edition. OIE, Paris. Dorson, M., and C. Touchy. 1981. The influence of fish age and water temper- Okamoto, N., T. Tayama, M. Kawanobe, N. Fujiki, Y. Yasuda, and T. Sano. ature on mortalities of Rainbow Trout, Salmo gairdneri Richardson, caused 1993. Resistance of a Rainbow Trout strain to infectious pancreatic necrosis. by a European strain of infectious pancreatic necrosis virus. Journal of Fish Aquaculture 117:71–76. Diseases 4:213–221. Pietsch, J. P., D. F. Amend, and C. M. Miller. 1977. Survival of infectious Felsenstein, J. 1985. Confidence limits on phylogenies: an approach using the hematopoietic necrosis virus held under various environmental conditions. bootstrap. Evolution 39:783–791. Journal of the Fisheries Research Board of Canada 34:1360–1364. Fijan, N., D. Sulimanovic,´ M. Bearzotti, D. Muzinic,´ L. O. Zwillenberg, S. Reno, P. W. 1998. Factors involved in the dissemination of disease in fish Chilmonczyk, J. F. Vautherot, and P. de Kinkelin. 1983. Some properties populations. Journal of Aquatic Animal Health 10:160–171. of the Epithelioma papulosum cyprini (EPC) cell line from carp Cyprinus Reno, P. W. 1999. Infectious pancreatic and associated aquatic birnaviruses. carpio. Annales de l’Institut Pasteur Virologie 134:207–220. Pages 1–55 in P. T. K. Woo and D. W. Bruno, editors. Fish diseases and Frantsi, C., and M. Savan. 1971. Infectious pancreatic necrosis virus: tempera- disorders. CAB International, Oxfordshire, UK. ture and age factors in mortality. Journal of Wildlife Diseases 7:249–255. Saitou, N., and M. Nei. 1987. The neighbor-joining method: a new method Hawley, L. M., and K. A. Garver. 2008. Stability of viral hemorrhagic septicemia for reconstructing phylogenetic trees. Molecular Biology and Evolution 4: virus (VHSV) in freshwater and seawater at various temperatures. Diseases 406–425. of Aquatic Organisms 82:171–178. Smail, D. A. 1999. Viral haemorrhagic septicaemia. Pages 123–147 in P. T. K. Hill, B. J. 1992. Impact of viral diseases of salmonid fish in the European Woo and D. W. Bruno, editors. Fish diseases and disorders. CAB Interna- Community. Pages 48–59 in T. Kimura, editor. Proceedings of the OJI in- tional, Oxfordshire, UK. ternational symposium on salmonid diseases. Hokkaido University Press, Smail, D. A., P. J. Huntly, and A. L. S. Munro. 1993. Fate of four fish pathogens Sapporo, Japan. after exposure to fish silage containing fish farm mortalities and conditions Jarp, J. 1999. Epidemiological aspects of viral diseases in Norwegian farmed for the inactivation of infectious pancreatic necrosis virus. Aquaculture 113: Atlantic Salmon (Salmo salar L.). Bulletin of the European Association of 173–181. Fish Pathologists 19:240–244. Tamura, K., M. Nei, and S. Kumar. 2004. Prospects for inferring very large phy- Jørgensen, P. E. V. 1982. Egtved virus: occurrence of inapparent infections with logenies by using the neighbor-joining method. Proceedings of the National virulent virus in free-living Rainbow Trout, Salmo gairdneri Richardson, at Academy of Sciences of the USA 101:11030–11035. low temperature. Journal of Fish Diseases 5:251–255. Tamura, K., D. Peterson, N. Peterson, G. Stecher, M. Nei, and S. Kumar. 2011. Lannan, C. N., J. R. Winton, and J. L. Fryer. 1984. Fish cell lines: establishment MEGA5: molecular evolutionary genetics analysis using maximum likeli- and characterization of nine cell lines from salmonids. In Vitro Cellular and hood, evolutionary distance, and maximum parsimony methods. Molecular Developmental Biology 20:671–676. Biology and Evolution 28:2731–2739. McAllister, P. E., and J. Bebak. 1997. Infectious pancreatic necrosis virus in the Thompson, J. D., D. G. Higgins, and T. J. Gibson. 1994. CLUSTAL W: im-

Downloaded by [Department Of Fisheries] at 18:31 20 January 2013 environment: relationship to effluent from aquaculture facilities. Journal of proving the sensitivity of progressive multiple sequence alignment through Fish Diseases 20:201–207. sequence weighting, position-specific gap penalties and weight matrix choice. Nishizawa, T., S. Kinoshita, and M. Yoshimizu. 2005. An approach for Nucleic Acids Research 22:4673–4680. genogrouping of Japanese isolates of aquabirnaviruses in a new genogroup, Toranzo, A. E., and F. M. Metricic. 1982. Comparative stability of two salmonid VII, based on the VP2/NS junction region. Journal of General Virology viruses and poliovirus in fresh, estuarine and marine waters. Journal of Fish 86:1973–1978. Diseases 5:223–231. Nishizawa, T., H. Savas¸, H. Is¸ıdan, C. Ust¨ unda¨ g,˘ H. Iwamoto, and M. Yoshimizu. Wolf, K. 1988. Fish viruses and fish viral diseases. Cornell University Press, 2006. Genotyping and pathogenicity of viral hemorrhagic septicemia virus New York. from free-living Turbot (Psetta maxima) in a Turkish coastal area of the Black Wolf, K., and M. C. Quimby. 1971. Salmonid viruses: infectious pancreatic Sea. Applied and Environmental Microbiology 72:2373–2378. necrosis virus—morphology, pathology, and serology of first European iso- Ogut, H., and A. Akyol. 2005. Prevalence and intensity of Hexamita salmonis lations. Archiv fur¨ die gesamte Virusforschung 34:144–156. in Rainbow Trout farms in the southeastern Black Sea and their relation- Wolf, K., M. C. Quimby, C. P. Carlson, and G. L. Bullock. 1968. Infectious ship to environmental factors. Israeli Journal of Aquaculture Bamidgeh 57: pancreatic necrosis: selection of virus-free stock from a population of carrier 97–104. trout. Journal of the Fisheries Research Board of Canada 25:383–391. This article was downloaded by: [Department Of Fisheries] On: 20 January 2013, At: 18:33 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Aquatic Animal Health Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/uahh20 Experimental Infection of Koi Carp with Viral Hemorrhagic Septicemia Virus Type IVb Emily R. Cornwell a , Sandra L. LaBuda a , Geoffrey H. Groocock a , Rodman G. Getchell a & Paul R. Bowser a a Aquatic Animal Health Program, Department of Microbiology and Immunology, College of Veterinary Medicine, Cornell University, Upper Tower Road, Ithaca, New York, 14853, USA Version of record first published: 04 Jan 2013.

To cite this article: Emily R. Cornwell , Sandra L. LaBuda , Geoffrey H. Groocock , Rodman G. Getchell & Paul R. Bowser (2013): Experimental Infection of Koi Carp with Viral Hemorrhagic Septicemia Virus Type IVb, Journal of Aquatic Animal Health, 25:1, 36-41 To link to this article: http://dx.doi.org/10.1080/08997659.2012.732653

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material. Journal of Aquatic Animal Health 25:36–41, 2013 C American Fisheries Society 2013 ISSN: 0899-7659 print / 1548-8667 online DOI: 10.1080/08997659.2012.732653

COMMUNICATION

Experimental Infection of Koi Carp with Viral Hemorrhagic Septicemia Virus Type IVb

Emily R. Cornwell,* Sandra L. LaBuda, Geoffrey H. Groocock, Rodman G. Getchell, and Paul R. Bowser Aquatic Animal Health Program, Department of Microbiology and Immunology, College of Veterinary Medicine, Cornell University, Upper Tower Road, Ithaca, New York 14853, USA

several smaller bodies of water within the Great Lakes basin, Abstract and one reservoir outside of the basin (Bain et al. 2010). The Viral hemorrhagic septicemia virus (VHSV) type IVb has a virus has been isolated from at least 28 fish species present wide host range that includes at least three cyprinid species: in the Great Lakes (OIE 2009) and was also isolated from the Fathead Minnow Pimephales promelas, Emerald Shiner Notropis atherinoides, and Bluntnose Minnow P. notatus. To date, VHSV amphipod Diporeia spp. (Faisal and Winters 2011) and leech IVb has only been found in wild fish. However, the possibility of Myzobdella lugubris (Faisal and Schulz 2009) collected in the infection in culture facilities remains. Koi Carp Cyprinus carpio Great Lakes. Recent evidence suggests that VHSV may infect are a major ornamental aquaculture species in the United States; other animals besides finfish: experimental infection of three however, their potential to become infected with VHSV IVb has different species of turtles with VHSV showed virus replication not yet been examined. In this study, we exposed Koi to 3 × 106 PFU VHSV Great Lakes isolate MI03 by intraperitoneal injec- in turtles fed infected Bluegill Lepomis macrochirus (Goodwin tion. While we observed low mortality (0–5%), VHSV was isolated and Merry 2011). in cell culture from the majority of fish up to 28 d postexposure In North America, reports of VHSV infection to date have (DPE) and was detected by a quantitative reverse transcription been confined to populations of wild fish. However, the possi- polymerase chain reaction (qRT-PCR) assay up to 90 DPE, when bility of VHSV spreading to cultured fish remains and has the the trial was terminated. The results of this study strongly suggest that Koi are at risk for VHSV infection, although their susceptibil- potential to be devastating given the high susceptibility of some ity by intraperitoneal injection appears to be low. This study also cultured species and the high density of fish in culture (VHSV provides more evidence of the sensitivity of qRT-PCR for detection Expert Panel and Working Group 2010). of VHSV IVb. Koi Carp Cyprinus carpio represent a major portion of U.S. ornamental fish aquaculture. Although other ornamental species Viral hemorrhagic septicemia virus (VHSV) is the causative have greater total sales, more farms raise Koi than any other agent of a devastating disease of finfish, viral hemorrhagic ornamental fish species in the USA. In 2005, there were 193 Downloaded by [Department Of Fisheries] at 18:33 20 January 2013 septicemia (VHS). Originally described in European Rainbow farms in the USA culturing Koi, with sales totaling more than Trout Oncorhynchus mykiss (Schaperclaus¨ 1938), the known US$6.5 million (USDA 2006). Koi are known to be susceptible host and geographic range of VHSV has expanded significantly to at least one other rhabdovirus, spring viremia of carp virus over the past 30 years (Kim and Faisal 2011). Currently, VHSV (Goodwin 2002). has been isolated from 82 fish species worldwide, including five Although to our knowledge VHSV has not been previously cyprinid species (OIE 2009). detected in Koi, several other cyprinid species have been shown There are four genotypes of VHSV that are correlated with to be susceptible to VHSV IVb. Laboratory infection is possi- geography (Elsayed et al. 2006). The most recent geographic ble in Fathead Minnow Pimephales promelas, where intraperi- expansion of VHSV occurred in the Laurentian Great Lakes, toneal injection resulted in 50% mortality and waterborne ex- where all isolates to date fall within genotype IVb (Thompson posure resulted in 13% mortality (Al-Hussinee et al. 2010). A et al. 2011). The IVb genotype of VHSV has been detected in 2006 surveillance effort for VHSV in New York State detected North America in all five Great Lakes (Cornwell et al. 2011), VHSV in Emerald Shiner Notropis atherinoides and Bluntnose

*Corresponding author: [email protected] Received May 9, 2012; accepted September 14, 2012

36 COMMUNICATION 37

Minnow P. notatus using virus isolation and a quantitative re- repeated between 16 and 22 times (representing the number of verse transcription polymerase chain reaction (qRT-PCR) assay fish to be allocated to that tank; Figure 1). The order of this list (Frattini et al. 2011). This qRT-PCR assay has been shown to was randomized using a randomizer program (Haahr 2011) to be more sensitive than virus isolation in cell culture, but it can- allow random allocation of fish into tanks. Two tanks of 18 ex- not determine whether detected virus is infective (Hope et al. posed fish and one tank of 16 control fish were used to establish 2010). Additionally, VHSV was detected by virus isolation in survivorship. In these tanks, five fish were sampled at 60 DPE cell culture from a Common Carp C. carpio collected in 2007 and the remaining fish were sampled at 90 DPE. Two tanks of 22 by our laboratory; however, this detection was not confirmed by exposed fish and one tank of 16 control fish were used for early the U.S. Department of Agriculture Animal and Plant Health In- periodic sampling. At day 0 (immediately after exposure), two spection Service National Veterinary Services Laboratory. This exposed fish from each tank and three control fish were sampled. unconfirmed detection, combined with the known susceptibility At 7 DPE, three control fish and four exposed fish from each of other cyprinid species, led us to hypothesize that experimental tank were sampled. At 14 and 21 DPE, three control fish and infection of Koi with VHSV IVb would be possible. five exposed fish from each tank were sampled, respectively. At 28 DPE, the remaining control and exposed fish were sampled. Strict biosecurity measures were taken during the infection METHODS trial. All personnel wore gloves and shoe covers, and used foot Study organisms.—Koi that did not pass initial color grading baths prior to entering rooms where fish were housed. All work were donated from a commercial Koi producer in New York with control fish was completed each day prior to work with fish State. Fish used in the experiment were: (1) between 50- and exposed to VHSV. The effluent from all tanks was treated with 90-mm TL (mean = 68-mm TL), (2) between 2.2 and 12.7 g 1 mg/L free chlorine residual for 10 min prior to discharge into (mean = 5.9 g), (3) from a long-term U.S. stock, and (4) approx- the city water treatment plant. Euthanasia was achieved with imately 60-d-old at the start of the experiment. Fish were fed approximately 500 mg/L tricaine methanesulfonate (Western Rangen starter number 3 for cultured fish (Buhl, Idaho) at ap- Chemical, Ferndale, Washington), buffered 1:1 (weight : proximately 1% body weight once daily. Fish were acclimated weight) with sodium bicarbonate (Sigma-Aldrich, St. Louis, to 10 ± 1◦C over a period of 2 weeks in a 700-L flow-through Missouri). All fish that died or were euthanized during the tank (Frigid Units). During this acclimation period, 10 fish were trial were carefully examined for gross external and internal euthanized, a complete necropsy as described in the American lesions prior to collection of tissues for VHSV detection. All Fisheries Society Blue Book (AFS-FHS 2010) was performed, experimental procedures described in this study were approved and tissue homogenates from all fish were inoculated onto by the Cornell Institutional Animal Care and Use Committee. Epithelioma papulosum cyprini (EPC) cells (Fijan et al. 1983; Detection of viral hemorrhagic septicemia virus.—Virus was Winton et al. 2010) to test for the presence of common fish detected using two methods: isolation in cell culture and qRT- viruses. PCR. Virus isolation in cell culture was achieved using a mod- Study design.—The MI03 isolate of VHSV was used to ification of that described in the American Fisheries Society expose Koi. This isolate was originally obtained from a Blue Book (AFS-FHS 2010). Heart, liver, spleen, and anterior Muskellunge Esox masquinongy (Elsayed et al. 2006) and has and posterior kidneys were aseptically dissected from each fish been used in multiple susceptibility trials for VHSV IVb (Kim at the time of death using sterile and nucleic acid-free instru- and Faisal 2010a,b,c). Confluent monolayers of EPC cells grown ments. Pooled tissue samples were homogenized with 9 vol- as described in Groocock et al. (2007) were inoculated at a mul- umes HMEM-10 in a sterile 2-mL homogenizing tube contain- tiplicity of infection of 0.005 PFU, incubated at 15◦C, and mon- ing approximately 0.25 g of 0.1-mm zirconium–silica beads Downloaded by [Department Of Fisheries] at 18:33 20 January 2013 itored daily until 90% cytopathic effect (CPE) was observed. (BioSpec Products, Bartlesville, Oklahoma) for 1 min using a The cells and media from each flask were harvested and cen- Mini-Beadbeater-16 (BioSpec Products). A portion of the re- trifuged at 5,000 × g to pellet cellular debris, the supernatant sulting supernatant was further diluted to 1:49 using HMEM-10 was filtered through a 0.22-µm sterile Acrodisc filter (Pall Life and filtered through a sterile 0.45-µm Acrodisc filter (Pall Life Sciences, Port Washington, New York),aliquoted into 1-mL cry- Sciences). An additional dilution was made from the filtered ovials, and frozen at –80◦C. The concentration of the stock virus 1:49 homogenate to 1:249. Each dilution (250 µL) was inocu- was determined by plaque assay (Batts and Winton 1989) imme- lated onto three wells on 48-well plates (CoStar-3548, Corning diately after freezing and immediately prior to exposure of Koi. Life Sciences, Lowell, Massachusettes), confluent monolayers Fish were exposed to 3 × 106 PFU of VHSV MI03 by of EPC cells containing 250 µL HMEM-5 (HMEM-10 sup- intraperitoneal injection (volume of injection = 0.1 mL). plemented with only 5% fetal bovine serum) resulting in final, All controls were injected with 0.1-mL sterile HMEM-10 in-well dilutions of 1:99 and 1:499. The plates were placed in a cell culture media (minimal essential medium with Hanks’ sealed container in an incubator at 15◦C and monitored at least salts supplemented with 10% fetal bovine serum, penicillin once a week for CPE. Wells showing CPE were passaged as [100 IU/mL], streptomycin [100 µg/mL], and HEPES buffer soon as CPE was detected, while wells not showing CPE were [1 M 0.015 mL/mL] [Gibco]). Immediately prior to exposure, blind passaged after 2 weeks. All wells were subjected to at a list was created with the destination 10-L tank (tanks 1–6) least two passages to confirm the presence or absence of CPE. 38 CORNWELL ET AL.

FIGURE 1. Experimental design used to determine whether Koi have the possibility of becoming infected with VHSV IVb. The number (#) of ﬁsh sampled from each tank on a speciﬁc day postexposure (DPE) is listed below each tank graphic. Any mortalities that occurred in survivorship and early sampled tanks are described in bold text.

Quantiﬁcation of virus via plaque assay or ranking of intensity sample and are reported as VHSV N gene copies/50 ng RNA. of CPE was not performed for ﬁsh tissues. A Kruskal–Wallis rank-sum test was used to test for differences Extraction of RNA was performed using a MagMax mag- in median viral load by day. An increase in mean viral load

Downloaded by [Department Of Fisheries] at 18:33 20 January 2013 netic bead extraction system and the MagMax-96 viral RNA between day 0 and sampling points during the first 2 weeks isolation kit (Life Technologies, Carlsbad, California) using the postexposure (7 and 14 DPE) was tested using a two-sample protocols described in the kit and the manufacturer-supplied Student’s t-test on log-transformed, standardized qRT-PCR re- extraction program AM1836 DW 50v2. The 1:9 dilution of tis- sults. A logistic regression model was fitted in R to determine sue homogenate described above was used to extract RNA. All the relationship between qRT-PCR quantity (measured in viral RNA was stored at –80◦C until the qRT-PCR assay was per- N gene copies/50 ng RNA) and the probability of isolation in formed. The qRT-PCR assay used was slightly modified from cell culture, expressed as that described by Hope et al. (2010) in that plates were run on an ABI 7500 detector, a linear regression was calculated from five Logit (cell culture test positive) ∼ qRT-PCR quantity. RNA standards ranging from 1.5 × 102 to 1.5 × 106, and all plates were run for 45 cycles. Quality and quantity of RNA was measured using a NanoVue spectrophotometer (GE Healthcare, RESULTS Piscataway, New Jersey). Over the course of the experiment survival was 100%, ex- Data analysis.—Data were analyzed using R version 2.10.1 cept for fish in three tanks where survival dropped to approx- (R Development Core Team 2011). Results of the qRT-PCR imately 95% within the first 2 d of exposure (one fish from assay were standardized to the amount of RNA present in the each of the early sampling tanks, and one fish from one of the COMMUNICATION 39

FIGURE 2. Survival and viral loads of Koi exposed to VHSV type IVb by intraperitoneal injection from both survivorship tanks and early sampling tanks. Survivorship is shown on the left axis and was 100% (solid black line) for all tanks except three, where survival reached approximately 95% on day 1 (one tank, dashed grey line) or day 2 (two tanks, dotted black line for early sampling tank and solid grey line for survivorship tank). Viral load as detected by qRT-PCR is shown on the right axis. Viral loads of the three dead fish can be seen on days 1 and 2; all other viral loads are from fish sampled periodically during the experiment. All controls tested negative for VHSV by qRT-PCR and virus isolation in cell culture (open squares; numbers tested at each sampling point shown in Figure 1); some exposed fish tested positive for VHSV by qRT-PCR and virus isolation in cell culture (solid circles), while some exposed fish tested positive for VHSV only by qRT-PCR (open circles). Median viral load by qRT-PCR is indicated by a horizontal line at each sampling point. Numbers below the arrows along the x-axis represent prevalence by qRT-PCR for that sampling timepoint.

survivorship–late sampling tanks; Figures 1, 2). Virus was de- There was a signiﬁcant difference in median viral load by

Downloaded by [Department Of Fisheries] at 18:33 20 January 2013 2 tected by qRT-PCR from all three dead fish, but virus was iso- DPE (χkw = 50.4, df = 6, P < 0.0001). When qRT-PCR results lated in cell culture from only two of the three dead fish. Virus were log transformed, the mean viral load was significantly was isolated in cell culture and detected by qRT-PCR from all lower on 0 DPE than on 7 and 14 DPE (t = –1.9973, P = exposed fish on sampling days 0, 7, and 14 (n = 22). On 21 and 0.029), giving evidence for VHSV replication in Koi. 28 DPE, virus was detected by qRT-PCR from all exposed fish, The median viral load of positive fish where virus isolation and virus was isolated in cell culture from 7 of 10 sampled fish was not possible was 0, while the median viral load of positive on 21 DPE and 9 of 10 fish on 28 DPE. At 60 DPE, VHSV was fish where virus isolation was possible was 3.9 × 103. A logis- detected by qRT-PCR in 8 of 10 fish and no virus was isolated tic regression model testing the association between the stan- from any fish. When the experiment was terminated, at 90 DPE, dardized quantity of virus detected by qRT-PCR and successful 14 of 25 exposed fish tested positive by qRT-PCR and no virus isolation of VHSV in cell culture was created, expressed as was isolated in cell culture from any fish. Gross clinical signs observed at necropsy in exposed fish during the experiment were Logit (cell culture test positive) very mild and included hemorrhage on abdominal organs (1/80, =−1.97 + 0.0034(qRT-PCR quantity) + error. 7 DPE) and scleral hemorrhage (1/80, 90 DPE). Throughout the experiment, all controls tested negative by qRT-PCR and virus This model had a residual deviance of 67.48 on 107 df and an isolation. Akaike’s information criterion (AIC) of 71.48 (compared with 40 CORNWELL ET AL.

TABLE 1. Results of null and full logistic regression models developed to the natural immune response of Koi. Future studies of Koi sus- test the association between virus isolation in cell culture and quantity detected ceptibility using more natural routes of exposure, such as wa- by qRT-PCR. terborne or oral challenge, will be necessary to examine the risk Model Predictor Estimate SE ZP-value AIC of VHSV infection in Koi in a culture or hobby setting. Although the reported host range of VHSV IVb is wide Null Intercept –0.63 0.20 –3.11 0.0019 142.96 and encompasses 13 families of finfish, susceptibility varies < Full Intercept –1.97 0.34 –5.78 0.0001 71.48 greatly between species. Muskellunge have a reported median qRT-PCR 0.0034 0.0012 –2.75 0.0060 lethal dose by intraperitoneal injection of 2.21 PFU (Kim and Faisal 2010c), while rainbow trout experienced only a 4.4% mortality after injection with 106.5 50% tissue culture infective an AIC of 142.96 for a null model; Table 1). This model predicts dose (Al-Hussinee et al. 2010). that using the protocols described in this study a sample with a To date, no natural infections of Koi with VHSV have been qRT-PCR quantity of 1.5 × 103 has a 95% probability of also reported. There has been one unconfirmed report of VHSV in- testing positive by virus isolation in cell culture in these fish. fection in a wild Common Carp. However, the evidence for VHSV replication in Koi presented in this study suggests that Koi farms may be at some risk for VHSV infection. More studies DISCUSSION are needed to determine whether Koi are susceptible to VHSV In this study, 80 Koi were exposed to VHSV by intraperi- under natural conditions as well as their relative susceptibility toneal injection. Mortality was very low and occurred in the first to VHSV IVb compared with other species. 2 d of the experiment, suggesting that deaths could be due to handling stress and not VHSV infection. Nevertheless, the results ACKNOWLEDGMENTS of this experiment provide strong evidence that Koi have the We thank Gregory A. Wooster for technical assistance. potential to become infected with VHSV IVb. Two Koi showed Sandra L. LaBuda was the recipient of a Summer Research mild clinical signs associated with VHSV infection. The virus Fellowship from the AQUAVET Program, a program in aquatic quantity present in fish increased slightly but significantly be- veterinary medicine jointly sponsored by the School of Veteri- tween exposure and the first 2 weeks postexposure. In addition, nary Medicine of the University of Pennsylvania and the College infectious VHSV was isolated from many fish up to 28 DPE. of Veterinary Medicine of Cornell University. This report is a Virus was detected by qRT-PCR up to 90 DPE (Figure 2), albeit resulting product from project R/FTD-11 funded under award at very low levels, suggesting that Koi have the potential to be NA10OAR4170064 from the National Sea Grant College Pro- long-term carriers of VHSV viral genomic material, although it gram of the U.S. Department of Commerce’s National Oceanic is not known at this point whether virus remaining at these later and Atmospheric Administration, to the Research Foundation of time points is viable, infectious virus. State University of New York on behalf of New York Sea Grant. The quantity of virus where successful VHSV isolation The statements, findings, conclusions, views, and recommen- in cell culture could reliably be conducted in this study was dations are those of the authors and do not necessarily reflect 1.5 × 103 N gene copies/50 ng RNA, approximately one order the views of any of those organizations. This project was also of magnitude lower than the limit of detection reported in Hope supported in part by an Assistance Award to Cornell University et al. (2010). The discrepancy is possibly explained by the from grant 2007.883 made by the Great Lakes Fishery Trust to methods used in the current study, where tissue homogenates the Western Fisheries Research Center, U.S. Geological Survey, were always plated immediately onto cells and never frozen. Downloaded by [Department Of Fisheries] at 18:33 20 January 2013 Seattle, Washington. Freeze–thaw cycles have been shown to decrease the amount of infectious VHSV IVa present in tissue (Arkush et al. 2006), and it is likely that freeze–thaw cycles have a similar effect on IVb. REFERENCES While this study presents strong evidence that Koi may be AFS-FHS (American Fisheries Society-Fish Health Section). 2010. FHS blue susceptible to VHSV, there are several limitations to this study. book: suggested procedures for the detection and identification of certain Conducting histopathology would have allowed visualization finfish and shellfish pathogens, 2010 edition. AFS-FHS, Bethesda, Maryland. of tissue destruction caused by VHSV infection in Koi. In this Al-Hussinee, L., P.Huber, S. Russell, V.LePage, A. Reid, K. M. Young,E. Nagy, R. M. W.Stevenson, and J. S. Lumsden. 2010. Viral haemorrhagic septicaemia study the size of the fish used prohibited the use of organs for virus IVb experimental infection of Rainbow Trout, Oncorhynchus mykiss cell culture, qRT-PCR, and histopathology. Furthermore, due to (Walbaum), and Fathead Minnow, Pimphales promelas (Rafinesque). Journal the low mortalities experienced during this trial, the results from of Fish Diseases 33:347–360. the early sampling tanks and survivorship tanks were combined. Arkush, K. D., H. L. Mendonca, A. M. McBride, S. Yun, T. S. McDowell, and These two treatment groups had varying population densities, R. P. Hedrick. 2006. Effects of temperature on infectivity and of commercial freezing on survival of the North American strain of viral hemorrhagic and it is likely this would affect viral load kinetics. This ex- septicemia virus (VHSV). Diseases of Aquatic Organisms 69:145–151. periment used an unnatural route of infection (intraperitoneal Bain, M. B., E. R. Cornwell, K. M. Hope, G. E. Eckerlin, R. N. Casey, injection), which has the possibility of circumventing some of G. H. Groocock, R. G. Getchell, P. R. Bowser, J. R. Winton, W. N. Batts, COMMUNICATION 41

A. Cangelosi, and J. W. Casey. 2010. Distribution of an invasive aquatic Hope, K. M., R. N. Casey, G. H. Groocock, R. G. Getchell, P. R. Bowser, and pathogen (viral hemorrhagic septicemia virus) in the Great Lakes and its re- J. W. Casey. 2010. Comparison of quantitative RT-PCR with cell culture to lationship to shipping. PLoS (Public Library of Science) ONE [online serial] detect viral hemorrhagic septicemia virus (VHSV) IVb infections in the Great 5(4):e10156. DOI: 10.1371/journal.pone.0010156. Lakes. Journal of Aquatic Animal Health 22:50–61. Batts, W. N., and J. R. Winton. 1989. Enhanced detection of infectious Kim, R., and M. Faisal. 2010a. Experimental studies confirm the wide host hematopoietic necrosis virus and other fish viruses by pretreatment of cell range of the Great Lakes viral haemorrhagic septicaemia virus genotype IVb. monolayers with polyethylene glycol. Journal of Aquatic Animal Health Journal of Fish Diseases 33:83–88. 1:284–290. Kim, R., and M. Faisal. 2010b. Comparative susceptibility of representative Cornwell, E. R., G. E. Eckerlin, R. G. Getchell, G. H. Groocock, T. M. Great Lakes fish species to the North American viral hemorrhagic septicemia Thompson, W. N. Batts, R. N. Casey, G. Kurath, J. R. Winton, P. R. Bowser, virus sublineage IVb. Diseases of Aquatic Organisms 91:23–34. M. B. Bain, and J. W. Casey. 2011. Detection of viral hemorrhagic septicemia Kim, R., and M. Faisal. 2010c. The Laurentian Great Lakes strain (MI03) of virus by quantitative reverse transcription polymerase chain reaction from two the viral haemorrhagic septicaemia virus is highly pathogenic for juvenile fish species at two sites in Lake Superior. Journal of Aquatic Animal Health Muskellunge, Esox masquinongy (Mitchill). Journal of Fish Diseases 33: 23:207–217. 513–527. Elsayed, E., M. Faisal, M. Thomas, G. Whelan, W. Batts, and J. Winton. 2006. Kim, R., and M. Faisal. 2011. Emergence and resurgence of the viral hem- Isolation of viral haemorrhagic septicemia virus from Muskellunge Esox orrhagic septicemia virus (Novirhabdovirus, Rhabdoviridae, Mononegavi- masquinongy (Mitchill), in Lake St. Clair, Michigan, USA reveals a new rales). Journal of Advanced Research 2:9–23. sublineage of the North American genotype. Journal of Fish Diseases 29:611– OIE (World Organization for Animal Health). 2009. Viral haemor- 619. rhagic septicaemia. Pages 279–298 in Aquatic Animal Health Stan- Faisal, M., and C. A. Schulz. 2009. Detection of viral hemorrhagic septicemia dards Commission, editors. Manual of diagnostic tests for aquatic ani- virus (VHSV) from the leech Myzobdella lugubris Leidy, 1851. Parasites and mals, 6th edition. OIE, Paris. Available: www.oie.int/fileadmin/Home/eng/ Vectors 2:45–49. Healthstandards/aahm/2010/2.3.09.VHS.pdf. (January 2012). Faisal, M., and A. D. Winters. 2011. Detection of viral hemorrhagic septicemia R Development Core Team. 2011. R: a language and environment for statisti- virus (VHSV) from Diporeia spp. (Pontoporeiidae, Amphipoda) in the Lau- cal computing. R Foundation for Statistical Computing, Vienna. Available: rentian Great Lakes, USA. Parasites and Vectors 4:2–5. www.R-project.org. (January 2012). Fijan, N., D. Sulimanovic,´ M. Bearzotti, D. Muzinic,´ L. O. Zwillenberg, S. Schaperclaus,¨ W. 1938. Die Schadigungen¨ der deutschen Fischerei durch Fis- Chilmonczyk, J. F. Vautherot, and P. de Kinkelin. 1983. Some properties chparasiten und Fischkrankheiten. [The damage done to German fishery of the Epithelioma papulosum cyprini (EPC) cell line from carp Cyprinus by fish parasites and fish diseases.] Allgemeine Fischerei-Zeitung 41:256– carpio. Annales de l’Institut Pasteur Virologie 134E:207–220. 259. Frattini, S. A., G. H. Groocock, R. G. Getchell, G. A. Wooster, R. N. Casey, Thompson, T. M., W. N. Batts, M. Faisal, P. Bowser, J. W. Casey, K. Phillips, J. W. Casey, and P. R. Bowser. 2011. A 2006 survey of viral hemorrhagic K. A. Garver, J. Winton, and G. Kurath. 2011. Emergence of viral hemor- septicemia (VHSV) virus type IVb in New York State waters. Journal of Great rhagic septicemia virus in the North American Great Lakes region is as- Lakes Research 37:194–198. sociated with low viral genetic diversity. Diseases of Aquatic Organisms Goodwin, A. E. 2002. First report of spring viremia of carp virus (SVCV) in 96:29–43. North America. Journal of Aquatic Animal Health 14:161–164. USDA (U.S. Department of Agriculture). 2006. Census of aquaculture (2005), Goodwin, A. E., and G. E. Merry. 2011. Replication and persistence of VHSV volume 3, special studies, part 2. USDA, National Agricultural Statistics IVb in freshwater turtles. Diseases of Aquatic Organisms 94:173–177. Service, AC-02-SP-2, Washington, D.C. Available: www.agcensus.usda.gov/ Groocock, G. H., R. G. Getchell, G. A. Wooster, K. L. Britt, W. N. Batts, J. R. Publications/2002/Aquaculture/AQUACEN.pdf. (January 2012). Winton, R. N. Casey, J. W. Casey, and P. R. Bowser. 2007. Detection of viral VHSV Expert Panel and Working Group. 2010. Viral hemorrhagic septicemia hemorrhagic septicemia in Round Gobies in New York State (USA) waters virus (VHSV IVb) risk factors and association measures derived by expert of Lake Ontario and the St. Lawrence River. Diseases of Aquatic Organisms panel. Preventive Veterinary Medicine 94:128–139. 76:187–192. Winton, J., W. Batts, P. DeKinkelin, M. LeBerre, M. Bremont, and N. Fijan. Haahr, M. 2011. Random.org: true random number service list randomizer. Ran- 2010. Current lineages of the epithelioma papulosum cyprini (EPC) cell line domness and Integrity Services, Dublin. Available: www.random.org/lists. are contaminated with Fathead Minnow, Pimephales promelas, cells. Journal (July 2010). of Fish Diseases 33:701–704. Downloaded by [Department Of Fisheries] at 18:33 20 January 2013 This article was downloaded by: [Department Of Fisheries] On: 28 February 2013, At: 23:30 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Aquatic Animal Health Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/uahh20 Mixed Metazoan and Bacterial Infection of the Gas Bladder of the Lined Seahorse—A Case Report Paul A. Anderson a & Barbara D. Petty a a Program in Fisheries and Aquatic Sciences, School of Forest Resources and Conservation, Institute of Food and Agricultural Sciences, University of Florida, 7922 Northwest 71st Street, Gainesville, Florida, 32653, USA Version of record first published: 23 Jan 2013.

To cite this article: Paul A. Anderson & Barbara D. Petty (2013): Mixed Metazoan and Bacterial Infection of the Gas Bladder of the Lined Seahorse—A Case Report, Journal of Aquatic Animal Health, 25:1, 42-52 To link to this article: http://dx.doi.org/10.1080/08997659.2012.743932

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material. Journal of Aquatic Animal Health 25:42–52, 2013 C American Fisheries Society 2013 ISSN: 0899-7659 print / 1548-8667 online DOI: 10.1080/08997659.2012.743932

COMMUNICATION

Mixed Metazoan and Bacterial Infection of the Gas Bladder of the Lined Seahorse—A Case Report

Paul A. Anderson* and Barbara D. Petty Program in Fisheries and Aquatic Sciences, School of Forest Resources and Conservation, Institute of Food and Agricultural Sciences, University of Florida, 7922 Northwest 71st Street, Gainesville, Florida 32653, USA

All of the submitted specimens were wild-caught bycatch Abstract of live-bait shrimpers operating off the central west coast of Five wild-caught Lined Seahorses Hippocampus erectus from an Florida. After capture, seahorses were typically held in mul- aquarium system presented with altered buoyancy and distended tispecies systems containing both vertebrate and invertebrate upper trunks. Radiography of one specimen revealed a reduced air volume in the gas bladder. Pneumocystocentesis revealed a animals for several days to several weeks before being donated brown exudate of numerous leukocytes, parasite ova, and Gram- to an aquaculture research laboratory. The Lined Seahorses pre- and acid-fast-positive bacilli under wet mounts and stains. Necrop- sented with clinical signs after being housed in the research sies revealed enlarged, friable kidneys and distended gas bladders laboratory for 3.50–7.25 months. The fish were maintained in containing copious purulent exudate, necrotic tissue, and adult di- a 2,082-L, species-specific system consisting of a series of 76- geneans Dictysarca virens. Bacterial isolates from exudate cultures grown on Lowenstein–Jensen medium were identified as Gordonia L, bare-bottom holding tanks outfitted with plastic plants for sp. and Mycobacterium poriferae by high-performance liquid chro- holdfasts. The system featured wet/dry trickle filtration, passive matography and 16S ribosomal DNA sequencing. Histopathology flow over granular activated carbon media, foam fractionation, demonstrated a histiocytic response in kidney and gas bladder ex- and ultraviolet light sterilization. Debris was siphoned from the udate, inflammation of the gas bladder wall, and infection of the tanks twice daily, necessitating 10% weekly water changes. Wa- gas bladder lumen with parasite ova and acid-fast-positive and ◦ Gomori’s methenamine silver-positive bacilli. Praziquantel is pre- ter quality was as follows: temperature was 26–27 C, salinity scribed for digenean infections but dissolves incompletely in seawa- was 28 g/L, pH was 8.24, total ammonia nitrogen was 0 mg/L, ter and is toxic to this host. Eradication of intermediate host vectors nitrite was 0.3 mg/L, and nitrate ranged from 12.5 to 25.0 mg/L. is a management option. Treatment of Gordonia infection has not Animals were fed frozen mysids (Piscine Energetics brand) once been addressed in nonhuman animals, and there is no known ef- daily and were also given live brine shrimp Artemia sp. (enriched fective treatment for Mycobacterium spp. infection in fishes. This is the first case report of digenean infection of the gas bladder in with Roti-Rich) once daily. a syngnathid, Gordonia sp. infection in a nonhuman animal, and Several biosecurity measures were implemented in the M. poriferae infection in a fish. research laboratory. Incoming animals (including the submitted Downloaded by [Department Of Fisheries] at 23:30 28 February 2013 specimens) were quarantined for 1 month prior to inclusion Five wild-caught Lined Seahorses Hippocampus erectus in the system. During this time, animals were subjected to a were submitted to the Fish Health Extension Laboratory (Pro- 5-min freshwater dip; in addition, two chloroquine diphosphate gram in Fisheries and Aquatic Sciences, School of Forest Re- bath treatments were applied at 10.6 mg/L for 1 d each and sources and Conservation, Institute of Food and Agricultural were spaced 21 d apart. A 2-mg/L praziquantel bath for 7 d Sciences, University of Florida, Gainesville) between October was attempted; however, the researcher indicated concerns 2005 and January 2006, presenting with negative buoyancy (Fig- about the drug fully dissolving in the bath, so the efficacy of ure 1A) and laterally distended upper trunks (Figure 1B), except this treatment was questionable. Other biosecurity measures for one animal, which presented with positive buoyancy. All five included 10-mg/L chlorine treatment of incoming Atlantic individuals otherwise fed normally and exhibited normal activ- Ocean seawater for a minimum of 1 d, with subsequent neutral- ity levels. No increased respiration was observed. ization, salinity adjustment, and pH adjustment prior to use. All

*Corresponding author: [email protected] Received May 25, 2012; accepted October 12, 2012 Published online January 23, 2013

42 COMMUNICATION 43

FIGURE 1. (A) A live Lined Seahorse demonstrating negative buoyancy and (B) a freshly euthanized Lined Seahorse exhibiting a laterally distended upper trunk (at arrows). [Figure available online in color.] Downloaded by [Department Of Fisheries] at 23:30 28 February 2013

equipment was disinfected after use: plastics were immersed indicative of a reduced volume of air in the gas bladder in a 200-mg/L chlorine bath for 1 h followed by neutralization; (Figure 2B). nets were disinfected with a 10-min bath in a quaternary Pneumocystocentesis was performed on two affected ani- ammonium compound solution; and knives and metal imple- mals (Table 1). A brown effusion was recovered from both and ments were coated with 70% ethanol and subsequently air was examined under wet mount. Wet mount examination re- dried. vealed thick-walled, elliptical, clear parasite eggs (approximate dimensions of 27 × 15 µm); many motile and nonmotile bacteria; and many leukocytes (Figure 3A). Smears of the effusion DIAGNOSTICS were prepared and stained with Dif-Quik (both fish), Gram (one One clinically normal fish and one affected fish were sub- fish), and acid-fast (both fish) stains. The Dif-Quik stain enabled mitted for radiography. Radiography in the normal individual the visualization of many macrophages and many bacilli (Fig- demonstrated a normally filled gas bladder as a large radiolu- ure 3B). These bacilli were weakly Gram positive (Figure 3C) cent area in the upper trunk (Figure 2A). The affected ani- and acid-fast positive (Figure 3D). From these observations, we mal demonstrated a markedly reduced area of radiolucency, characterized the effusion as an exudate. 44 ANDERSON AND PETTY

FIGURE 2. Radiographs of (A) normal and (B) affected Lined Seahorses. The large radiolucent area (arrow) in the upper trunk of the normal specimen indicates the presence of a large gas bladder. In contrast, the markedly reduced area of radiolucency (arrow) in the affected specimen indicates a reduced volume of air in the gas bladder. [Figure available online in color.]

Downloaded by [Department Of Fisheries] at 23:30 28 February 2013 NECROPSY In the first two animals examined, friable gas bladder walls Affected animals were euthanized via immersion in tricaine were inadvertently torn while opening the body cavity, releasing methanesulfonate (MS-222) at 1,000 mg/L of seawater buffered a brown exudate into the coelom of each animal. Examination of with 2,000-mg/L sodium bicarbonate. Submitted animals pre- the gas bladders of all five animals under a dissecting microscope sented grossly with no remarkable external conditions except revealed thickened walls, exudate, unidentifiable necrotic tissue, distended upper trunks. Otherwise, the fish seemed to be in good and adult digeneans (Figure 5). Digeneans were preserved by condition and exhibited size-appropriate weights. Gross internal (1) pipetting into near-boiling saline for 30 s to kill them; (2) examination revealed a variety of anomalous findings, the most transferring them to a vial of 10% formalin, where they were prominent of which were an enlarged, hardened gas bladder; ex- fixed for 24–48 h; and (3) transferring them to 70% ethanol (A. tremely friable and enlarged posterior kidneys; and pale, friable Dove, Stony Brook University, personal communication). Fixed livers (Table 1; Figure 4). The one positively buoyant animal digenean specimens were submitted to A. Dove, who identified contained a vesicle filled with gas, serosanguinous fluid, and a them as Dictysarca virens (Linton 1910; Manter 1947). purulent effusion in the coelom adjacent to the posterior kidney, Wet mounts of internal organs (gills, heart, gallbladder, gas which we presumed to be causative of that individual’s positive bladder, liver, gastrointestinal tract, and gonads) of all five Lined buoyancy. Seahorses were largely unremarkable. The exceptions were that COMMUNICATION 45

TABLE 1. Summary of pertinent diagnostic procedures performed on submitted specimens of Lined Seahorse, and the results of those procedures (+=positive; −=negative; NE = not examined; text = see text for full description).

Case number Procedure or observation 1 2 3 4 5 Date examined 11 Oct 2005 18 Oct 2005 22 Dec 2005 13 Jan 2006 25 Jan 2006 Sex Male Male Male Male Female Total length (mm) 170 141 152 175 146 Weight (g) NE 11.9 15.9 29.0 15.5 Laterally distended upper trunk ++ + + + Buoyancy −− − + − Pneumocystocentesis Brown exudate with ova, granulocytes, bacilli + NE NE NE + Gram stain + NE NE NE NE Acid-fast stain + NE NE NE + Necropsy gross findings Coelom: Exudate released from gas bladder into ++ − − − coelom Gas-filled vesicle adjacent to posterior kidney −− − + − Gas bladder enlarged, hardened, containing ++ + + + brown exudate and adult Dictysarca virens Posterior kidney Enlarged NE Enlarged Liquefied Very friable Liver Pale NE Pale, fatty Enlarged Pale, friable Wet mounts Gas bladder with adult digenetic trematodes, ++ + + + many parasite ova, many bacteria, many granulocytes Liver with many enlarged vacuoles in ++ + + + hepatocytes Microbiology Gas bladder: Blood agar medium Pasteurella − Staphylococcus + NE multocida sp. Lowenstein–Jensen medium − Mycobacterium Gordonia sp. − Mycobacterium poriferae poriferae Marine agar medium −− −NE NE Downloaded by [Department Of Fisheries] at 23:30 28 February 2013 Posterior kidney: Blood agar medium + NE NE −+ Lowenstein–Jensen medium NE NE NE NE NE Marine agar medium − NE NE NE NE Coelom: Blood agar medium − NE NE + NE Lowenstein–Jensen medium NE NE NE NE NE Marine agar medium − NE NE NE NE Histopathology Hematoxylin and eosin stain NE Text Text NE NE Acid-fast stain NE ++NE NE Gomori’s methenamine silver stain NE + NE NE NE Periodic acid–Schiff stain NE + NE NE NE 46 ANDERSON AND PETTY

FIGURE 3. Photomicrographs of Lined Seahorse gas bladder exudate: (A) wet mount, revealing thick-walled parasite ova (o), many motile and nonmotile bacteria (b), and many leukocytes (l); (B) dry mount with Dif-Quik stain, revealing many macrophages (m) and bacilli (b); (C) dry mount with Gram stain, revealing many positive-staining bacilli in blue; and (D) dry mount with acid-fast stain, showing numerous positive bacilli in red. [Figure available online in color.]

many enlarged vacuoles were visible in hepatocytes, and di- Three of the ﬁve gas bladder exudate samples demonstrated genetic trematode adults, many eggs, leukocytes, and bacterial positive growth, visualized as strongly orange-colored colonies, rods were present in gas bladder exudate. on Lowenstein–Jensen medium at both incubation temperatures. These samples were submitted to the National Jewish Medical and Research Center (Denver, Colorado) for identiﬁcation by high-performance liquid chromatography and 16S ribosomal MICROBIOLOGY

Downloaded by [Department Of Fisheries] at 23:30 28 February 2013 DNA sequencing. One sample was identified as Gordonia sp. We sampled gas bladder lumen exudate from all five Lined ◦ (incubated at 25 C), and two others were identified as Mycobac- Seahorses (Table 1). Samples from three animals were cultured ◦ terium poriferae (one sample identified from a culture incubated on marine agar plates for incubation at 25 C. Samples from four ◦ ◦ at 25 C and another sample identified from cultures incubated fish were cultured on blood agar plates for incubation at 25 C. ◦ ◦ at 25 C and 37 C). Samples from all five animals were cultured on two Lowenstein– Jensen slants each—one for incubation at 25◦C and one for incubation at 37◦C. HISTOPATHOLOGY Positive growth was observed on two of three blood agar Organs of two animals (one fish that tested positive for M. cultures from posterior kidneys, but these cultures were not poriferae infection and another fish that tested positive for Gor- submitted for identification. Positive growth was also observed donia sp. infection; Table 1) were submitted to the histopathol- on three of the four blood agar cultures from gas bladder exu- ogy laboratory at the College of Veterinary Medicine, University dates. One culture was identified as Pasteurella multocida and of Florida, Gainesville, for histopathological preparation with a second was identified as Staphylococcus sp. by All Florida hematoxylin and eosin, acid-fast, Gomori’s methenamine silver Veterinary Laboratories (Archer, Florida); the third culture was (GMS), and periodic acid–Schiff stains. Prepared slides were not submitted for identification. then submitted to Wildlife Diagnostics (Kissimmee, Florida) for COMMUNICATION 47

FIGURE 4. Internal organs of (A) normal and (B) affected Lined Seahorses. The normal specimen exhibits a transparent, gas-ﬁlled gas bladder. In contrast, the affected specimen exhibits an enlarged, hardened, opaque gas bladder; this specimen’s pale, friable liver is also visible in the lower-left quadrant of the coelom (arrow). [Figure available online in color.]

evaluation. Histopathological findings for both animals revealed Gaevskaya 2004; Mazzi 2004), and digenean metacercariae a markedly distorted gas bladder: the gas bladder wall was mul- have been observed in other Lined Seahorses (L. Croft, tifocally lost or effaced by a dense, histiocytic, inflammatory Sea World Orlando, personal communication; our personal infiltrate intermixed with abundant necrotic cellular debris that observations). Of particular concern is the prevalence of extended into the adjacent coelom and adjacent to organs such mycobacteriosis in aquarium-housed syngnathids, usually as the liver and kidney (Figure 6). The inflammation was com- caused by Mycobacterium marinum, M. chelonae, and M. posed primarily of macrophages containing intracytoplasmic fortuitum (Greenwell 2002). This is the first published case Downloaded by [Department Of Fisheries] at 23:30 28 February 2013 acid-fast-positive bacteria, with lesser numbers of heterophils report of D. virens infection of the gas bladder in a syngnathid, and lymphocytes. Dense, radiating colonies of bacterial the first published report of mycobacteriosis caused by M. bacilli were common within the inflammation, as were small, poriferae in a fish, and the first published report of Gordonia thick-walled parasite eggs (Figure 7A). Internal structures of sp. infection in a nonhuman animal. the eggs stained positive with periodic acid–Schiff. Bacteria stained acid-fast and GMS positive (Figure 7B). Examination Dictysarca virens of the renal interstitium revealed moderate, diffuse hemoblast Dictysarca virens was first reported in the gas bladders of or hemocyte hyperplasia (Figure 6). morays Gymnothorax spp. (Linton 1910; Manter 1947). The presentation that Linton (1910) and Manter (1947) described is DISCUSSION virtually identical to our findings, with D. virens being charac- Among aquarists and aquatic animal veterinarians and terized by large size (up to 10 mm long in Manter’s collections), pathologists, syngnathids are generally known for their opaque white testes and bright yellow-orange ovaries, and large susceptibility to infectious diseases (S. Frasca, University of caeca filled with finely granular material. The life cycles of Connecticut, personal communication). Digenean infestation most hemiurid trematodes are not well known; in the case of of syngnathids is occasionally encountered (Korniychuk and Dictysarca, the life cycle is unknown. The hemiuroid trematode 48 ANDERSON AND PETTY

in partially evaporated seawater. For physostomous fishes (e.g., morays Gymnothorax spp.; Tesch 2003), this observation is consistent with release of eggs into the environment via the pneumatic duct and subsequently the esophagus, where they may go on to hatch and continue the life cycle (as in I. hypselobagri). However, for physoclistous fishes (e.g., advanced teleosts, such as Hippocampus spp.; Helfman et al. 1997), which do not possess a duct for transport from the lumen of the gas bladder to the external environment, the requirement for seawater exposure to induce hatching of digenean ova makes it unlikely that the eggs would hatch inside the body of the fish, which has a markedly lower salinity than environmental seawater. This suggests (1) that the Lined Seahorse is an accidental host of D. virens,(2) that the death and subsequent decomposition of the Lined Sea- horse host are necessary to liberate eggs into the environment for hatching and subsequent life cycle development, or (3) both. The numerous cases that we received from this single aquaculture facility prompted us to consider the possibility that the digenean life cycle was perpetuating in the aquaculture system. The aquaculture system from which these animals were obtained was species specific for Lined Seahorses. No other fish species or invertebrates were intentionally kept, and there were no mollusks present in the system. However, the research collection was fed live Artemia sp. daily, which presents a potential vector of disease. The researcher also noted that numerous harpacticoid copepods were present in the aquaria, and these were presum- ably brought in with the Artemia sp. We collected a sample of 2,616 copepods from the system and examined them for metacercariae, but we found none. The multispecies systems in which these host specimens were held before they were acquired by the aquaculture research facility may have enabled proliferation of D. virens because of the presumed presence of intermediate hosts required for completion of the digenean life cycle. Alter- natively, infection of these Lined Seahorses may have originated from the wild and occurred prior to the fish’s capture. Because of the phenomenon of co-infection (see Co-infection below), it is difficult to tease apart the host’s response to each FIGURE 5. Photomicrograph of the digenean Dictysarca virens. [Figure avail- infective agent. For digenean metacercarial infestations of syng- able online in color.] nathids, Mazzi (2004) noted the formation of “pigmented cysts,” Downloaded by [Department Of Fisheries] at 23:30 28 February 2013 which were likely granulomas, in the muscle and skin of Broad- Isoparorchis hypselobagri exhibits a life cycle consistent with Nosed Pipefish Syngnathus typhle. In digenean (usually metac- our observations of D. virens; adult I. hypselobagri parasitize the ercarial) infestations of other fish species, granuloma formation gas bladders of freshwater physostomous siluroid teleosts (Rai is commonly observed (Hunter and Hamilton 1941). The eggs and Pande 1965; Mahajan et al. 1978), and eggs are believed to of many aporocotylids elicit a variety of responses depending exit the gas bladder via the pneumatic duct (Cribb 1988). An- upon the location within the host and the species of host that other strategy of adult trematodes that parasitize fish includes the is parasitized (Bullard and Overstreet 2008). The eggs are re- strategy employed by blood flukes of the family Aporocotylidae, ported to cause infarcts when they lodge in blood vessels. When whose eggs hatch in the gills after being trapped there (Bullard found in organs, the eggs are often encapsulated with a mix and Overstreet 2008). of macrophages, neutrophils, and eosinophilic granular cells. We question the fate of the numerous eggs found in the gas In gills, the response varies from none to extensive, depending bladders of our specimens. Digenean adults and eggs were only on the host species. Our observations describe a histiocytic re- found within the confines of the Lined Seahorses’ gas blad- sponse, which we presume to be largely in response to bacterial ders. Manter (1947) reported that the eggs of D. virens hatched co-infection (see next section). COMMUNICATION 49

FIGURE 6. Photomicrograph of a histological section of the posterior kidney and gas bladder wall of an affected Lined Seahorse (hematoxylin and eosin stain). The gas bladder wall (w) is effaced by inﬂammatory inﬁltrate and abundant necrotic cellular debris. The renal interstitium (i) reveals moderate, diffuse hemoblast or hemocyte hyperplasia. [Figure available online in color.]

Mycobacterium poriferae and macrophages with granuloma formation (Lansdell et al. Mycobacteriosis is the most common chronic disease affect- 1993). Our presentations of mycobacteriosis are atypical in sev- ing aquarium fish (Noga 2000). It is generally characterized eral aspects. First, we did not find systemic mycobacteriosis in as a systemic disease affecting several organ systems. Host re- any of our cases. Histopathology showed that acid-fast-positive sponse is usually characterized by infiltration of lymphoid cells bacilli were always confined to the lumen and wall of the gas Downloaded by [Department Of Fisheries] at 23:30 28 February 2013

FIGURE 7. Photomicrographs of histological sections of the gas bladder lumen in an affected Lined Seahorse, revealing (A) a dense inﬂammatory inﬁltrate composed of many leukocytes (l) and many thick-walled parasite ova (o) (hematoxylin and eosin stain); and (B) dense, radiating bacterial colonies of positive- staining bacilli (in red; acid-fast stain). [Figure available online in color.] 50 ANDERSON AND PETTY

bladder and did not infiltrate any other organs. Second, while GMS-positive, and Gram-positive bacilli; rapid growth; growth we observed a histiocytic response characterized by enhanced at 37◦C (Kerr and Barrett 1994); and previous reports of M. leukocyte production in the posterior kidney and leukocyte in- fortuitum in syngnathids (Greenwell 2002). However, strongly filtration of the gas bladder lumen, granulomas were noticeably orange-colored growth on Lowenstein–Jensen medium (in- absent. The nongranulomatous response of syngnathids to my- dicative of scotochromogenic species) suggested that the bac- cobacteriosis has also been noted by other aquatic animal health terium was M. poriferae (Padgitt and Moshier 1987), and high- professionals (I. Berzins, The Florida Aquarium, personal com- performance liquid chromatography and 16S ribosomal DNA munication) and has been seen in some other infectious diseases analysis confirmed its identification as such. Clinicians should of fish, such as proliferative kidney disease (Feist and Bucke make note of M. poriferae and should consider this species in 1993; Chilmonczyk et al. 2002; Morris et al. 2003; Sterud et al. a differential diagnosis of mycobacteriosis in fish from which 2007; Bettge et al. 2009) and Edwardsiella ictaluri infection rapidly growing, acid-fast-positive bacilli have been isolated. (Miyazaki and Plumb 1985; Iwanowicz et al. 2006). Mycobacterium spp. that cause disease in fish are known Yanong et al. (2003) reported a similarly atypical presenta- to cause disease in humans as well. Kerr and Barrett (1994) tion of mycobacteriosis (M. marinum) in Striated Frogfish An- described M. marinum, M. fortuitum, and M. chelonae infec- tennarius striatus. Although infection was systemic with mul- tions in humans; these infections were induced by contact with tiple organ involvement, there was no granulomatous response; infected organisms or aqueous environments or were postop- however, there was a histiocytic response. Yanong et al. (2003) erative infections arising from improperly sterilized surgical speculated that (1) the infected fish may have been immunocom- tools. In most people, these nontuberculous mycobacterial in- promised because of stress caused by aquarium conditions, (2) fections are usually localized to the area of contact and often the family or genus to which Striated Frogfish belong may differ involve granulomatous response. There is a greater chance of from other fish taxa in the immunologic response to mycobacte- disseminated disease with species that can grow at 37◦C (hu- rial organisms, (3) the strain of M. marinum involved may differ man body temperature) and especially in immunocompromised from other strains in terms of antigenicity or pathogenicity, or patients. Because of the potential for zoonosis, workers that are (4) the infection may have been acute but lethal, leading quickly in contact with fish in aquaculture or aquarium settings should to disease and subsequent mortality before typical lesions devel- employ biosecurity protocols to minimize human exposure to oped. In comparing these hypotheses with our case, we cannot Mycobacterium spp. and also to minimize their spread among rule out stress-induced immunosuppression caused by aquar- tanks. Jacobs et al. (2004) and Mainous and Smith (2005) ad- ium conditions, as the submitted animals also came from an dressed prevention and disinfection of Mycobacterium spp. in aquarium environment. Phylogenetically, frogfishes (Paracan- aquaculture. thopterygii) are far removed from seahorses (Acanthopterygii), so if the nongranulomatous immune response is taxon specific, Gordonia sp. this phenotype must have evolved at least twice on the phy- Along with Mycobacterium spp., Gordonia sp. is another logenetic tree. We argue against strain-specific effects, as this member of the actinomycetes, and its presence in one of response has been seen in syngnathids against M. poriferae (this the affected individuals is noteworthy. The genus name of paper) as well as against M. marinum and M. chelonae (Berzins, Gordonia has only recently been implemented (in 1997) to personal communication). A quick onset of disease leading to describe species that were previously assigned to the genus mortality before a granulomatous response can occur is possible Rhodococcus and later to the genus Gordona (Arenskotter¨ et al. in our case. The acute nature of the mycobacteriosis is supported 2004). The genus Gordonia represents a group of bacteria that by infection being limited to the gas bladder. Furthermore, M. are also distributed ubiquitously in nature and in industrial Downloaded by [Department Of Fisheries] at 23:30 28 February 2013 poriferae is a rapidly growing species. We witnessed growth or polluted environments; for example, these bacteria can be on Lowenstein–Jensen medium by 72 h at 37◦C and by day 10 found in soil (mangrove rhizosphere), oil-producing wells, at 25◦C (growth may have been evident before day 10, but we hydrocarbon-contaminated soil, and wastewater treatment did not check after 72 h). This is consistent with Padgitt and bioreactors or biofilters (Arenskotter¨ et al. 2004). This genus Moshier’s (1987) observations; they witnessed growth of M. is of interest to biotechnology; Gordonia bacteria catalyze poriferae in 4 d on Lowenstein–Jensen medium at 28–30◦C. a range of enzymatic reactions that is rarely (if ever) seen This is the first report of M. poriferae infection in a fish. My- in other organisms, and they have potential applications cobacterium poriferae was first isolated from a marine sponge in bioremediation, biodegradation, biotransformation, and and was characterized by Padgitt and Moshier (1987). It had production of biosurfactants (Arenskotter¨ et al. 2004). In recent since been reported in granulomatous lesions of Chevron Snake- years, Gordonia spp. have also been implicated in human heads Channa striata that were fed marine offal (Tortoli et al. infections, although the requirement of genomic sequencing for 1996), but the isolate was later identified as M. fortuitum (Putti- identification has limited recognition of this genus in human naowarat et al. 2002). We originally suspected that our strain disease to the last 13 years. In humans, clinical manifestations of Mycobacterium was consistent with M. fortuitum based on are variable and include primary and secondary bacteremia, the following lines of evidence: long, thin, acid-fast-positive, soft tissue infection, and infective keratitis (Lai et al. 2010). COMMUNICATION 51

As with Mycobacterium spp., Gordonia infections in humans Laboratory of the Program in Fisheries and Aquatic Sciences, are commonly associated with underlying conditions that cause School of Forest Resources and Conservation, Institute of Food immunosuppression (Arenskotter¨ et al. 2004; Blaschke et al. and Agricultural Sciences, University of Florida, Gainesville. 2007; Gupta et al. 2010; Lai et al. 2010). In humans, Gordonia P. Anderson was supported by the University of Florida Pro- infections respond to antibiotic therapy (Blaschke et al. 2007; gram in Fisheries and Aquatic Sciences Alumni Fellowship. Aoyama et al. 2009; Jannat-Khah et al. 2009; Gupta et al. 2010; Lai et al. 2010). To our knowledge, this is the first report of a REFERENCES Gordonia sp. infection in a nonhuman animal. Aoyama, K., Y. Kang, K. Yazawa, T. Gonoi, K. Kamei, and Y. Mikami. 2009. Characterization of clinical isolates of Gordonia species in Japanese clinical Co-infection samples during 1998–2008. Mycopathologia 168:175–183. We note the prevalence of co-infection in these specimens. Arenskotter,¨ M., D. Broker,¨ and A. Steinbuchel.¨ 2004. Biology of the metabol- Mycobacterium was isolated in two of the five cases submitted; ically diverse genus Gordonia. Applied and Environmental Microbiology Gordonia sp. was isolated in one case. In other cases, Pasteurella 70:3195–3204. Bettge, K., T. Wahli, H. Segner, and H. Schmidt-Posthaus. 2009. Proliferative multocida and Staphylococcus sp. were isolated, but these were kidney disease in Rainbow Trout: time- and temperature-related renal pathol- suspected contaminants. Other bacteria species may have been ogy and parasite distribution. Diseases of Aquatic Organisms 83:67–76. present, as suggested by growth on blood agar plates, but some Blaschke, A. J., J. Bender, C. L. Byington, K. Korgenski, J. Daly, C. A. Petti, of these cultures were not submitted for identification. In all A. T. Pavia, and K. Ampofo. 2007. Gordonia species: emerging pathogens in five cases, D. virens was present. In considering the prevalence pediatric patients that are identified by 16S ribosomal RNA gene sequencing. Clinical Infectious Diseases 45:483–486. of bacterial co-infection with the presence of this gas bladder Bullard, S. A., and R. M. Overstreet. 2008. Digeneans as enemies of fishes. parasite, we hypothesize that digeneans may harbor bacteria Pages 817–976 in J. C. Eiras, H. Segner, T. Wahli, and G. B. Kapoor, editors. within their caeca and transport them to new hosts. Alternatively, Fish diseases, volume 2. Science Publishers, Enfield, New Hampshire. the lesions imposed by D. virens may have rendered the host Chilmonczyk, S., D. Monge, and P. de Kinkelin. 2002. Proliferative kidney susceptible to secondary bacterial infection. disease: cellular aspects of the Rainbow Trout, Oncorhynchus mykiss (Wal- baum), response to parasitic infection. Journal of Fish Diseases 25:217–226. Cribb, T. H. 1988. Two new digenetic trematodes from Australian freshwater Treatment fishes with notes on previously described species. Journal of Natural History The collection at the aquaculture research laboratory where 22:27–43. the affected specimens were held would be benefited by treat- Feist, S. W., and D. Bucke. 1993. Proliferative kidney disease in wild salmonids. ment with praziquantel to eradicate digeneans. However, the Fisheries Research 17:51–58. researcher at that facility noted problems with dissolution of Greenwell, M. G. 2002. Syngnathid health management. Pages 15–24 in C. D. Bull and J. S. Mitchell, editors. Seahorse husbandry in public aquaria. Project praziquantel into saltwater and with the drug’s toxicity. Be- Seahorse, University of British Columbia, Vancouver. cause most digeneans exhibit an indirect life cycle that requires Gupta, M., D. Prasad, H. S. Khara, and D. Alcid. 2010. A rubber-degrading multiple hosts, the identification and eradication of intermediate organism growing from a human body. International Journal of Infectious host vectors should be considered as a management option. Diseases 14:e75–e76. Currently, there is no effective treatment against mycobacte- Helfman, G. S., B. B. Collette, and D. E. Facey. 1997. The diversity of fishes. Blackwell Scientific Publications, Malden, Massachusetts. riosis in fishes. Depopulation and disinfection constitute the only Hunter, G. W., I, and J. M. Hamilton. 1941. Studies on host–parasite reactions effective method to eradicate Mycobacterium spp. from aquacul- to larval parasites: IV. the cyst of Uvulifer ambloplitis (Hughes). Transactions ture systems. However, chronic exposure to mycobacteria can of the American Microscopical Society 60:498–507. be minimized by utilizing ultraviolet light sterilization in filter Iwanowicz, L. R., A. R. Griffin, D. D. Cartwright, and V. S. Blazer. 2006. Mor- design (minimum recommended exposure of 10 µW·cm−2·s−1; tality and pathology in Brown Bullheads Ameiurus nebulosus associated with a spontaneous Edwardsiella ictaluri outbreak under tank culture conditions.

Downloaded by [Department Of Fisheries] at 23:30 28 February 2013 Jacobs et al. 2004) and by using effective disinfectants for nets, Diseases of Aquatic Organisms 70:219–225. brushes, and other equipment (Mainous and Smith 2005). To Jacobs, J. M., A. Lazur, and A. Baya. 2004. Prevention and disinfection of date, no guidelines have been established for the prevention or Mycobacterium sp. in aquaculture. Maryland Sea Grant Extension, Finfish management of Gordonia sp. in aquaculture systems. Worksheet 9, Publication UM-SG-SGEP-2004-01, College Park. Jannat-Khah, D. P., E. S. Halsey, B. A. Lasker, A. G. Steigerwalt, H. P. Hin- rikson, and J. M. Brown. 2009. Gordonia araii infection associated with an ACKNOWLEDGMENTS orthopedic device and review of the literature on medical device-associated We thank A. Dove (Stony Brook University) for digenean Gordonia infections. Journal of Clinical Microbiology 47:499–502. Kerr, J. H., and T. L. Barrett. 1994. Atypical mycobacterial diseases. Pages identification; T. Crosby, R. Francis-Floyd, and J. Holloway 391–422 in W. D. James, editor. Military dermatology. Walter Reed Army (University of Florida) for assistance with diagnostic work; the Medical Center Borden Institute, Washington, D.C. Diagnostic Imaging staff of the College of Veterinary Medicine, Korniychuk, Y. M., and A. V. Gaevskaya. 2004. The first record of Aphallus University of Florida, for radiography; and P. Perkins (The tubarium (Trematoda, Cryptogonimiidae) in the Black Sea. Vestnik Zoologii Florida Aquarium) for assistance with figure preparation. Hus- 38:79–80. Lai, C. C., C. Y. Wang, C. Y. Liu, C. K. Tan, S. H. Lin, C. H. Liao, C. H. Chou, bandry protocols were authorized by the University of Florida Y. T. Huang, H. I. Lin, and P. R. Hsueh. 2010. Infections caused by Gordonia Institutional Animal Care and Use Committee Protocol Num- species at a medical centre in Taiwan, 1997 to 2008. Clinical Microbiology ber D-432. This work was funded by the Fish Health Extension and Infection 16:1448–1453. 52 ANDERSON AND PETTY

Lansdell, W., B. Dixon, N. Smith, and L. Benjamin. 1993. Isolation of several rine sponge. International Journal of Systematic Bacteriology 37:186– Mycobacterium species from fish. Journal of Aquatic Animal Health 5:73–76. 191. Linton, E. 1910. Helminth fauna of the Dry Tortugas: II. trematodes. Carnegie Puttinaowarat, S., K. D. Thompson, A. Kolk, and A. Adams. 2002. Identification Institution of Washington Publication 4:11–98. of Mycobacterium spp. isolated from Snakehead, Channa striata (Fowler), Mahajan, C. L., N. K. Agrawal, M. J. John, and V.P.Katta. 1978. Parasitization of and Siamese Fighting Fish, Betta splendens (Regan), using polymerase chain Isoparorchis hypselobagri Billet in Channa punctatus Bloch. Current Science reaction–reverse cross blot hybridization (PCR-RCBH). Journal of Fish Dis- 47:835–836. eases 25:235–243. Mainous, M. E., and S. A. Smith. 2005. Efficacy of common disinfectants against Rai, P., and B. P.Pande. 1965. Isoparorchid infection in some hitherto unrecorded Mycobacterium marinum. Journal of Aquatic Animal Health 17:284–288. fish species. Current Science 34:586–588. Manter, H. W. 1947. The digenetic trematodes of marine fishes of Tortugas, Sterud, E., T. Forseth, O. Ugedal, T. T. Poppe, A. Jørgensen, T. Bruheim, H. P. Florida. American Midland Naturalist 38:257–416. Fjeldstad, and T. A. Mo. 2007. Severe mortality in wild Atlantic Salmon Salmo Mazzi, D. 2004. Parasites make male pipefish careless. Journal of Evolutionary salar due to proliferative kidney disease (PKD) caused by Tetracapsuloides Biology 17:519–527. bryosalmonae (Myxozoa). Diseases of Aquatic Organisms 77:191–198. Miyazaki, T., and J. A. Plumb. 1985. Histopathology of Edwardsiella ictaluri in Tesch, F. W. 2003. The eel, 3rd edition. Translated from the German by R. J. Channel Catfish, Ictalurus punctatus (Rafinesque). Journal of Fish Diseases White, edited by J. E. Thorpe. Blackwell Scientific Publications, Ames, Iowa. 8:389–392. Tortoli, E., A. Bartoloni, E. Bozzetta, C. Burrini, C. Lacchini, A. Mantella, V. Morris, D. J., M. Longshaw, and A. Adams. 2003. California Golden Trout On- Penati, M. Tullia Simonetti, and C. Ghittino. 1996. Identification of the newly corhynchus mykiss aguabonita are susceptible to proliferative kidney disease. described Mycobacterium poriferae from tuberculous lesions of Snakehead Journal of Aquatic Animal Health 15:184–187. Fish (Channa striatus). Comparative Immunology, Microbiology and Infec- Noga, E. J. 2000. Fish disease: diagnosis and treatment, 2nd edition. Iowa State tious Diseases 19:25–29. University Press, Ames. Yanong, R. P. E., E. W. Curtis, S. P. Terrell, and G. Case. 2003. Atypical presen- Padgitt, P. J., and S. E. Moshier. 1987. Mycobacterium poriferae sp. tation of mycobacteriosis in a collection of frogfish (Antennarius striatus). nov., a scotochromogenic, rapidly growing species isolated from a ma- Journal of Zoo and Wildlife Medicine 34:400–407. Downloaded by [Department Of Fisheries] at 23:30 28 February 2013 This article was downloaded by: [Department Of Fisheries] On: 28 February 2013, At: 23:33 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Aquatic Animal Health Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/uahh20 Effects of Aquarium-Related Stressors on the Zebrafish: A Comparison of Behavioral, Physiological, and Biochemical Indicators David Gronquist a & John A. Berges a a Department of Biological Sciences, University of Wisconsin–Milwaukee, Lapham Hall, 3209 North Maryland Avenue, Milwaukee, Wisconsin, 53211, USA Version of record first published: 22 Jan 2013.

To cite this article: David Gronquist & John A. Berges (2013): Effects of Aquarium-Related Stressors on the Zebrafish: A Comparison of Behavioral, Physiological, and Biochemical Indicators, Journal of Aquatic Animal Health, 25:1, 53-65 To link to this article: http://dx.doi.org/10.1080/08997659.2012.747450

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material. Journal of Aquatic Animal Health 25:53–65, 2013 C American Fisheries Society 2013 ISSN: 0899-7659 print / 1548-8667 online DOI: 10.1080/08997659.2012.747450

ARTICLE

Effects of Aquarium-Related Stressors on the Zebraﬁsh: A Comparison of Behavioral, Physiological, and Biochemical Indicators

David Gronquist and John A. Berges* Department of Biological Sciences, University of Wisconsin–Milwaukee, Lapham Hall, 3209 North Maryland Avenue, Milwaukee, Wisconsin 53211, USA

Abstract Fishes in aquaria and aquaculture settings may experience a variety of stressors including crowding, different lighting, periods of food deprivation, and vibrations from sources including pumps and tapping of tank sides. The effects of such low-level chronic stress are poorly explored. We used replicate sets of six Zebrafish Danio rerio in four series of experiments to compare the effects of (1) stocking densities ranging from 0.13 to 1.2 fish/L, (2) cool white (6,500 K), warm white (4,100 K), and ultraviolet-enhanced (420 actinic) fluorescent lighting, (3) food deprivation for up to 9 d, and (4) random mechanical tapping on the tank side sufficient to induce a startle response on specific behaviors (fin display, body fluttering, aggression, mouth gaping, and chattering), dissolved cortisol released into aquarium water (collected on a chromatography column and analyzed with an immunoassay), and heat-shock proteins (HSPs 27, 40, 60, and 70) detected immunochemically in western blots of muscle tissue. Of all the treatments, only food deprivation resulted in significant differences between control and treatment fish; dissolved cortisol declined after 120 h of starvation and HSP40 and HSP60 in muscle tissue increased significantly after 216 h. High variability in behaviors and HSP measurements was noted within all controls and treatments, suggesting that effects of treatments were experienced unequally by individuals within a treatment. Social stressors resulting from dominance hierarchies may play a critical role in modifying the effects of aquarium and aquaculture stressors on captive fish.

Fish on display in public or in home aquaria, retail pet stores, stress occur at many different levels including changes in and aquaculture production systems may be subject to a number stress hormones, metabolic changes such as an increase of the of environmental factors known to induce stress, including high amount of glucose and a decrease of tissue glycogen, cellular stocking densities (Haukenes and Barton 2004; Ramsay et al. changes such as heat-shock protein production, changes in Downloaded by [Department Of Fisheries] at 23:33 28 February 2013 2006), various types of artificial lighting (Head and Malison immune function, hematological features, and osmoregulatory 2000; Karakatsouli et al. 2010), periods of food deprivation disturbance, as well as tertiary responses in behavior, growth, (Cara et al. 2005; Weber and Bosworth 2005), and varying swimming, and disease resistance (Barton 2002; Iwama et al. levels of noise (Smith et al. 2004; Wysocki et al. 2004). However, 2006). Considerable emphasis has been placed on diagnosing despite recognition of these issues in the industry (e.g., Bartelme stress using changes in cortisol levels in the fish (Iwama 2004), relatively few systematic studies have been carried out et al. 2006), while understanding behavioral responses in using chronic low-level stressful conditions rather than acute aquaculture has been increasingly favored for assessing stress conditions (e.g., Ashley 2007). and fish welfare (Ashley 2007). However, given the multiple Stress is most generally defined in terms of effects on levels of response, a comparison of measurements made at organism function to the point where chances of survival different levels seems a valuable approach (e.g., Anderson et al. are reduced (Brett 1958; Barton 1997). Fish responses to 2011).

*Corresponding author: [email protected] Received March 14, 2012; accepted November 4, 2012 Published online January 22, 2013 53 54 GRONQUIST AND BERGES

Direct observation of fish behavior is a relatively straightfor- In this study, we examined four potential stressors that would ward and noninvasive means to assess the effect of a stressor. be commonly experienced by fish in aquaria: different stocking Fish in aquaria generally respond to disturbances by altering fre- densities, variations in tank lighting spectrum, deprivation of quency of behaviors such as erratic movements, chasing other food, and effects of sound (pressure) waves created by tapping fish, nipping at fins, different fin displays, or gaping of the mouth on the side of the tank. Three responses were measured: fish (e.g., Chervova 1997; Haukenes and Barton 2004). behavior, cortisol released into aquarium water, and levels of As a means of quantifying the stress response, levels of four HSP proteins in muscle tissue. cortisol in the bloodstream are commonly accepted as reliable Fishes in display aquaria and aquaculture settings are of- (Wendelaar Bonga 1997; Scott et al. 2001; Ramsay et al. 2006), ten stocked at higher densities than found in nature (Bartelme though the response can be transient, so care needs to be taken 2004). The aquaculture literature is rich with studies of stock- in interpretation (e.g., Wendelaar Bonga 1997; Davis and Small ing density for optimized growth, and some of these have also 2006). However, such a measurement is highly invasive and assessed stress measures (see Ellis et al. 2002; Aksakal et al. sampling can induce stress itself. Scott and Sorensen (1994) 2011; Castranova et al. 2011). Based on results from this lit- developed the alternative of measuring steroids excreted into erature, we hypothesized that increasing stocking densities in the water; subsequent work has demonstrated good correlations Zebrafish would result in alterations in behavior, increased cor- between steroid concentration in the water and in plasma (Scott tisol release, and elevated tissue HSPs. et al. 2001), and the method has now been used in several species Lighting spectra for aquaria differ from that of natural (Scott and Ellis 2007). Transient responses in cortisol remain a environments. In particular, display aquaria lighting can be concern for interpretation, however. enhanced at the ultraviolet (UV) end of the spectrum. Zebrafish Measurements of tissue proteins may provide a longer-term are highly visual fish and have sensitivity to light in the UV and more integrative measure of stress. One group of proteins range (Robinson et al. 1993). We hypothesized that there would that has received particular attention are the so-called heat-shock be no differences in responses between fish in aquaria lighted proteins (HSPs), a set of highly conserved proteins defined by with typical “cool white” spectrum fluorescent indoor bulbs their molecular masses (e.g., HSP70s have molecular masses (color temperature, 4,100 K) and enhanced daylight-spectrum near 70 KDa) that have diverse functions and are expressed at lamps (color temperature, 6,500 K), but that UV-enhanced increased levels under stress (e.g., Ackerman et al. 2000; Basu bulbs (color temperature, 7,100 K) would result in changes in et al. 2002; Iwama et al. 2004). Measurements of HSPs are behaviors and elevated cortisol and HSPs. most commonly made at the level of transcription (i.e., messen- Fishes in aquaria may be fed to maximize growth rate ger RNAs) or at the protein level (immunochemically). These (Jobling 1981) or minimize aquarium maintenance (Ashley methods are necessarily invasive and typically involve dissection 2007); it seems likely that in home aquaria, delayed and missed of the organism. There are concerns about the interpretability of feedings are common. Food deprivation is correlated with stress tissue levels of HSPs with respect to stress (Iwama et al. 1999, responses in the literature (e.g., Brann¨ as¨ et al. 2003; Bayir et al. 2004), but they continue to be widely used. 2011); thus, we hypothesized that depriving Zebrafish of food Surprisingly few studies have attempted to monitor stress at over a 10-d period would result in changes in behavior and an more than one or two different levels of biological organization increase in cortisol levels in water and HSPs in tissues with and to compare the responses. In the present study, we chose longer deprivation. to monitor behavior, a physiological response (cortisol release), Finally, to judge from the prevalence of signs in pet shops and a biochemical response (tissue HSP levels). asking customers to refrain from touching aquaria, it seems Zebrafish Danio rerio is a species that represents a conve- likely that fishes are subject to the intermittent low-frequency Downloaded by [Department Of Fisheries] at 23:33 28 February 2013 nient model fish system for studying aquarium and aquaculture vibrations that “tapping” on aquarium glass would produce (cf., stress, is easy to rear and work with experimentally, and Leong et al. 2009). The issue of acoustic and vibrational stress has a well-established biochemical and genetic background. due to factors like pump noise and tank echoes is well appreci- Zebrafish are kept in home and public aquaria worldwide ated in aquaculture (Craven et al. 2009), though there are also and there is a well-developed commercial trade for them positive aspects to sound (Papoutsoglou et al. 2010). Zebrafish (Westerfield 1993; Lawrence 2007). The commercial stocks do exhibit clear startle responses to acoustic stimuli (Colwill originate from fish collected in rice fields and quiet streams et al. 2011) and we hypothesized that fish experiencing tap- in the Ganges River watershed of India and Bangladesh ping on aquarium glass would exhibit changes in behavior and (Lawrence 2007). Although relatively little is known about elevated cortisol and HSPs relative to those that did not. their ecology and behavior in nature (Spence et al. 2008), Zebrafish have been used extensively in fields ranging from METHODS toxicology and biochemistry to medicine, genetics, and even social behavior (Ton et al. 2002; Airaksinen et al. 2003; Fish and General Maintenance Bosworth et al. 2005; Chizinski et al. 2008; Saverino and Gerlai Adult Zebrafish (90–120 d old) were obtained from a 2008). commercial supplier (Fish2U.com, Gibsonton, Florida). To AQUARIUM-RELATED STRESSORS ON THE ZEBRAFISH 55

minimize fungal infections or parasites, the fish were treated were the same in all cases (12 h light : 12 h dark). Hanging di- with Paracide-F (Argent Chemical Laboratories, Redmond, viders of thick black felt were used to isolate treatments. Series Washington) for 1 h at 175 mg/L for three consecutive days A used enhanced daylight-spectrum lighting (813152, Philips, (Francis-Floyd 1996). Zebrafish were maintained in a 450-L Philips Electronics, Markham, Ontario; color temperature of aquarium under a 12 h light : 12 h dark photoperiod in a 6,500 K), series B used lamps more strongly weighted to a flow-through system with dechlorinated water at 28◦C and fed shorter wavelength, typical of those used in display aquaria to Grower BioDiet fish food (Bio-Oregon, Warrenton, Oregon) enhance fish colors (TL 40W/03RS, Philips, Philips Electron- daily until satiation. Fish were held for a minimum of 3 weeks ics; color temperature of 7,100 K, 420 actinic), and series C had after the Paracide-F treatment before use in experiments and cool white fluorescent lighting, typical of many public build- were used only once for each experiment. ings (813157, Philips, Philips Electronics; (color temperature of 4,100 K). Spectra of the lamps used, in terms of relative Experimental Design radiant power, are shown in Figure 2. In all cases, quantum Four different series of experiments (manipulating stocking irradiance was measured (Li-Cor LI-250A witha2pisensor, density, tank-lighting spectra, time of food deprivation, and tap- LI-COR Biosciences, Lincoln, Nebraska) and adjusted to an − − ping), each lasting 10 d, were conducted sequentially using sim- average of 210 µmol quanta·m 2·s 1 for all treatments. ilar experimental designs. In each series, sets of six Zebrafish Food deprivation.—To examine the effects of withholding were randomly assigned to each of nine 45-L aquaria, which food, Zebrafish were acclimated to the experimental tank for were divided into three treatments (sets A, B, and C; Figure 1). 48 h on the same feeding regime as in the common holding Flow into each aquarium was adjusted to 3.0 L/min. Fish were tank, and then feeding was suspended. Series A were sampled acclimated in aquaria for 48 h before the experiments began. after 24 h without food, series B after 72 h, and series C after Stocking density.—To examine the effects of stocking den- 216 h without food. sity, water volumes in which Zebrafish were free to move were Tapping.—To examine the effects of percussion on the restricted using fine-mesh tank dividers (cat. no. 10600, Lee’s tank wall, “tapping” devices (Figure 3) were designed and Aquarium and Pet Products, San Marcos, California) and aquar- constructed (T. Consi and K. Verhein, University of Wisconsin– ium gravel at the base of the divider. Volumes included the full Milwaukee, unpublished). The computer-controlled, pro- tank (45 L for set C, for a density of 0.13 fish/L), approximately grammable devices consisted of a microcontroller chip one-half the tank (20 L for set B, 0.30 fish/L), and approximately (PIC18F1320-I/P-ND, Digi-key Corporation, Thief River Falls, one-eighth of the tank (5 L for set A, 1.2 fish/L). Tank dividers Minnesota), programmed through an RS232 serial converter were present in all tanks; they were placed against one wall of (MAX233CPP-ND, Digi-key Corporation), that controlled a the aquaria in the 45-L treatment. 12-V DC, pull-type solenoid (2137941, Jameco Electronics, Tank lighting.—To examine the effects of different light qual- Belmont, California) with a rubber stopper fixed to the end. ities, each group of three aquaria (series A, B, and C) was exposed to different lighting by suspending separate fluorescent fixtures (part no. 758485, Lithonia Lighting MWDC, Hanover Park, Illinois) with different tubes over each tank. Photoperiods Downloaded by [Department Of Fisheries] at 23:33 28 February 2013

FIGURE 1. General experimental design of experiments, consisting of nine 45-L aquaria in three series (A, B, and C), all fed from a common header FIGURE 2. Relative spectra of fluorescence lamps used in lighting experi- tank. Flow-through of water was controlled independently for each tank using a ments. These include enhanced daylight-spectrum lighting (color temperature, valve, and water level was maintained by a drain standpipe. Each tank contained 6,500 K), lamps strongly weighted to shorter wavelength to enhance fish colors six Zebrafish. Water temperature was maintained at 28◦C using aquarium im- (color temperature, 7,100 K, 420 actinic), and cool white fluorescent lighting, mersion heaters, and lighting was provided from above. Each tank series was typical of many public buildings (color temperature, 4,100 K). Data obtained isolated from each other using hanging dividers of heavy black felt. from http://www.usa.lighting.philips.com/connect/tools literature/ 56 GRONQUIST AND BERGES

individual fish. Statistical analyses were performed on these counts of behavior, as described below. Water samples for analysis of excreted cortisol were collected on days 5 and 10 of each experiment, except in the case of the feeding experiments where samples were collected at the end of the experiments for series A and B (24 and 72 h, respectively) and at two points (120 h and 216 h) for series C. One hour before collection of water, the main valve on the header tank was turned off to stop the flow of water and allow excreted cortisol to accu- mulate in each aquarium. During this time, water temperature was monitored and did not change by more than 0.2◦C. After 1 h, 2 L of water was collected from each tank by siphon. Cortisol in the water was extracted following Scott and Sorensen (1994) and Scott et al. (2001). Briefly, water samples were prefiltered through glass fiber filters (25 mm, type A/E, Pall, Ann Arbor, Michigan) to remove particulates and then pumped at a rate of 1 L/h using a peristaltic pump (Masterflex 7553-70, Cole-Parmer Instrument, Vernon Hills, Illinois) through September-Pak Plus C-18 cartridges (WAT020515, Waters, Milford, Massachusetts), which had been activated immediately before use with 5 mL of 100% methanol followed by 5 mL of distilled water. Cartridges FIGURE 3. Design of tank tapping devices (K. Verhein, University of were rinsed with 5 mL of distilled water and excess water was Wisconsin–Milwaukee, unpublished). The computer-controlled, programmable removed by forcing 5 mL of air though samples with a syringe. units consisted of a Microchip PIC18F1320 microcontroller, an RS232 serial Cartridges were stored at −70◦C for later analysis. converter (operated at 9,600 baud 8N1), and a 12-V DC pull-type solenoid (“tapper”) with a rubber stopper fixed to the tapping end. Each tapper was mounted In order to quantify cortisol recovery from aquarium water to the plastic frame of an aquarium and the distance adjusted to provide the four control aquaria were set up with additions of 0, 0.5, 5, minimum force needed to evoke a startle response in the Zebrafish. or 50 ng/L of cortisol solution (H6909, hydrocortisone, Sigma- Aldrich, St. Louis, Missouri). Water samples were taken and processed as described above. Power was provided by a regulated DC supply (PSR-10, Speco Tissue samples were taken on the final day of the exper- Technologies, Amityville, New York). The solenoid allowed a iments. Zebrafish were removed from their treatment tanks, spring-driven piston to strike the tank at specified intervals and rapidly anesthetized with an overdose of tricaine methanesul- frequency. The force of each “tap” was controlled by adjusting fonate (MS-222, Argent Chemical Laboratories; (>100 mg/L, distance from the tank to provide the minimum needed to neutralized with 1 M NaOH), flash-frozen in cryotubes in liquid induce a startle response in the fish. The tappers were controlled nitrogen, and stored at −70◦C for later analysis. using a computer and custom software written in BASIC (J. A. Berges, unpublished). Series A had 10-s bursts of 1 tap/s Cortisol Analysis delivered randomly for a total of 10 min in every 24 h. Series Cortisol analysis followed that of Scott and Sorensen (1994) B had 10-s bursts of 2 taps/s delivered randomly for a total of with some modifications. Cortisol was extracted from thawed Downloaded by [Department Of Fisheries] at 23:33 28 February 2013 20 min in 24 h. Tank series C received no tapping. Care was C18 cartridges in 4 mL of ethyl acetate, followed by a purge taken to ensure that vibrations were not transmitted among tank with 10 mL of air to ensure complete collection. Ethyl acetate series; we verified that taps on one series of tanks evoked no was evaporated by placing samples in a 45◦C water bath and detectible response in other tank series. drying under a stream of N2 gas. Dried samples were frozen (−70◦C) and accumulated for enzyme immunoassays (EIA). Observations and Sampling Cortisol was quantified using a Cortisol EIA kit based on Behavioral observations were made for 5-min periods a 96-well microplate (58212, Cayman Chemical, Ann Arbor, sequentially for each tank, midafternoon on each day of the Michigan) according to manufacturer’s directions and using the experiments (essentially the “sampling all occurrences of some kit standard. Dried samples (including recovery controls) were behaviors” described by Altmann 1974). Black felt dividers redissolved in 100 µL of the provided EIA buffer. Samples hung between tanks kept disturbance to a minimum. The most and standards were analyzed using a VERSAmax microplate frequent behaviors were identified based on initial observations reader (Molecular Devices, Sunnyvale, California) and Softmax during acclimation at the beginning of the experiment. During Pro software (version 4.8, Molecular Devices). Calculations of each 5-min period, each occurrence of each of the behaviors cortisol concentrations were performed as described in the kit was counted with no attempt made to assign behaviors to instructions. Where samples produced results that were out of AQUARIUM-RELATED STRESSORS ON THE ZEBRAFISH 57

range, a 100-fold dilution of the extracts were performed and Probing and detection followed guidelines in the Pierce Su- the samples rerun. persignal West Pico Chemiluminescent Substrate kit (Pierce). Membranes were blocked with 5% (w/v) skim milk powder in HSP Analysis tris-buffered saline (TBS) for 1 h. Each membrane was probed Frozen Zebrafish were thawed on ice and bilaterally dis- sequentially with rabbit polyclonal antibodies in the follow- sected, and muscle tissue was removed and placed into tared ing order: anti-HSP70 (spa-812), anti-HSP60 (spa-805), anti- microcentrifuge tubes. Extraction buffer, consisting of tris-SDS HSP40 (spa-400), and anti-HSP27 (spa-803) (Stressgen Biore- buffer (50 mM tris-HCl, 4% [w/v] sodium dodecyl sulfate, pH agent, Victoria, British Columbia). In order to ensure proper 7.5; Fisher Scientific, Fairlawn, New Jersey) with protease in- scaling of HSP signals, replicate samples were also blotted and hibitor (78415, HaltTM EDTA-free, Pierce Chemical, Rockford, probed with an anti-actin antibody (A 2066, Sigma Chemical, Illinois; 100 µL inhibitor to 10 mL buffer) was added in the pro- St. Louis, Missouri); this provided a specific index of muscle tis- portion 1 mL buffer : 0.2 g tissue. Samples were homogenized sue. Between antibodies, blots were stripped of immunoglobins on ice using a rotor-stator-type blender (PRO 200, Pro Scientific, (21062, Restore Western Blot Stripping Buffer, Pierce). In each Oxford, Connecticut) for three cycles of 20 s, then centrifuged case, single antibodies were diluted 1:2,000 with skim milk- for 2 min at 16,000 × g. From each supernatant, subsamples TBS block and incubated with membranes for 1 h. After wash- were aliquoted for electrophoresis and protein determinations ing, secondary antibodies (31460, HRP-conjugated goat anti- and frozen (−70◦C) for later analysis. rabbit immunoglobin, Pierce) were applied at 1:50,000 dilution Protein content of homogenates was measured using a BCA in skim milk-TBS block and incubated for 1 h. After a sec- Protein Assay Kit (23225, Pierce, Rockford, Illinois) and the ond washing series, membranes were rinsed in TBS and de- recommended 37◦C microplate protocol, and absorbance was tection performed using a chemiluminescent substrate (34080, corrected using blanks for the buffer. Homogenate samples Supersignal West Pico Chemiluminescent Substrate, Pierce) were prepared for electrophoresis following Greene et al. (1991) prepared according to manufacturer’s directions. Chemilumi- adding one volume of 0.2 M dithiothreitol and two volumes of nescence was detected by exposure to CL-Xposure Film (34090, 4% (w/v) SDS, 15% (v/v) glycerol, and 0.05% (w/v) bromoth- Pierce) and film was imaged and quantified using Kodak 1D ymol blue. Samples were boiled for 5 min and centrifuged at Digital Analysis system (version 4.0, Eastman Kodak). Area- 16,000 × g for 1 min. density values were determined for each band and scaled to con- For analyses, four fish from each treatment (out of a set of venient numerical values (0–50); because all treatments were three tanks × six fish = 18 fish exposed to each treatment) run on a single gel or blot, variation between exposures was were randomly selected for analysis. Homogenates were loaded minimized. Because antibody standards were derived from dif- on an equal protein basis (20 µg) alongside a blank, a molecular ferent species, no attempt was made to scale blots to abso- mass standard (product no. 26681, Blue Ranger Pre-stained lute protein or to compare quantity of protein among different Protein Molecular Weight Marker Mix, Pierce) and an HSP HSPs. standard on 15-well 10% Precise Protein gels (25241, Pierce). The HSP standard consisted of a mixture of 100 ng each of the following proteins: HSP27 (spp-715), HSP40 (spp-400), HSP60 Statistical Analyses (spp-741), and HSP70 (nsp-555), all obtained from Stressgen Analyses were carried out using SigmaPlot (version 12.0, Sy- Bioreagents, Victoria, British Columbia. Proteins were sepa- stat Software, San Jose, California). We used two approaches to rated by polyacrylamide gel electrophoresis in a HEPES-tris analyze results. First, using mean responses, results were com- buffer solution (recommended by the gel manufacturer) in a pared among treatments within each experimental series for Downloaded by [Department Of Fisheries] at 23:33 28 February 2013 Mini Protean 3 system (Bio-Rad Laboratories, Hercules, Cal- each parameter using ANOVA techniques, followed by Tukey’s ifornia). In each case, duplicate gels were run, one for staining tests to identify which treatments differed from each other. In and one for blotting. Gels were run at 130 V for approximately the case of behavioral data, where multiple measurements were 45 min or until tracking buffer reached the bottom of the gel. made over time, we used two-way, repeated-measures (RM) One of the gels was stained with Coomassie Blue R-250, ANOVA in which treatment and time were used as factors, and detained, and imaged using a Kodak 1D Digital Analysis sys- aquarium tanks as observation units (comparable to the model tem (version 4.0, Eastman Kodak, Rochester, New York). The used for ethological data in Anderson et al. 2011). Secondly, second gel was blotted onto 0.2-µm nitrocellulose membrane recognizing the importance of changing variance of measure- (162-0212, Bio-Rad Laboratories, Hercules, California) using a ments as indicators (Orlando and Guillette 2001), we tested for Mini Trans-blot cell (Bio-Rad Laboratories) and Towbin buffer heterogeneity of variance by calculating the absolute values of (Towbin et al. 1979) and blotting at 100 V for 60 min using a the difference of each individual measurement from the mean Powerpac 200 power supply (Bio-Rad Laboratories). To verify and running ANOVA on these data (as in Anderson et al. 2011). complete transfer, membranes were immersed in Ponceau S Where possible (i.e., where assumptions of normality and equal stain agitated on a rocker table for 5 min and then rinsed with variance were met), raw data were used; otherwise, data were distilled water until bands could be visualized. Membranes were transformed using square-root or common-logarithm transfor- allowed to dry overnight before being probed with antibodies. mations. In cases where transformed data still did not meet the 58 GRONQUIST AND BERGES

requirements for parametric tests, Kruskal–Wallis ANOVA on ranks was performed instead.

RESULTS Behavioral Measurements Five discrete behaviors were used in analyses after the ethogram was developed from initial observations (Table 1). Behaviors were consistent among experimental series and we did not observe novel behaviors under particular treatments.

Cortisol Measurements Using the column extraction method and EIA kit, cortisol was readily detected in the tank water; the effective detection limit of the method for aquarium water was approximately 10 pg/L. Based on tanks with known additions of cortisol, the recovery of cortisol was complete (i.e., 100%) within the range 0.5–50 ng/L, and no cortisol was detected in the water in the tank without additions (data not shown). Across all controls, cortisol in tank water averaged about 0.53 ng/L. Since this represented the amount accumulated in1hwithanaveragebiomass of 2.5 g/45-L tank, and since no cortisol was detected in the inﬂowing water (i.e., the tank without cortisol additions), we estimated an average rate of cortisol release of 9.6 ng·g−1·h−1. This rate was an overestimate, since it neglected residual cortisol, but for a 45-L tank with an inﬂow of 3.0 L/min, the error should be less than 10%.

Stocking Density Behaviors did not differ among stocking densities (two-way RM ANOVA: fin display, P > 0.9; flutter, P > 0.7; aggression, P > 0.4; gape, P > 0.1; chatter, P > 0.8; Figure 4A); there was no effect of time on behavior (P > 0.1 in all cases) and no > interaction between treatment and time (P 0.3 in all cases). FIGURE 4. Effects of three stocking densities of Zebrafish on (A) behavior, For each behavior, there was high variability among replicate (B) cortisol measured in aquarium water, and (C) HSPs measured in muscle tanks. Gaping was least displayed, while chattering was seen tissue. In each series of three 45-L aquaria, volumes were manipulated using most often overall. When analysis was carried out to examine mesh dividers resulting in six Zebrafish in 45 L (0.13 fish/L), 20 L (0.30 fish/L), ◦ heterogeneity of variance, no differences between treatment and or 5 L (1.2 fish/L). Aquaria were maintained at 28 C on a 12 h light : 12 h dark > photoperiod, and fish were fed once a day ad libitum. Average occurrence of time and no interactions were found (P 0.1 in all cases). behaviors is based on daily 5-min observations over a 10-d period; definitions Downloaded by [Department Of Fisheries] at 23:33 28 February 2013 of behaviors are given in the Methods and in Table 1. Cortisol was measured TABLE 1. Ethogram of Zebrafish observed in experiments. in 2-L water samples, concentrated using C-18 cartridges, and assayed using enzyme immunoassay following methods of Scott and Sorensen (1994). The Behavior Description HSPs (27, 40, 60, and 70) were extracted, separated using polyacrylamide gel electrophoresis, blotted onto nitrocellulose membranes, probed sequentially Fin display Fish extends pelvic and pectoral fins outward with anti-HSP27, HSP40, HSP60, and HSP70 anti-bodies, and quantified using to the limit of motion. chemiluminescent methods as described in Methods. Bars indicate relative HSP Flutter Propagation of a waving motion along the present in four fish randomly selected from each treatment. In each case, error entire body. bars represent SD. Aggression Chasing or nipping at the fins of another fish, or both. Cortisol averaged 0.53 ng/L across all tanks and no sig- Gape Opening the mouth for a period of more nificant differences were found among treatments or between than 2 s. sampling times (two-way RM ANOVA: P > 0.8 for treatment, Chatter Rapid and repeated opening and closing of the P > 0.5 for time, and P > 0.8 for interaction of treatment and mouth. time; Figure 4B). No differences in variability were detected AQUARIUM-RELATED STRESSORS ON THE ZEBRAFISH 59

(P > 0.2 for treatment, time, and interaction). The 0.13-fish/L there were some differences in behaviors over time. Frequencies density tanks were not sampled on day 5. of fin display and flutter were significantly lower for the first Relative protein did not differ among treatments for any of three versus the later days, and frequency of aggression was the HSPs tested (one-way ANOVA: HSP27, P > 0.2; HSP40, lower on the first day compared with other days (P < 0.05 in P > 0.8; HSP60, P > 0.2; HSP70, P > 0.7; Figure 4C). Vari- all cases). However, these differences were consistent across ability among individual fish, both within and among replicate treatments; no interactions were observed between treatment tanks, was high, but no systematic differences related to treat- and time (P > 0.2 in all cases). In terms of relative frequencies, ment were found (one-way ANOVA: P > 0.3 in all cases, except gaping was not observed as often and fin display and chattering in the case of HSP40 where Kruskal–Wallis ANOVA on ranks appeared to be the most common. Variability of behaviors did was performed). not differ with treatment, and in general did not change with time (P > 0.07 in all cases), though in the case of the fluttering, Tank Lighting frequencies on the first day were higher only for the 4,100-K Behaviors did not differ among lighting treatments (two-way lighting treatment (P < 0.05). RM ANOVA: fin display, P > 0.8; flutter, P > 0.06; aggression, As was the case for stocking density, cortisol in tank water P > 0.3; gape, P > 0.9; chatter, P > 0.8; Figure 5A); however, averaged approximately 24 ng/L and no differences were found among treatments (two-way RM ANOVA: P > 0.3; Figure 5B), though cortisol levels on day 10 were significantly lower than on day 5 (P < 0.03). No changes in variability of data among treatments or over time were detected (P > 0.2 in all cases). Similar to the stocking density experiment, no significant differences were found among HSPs (one-way ANOVA: HSP27, P > 0.7; HSP40, P > 0.9; HSP60, P > 0.7; HSP70, P > 0.6; Figure 5C). Variability was high among individual fish and treatments, but no differences in variability could be attributed to treatment for any HSP (P > 0.4 in all cases; HSP60 and HSP70 required analyses with Kruskal–Wallis ANOVA on ranks).

Food Deprivation Most behaviors did not differ among treatments (two-way RM ANOVA: fin display, P > 0.3; aggression, P > 0.2; gape, P > 0.1; chatter, P > 0.2; Figure 6A). However, fluttering increased significantly in frequency for fish deprived of food for 72 or 216 h versus 24 h (Tukey’s test: P < 0.04). In terms of relative frequencies of behaviors, aggression was least common and fin display most common overall. There were no changes in frequencies of behavior with time (P > 0.06 in all cases), and variability did not change with treatment or time (P > 0.05 in all cases), except in the case of gaping where frequencies on the final day of the experiments were significantly higher across all

Downloaded by [Department Of Fisheries] at 23:33 28 February 2013 treatments (P < 0.03). There were signiﬁcant differences in cortisol concentrations found in the water, with lower cortisol at 120 and 216 h versus 24 or 72 h (Tukey’s tests: P < 0.001; Figure 6B). All HSPs except HSP70 showed differences among treatments (one-way ANOVA: HSP27, P < 0.04; HSP40, P < 0.001; HSP60, P < 0.01; HSP70, P > 0.5; Figure 6C). Tukey’s tests generally showed that HSPs at 216 h were higher than at 24 or 72 h. No differences in variability were seen across treatments (P > 0.3 in all cases; HSP27 required analysis with Kruskal–Wallis ANOVA on ranks).

FIGURE 5. Effects of three different tank lighting systems (6,500-K bulbs, Tapping 420 actinic bulbs, and 4,100-K bulbs) for Zebrafish on (A) behavior, (B) cortisol measured in aquarium water, and (C) HSPs measured in muscle tissue. All tanks Frequency of behaviors did not differ among treatments received equal quantum irradiance. Other details are as in Figure 3. (two-way RM ANOVA: fin display, P > 0.6; flutter, P > 0.07; 60 GRONQUIST AND BERGES

FIGURE 6. Effects of food deprivation on Zebrafish for 24, 72, and 216 h on FIGURE 7. Effects of random tapping of aquaria of Zebrafish on (A) behavior, (A) behavior, (B) cortisol measured in aquarium water, and (C) HSPs measured (B) cortisol measured in aquarium water, and (C) HSPs measured in muscle in muscle tissue. Other details are as in Figure 3. tissue. Aquaria were tapped using a custom-built programmable device (see Methods). Tanks tapped at a “low rate” received a total of 10 min in every 24 h of random taps, delivered in 10-s bursts at a rate of 1 tap/s. Tanks tapped at a aggression, P > 0.4; gape, P > 0.1; chatter, P > 0.3; Figure 7A). “high rate” received a total of 20 min in every 24 h of random taps, delivered in 10-s bursts at a rate of 2 taps/s. The final series of tanks did not receive tapping. Downloaded by [Department Of Fisheries] at 23:33 28 February 2013 In general, fin display was more frequent than the other behav- Other details are as in Figure 3. iors, while gaping appeared to be the least displayed. However, there were some differences in behaviors over time. Frequencies of fin display were higher in the last 4 d compared with the first HSP70, P > 0.12; Figure 7C). Levels of HSP40 were higher < 3d(P 0.05), and both chattering and fluttering frequencies in fish from tanks exposed to high tapping rates versus low or were greater on the fourth day of the experiment compared with no tapping (Tukey’s test). No differences in variability among < other days (P 0.05). No changes in variability of behaviors treatments were found for any HSP (P > 0.05 in all cases). were observed related to treatment or time (P > 0.1 in all cases). Cortisol averaged 0.54 ng/L over all tanks and no differences were apparent among treatments or over time (two-way DISCUSSION ANOVA: P > 0.13, P > 0.06; Figure 7B). No changes in variability were found with respect to treatment or time (P > 0.7 in Effects of Specific Stressors all cases). Remarkably few significant differences were found between Only HSP40 showed differences among treatments (one-way treatments and controls for any of the aquarium-related stressors ANOVA: HSP27, P > 0.16; HSP40, P < 0.05; HSP60, P > 0.9; investigated; only food deprivation showed consistent results. AQUARIUM-RELATED STRESSORS ON THE ZEBRAFISH 61

The hypothesis that increasing stocking densities in Zebrafish much higher doses than are feasible using commercially avail- would result in alterations in behavior, increased cortisol release, able lighting. For other species, light spectra appear to have and elevated tissue HSPs was rejected, suggesting that the stock- inconsistent effects that include red light spectra decreasing ox- ing densities used were not stressful to Zebrafish. These results idative stress relative to white light (in Yellowtail Clownfish contrast with some in the literature. For example, Moretz et al. Amphiprion clarkii, Shin et al. 2011), red light spectra or dark- (2007) found Zebrafish showed increases in aggression behav- ness decreasing plasma cortisol relative to white light in Tench iors at a density of 1.4 fish/L (though more fish were used in Tinca tinca (Owen et al. 2010), blue light spectra increasing their study; 30 fish in 21-L tanks), and Spence and Smith (2005) growth and size heterogeneity and lowering cortisol (in Rain- found male Zebrafish displayed increased territorial behaviors bow Trout and Common Carp Cyprinus carpio, Karakatsouli at a density of 0.25 fish/L (just twice our lowest density of et al. 2008, 2010), blue light spectra increasing plasma cortisol 0.13 fish/L). Larson et al. (2006) demonstrated development (in Atlantic Salmon postsmolts, Migaud et al. 2007), or no ef- of dominant–subordinate behaviors in Zebrafish at densities of fects of spectra on cortisol (in Red Porgy Pagrus pagrus, Szisch only 0.025 fish/L (though the frequencies of such behaviors in et al. 2002). Head and Malison (2000) report a synergistic effect their study were an order of magnitude greater than in the present of blue light and handling stress on plasma cortisol and growth study; “bites” and “chases,” contained in our aggression cate- rate in Yellow Perch Perca flavescens. gory, were observed at an equivalent frequency of between 27 We hypothesized that depriving Zebrafish of food over a and 90 events per 5-min observation; we averaged only eight at 10-d period would cause changes in behavior and uniformly the highest stocking densities). For other indicators, much higher increase cortisol in water and HSPs in tissues. Elements of the densities may be required for effects to be seen. For example, hypothesis were supported, but while tissue HSP40 and HSP60 Castranova et al. (2011) saw no effects on Zebrafish reproduc- increased in response to deprivation, cortisol in tank water action or viability until densities of 12 fish/L were reached, and tually decreased. Reports in the literature of the effects of food Goolish et al. (1998) did not observe changes in egg produc- deprivation on stress in fishes vary considerably. In terms of tion, percentage hatch, or larval length until densities reached behavior, Novak et al. (2005) found fasting Zebrafish had an 60 fish/L. Ramsay et al. (2006) found that whole-body corti- increase in physical activity after 2 d followed by a sharp drop sol responses (four-fold increases) were observed at 40 fish/L. until fish were fed after 15 d; this was not reflected in frequency Few data are available for HSP responses to stocking density in of behaviors in the present study, though both “flutter” and Zebrafish, but results for other species are variable. Zarate and “aggression” behaviors, while not significantly different among Bradley (2003) found HSP70 was not an indicator of crowding treatments, were qualitatively more variable than what was seen stress in Atlantic Salmon Salmo salar, but Gornati et al. (2005) in other experimental series (Figure 6). Food deprivation can did see increased HSP in Sea Bass Dicentrarchus labrax (also increase cortisol; for example, Olsen et al. (2005) found cor- known as European Bass Morone labrax) at high stocking den- tisol increased detectibly in Rainbow Trout after 3 d of food sities. Aksakal et al. (2011) saw increases in muscle HSP70 in deprivation, which contrasts with results of the present study. Rainbow Trout Oncorhynchus mykiss in response to increased A common pattern in fishes in response to stress is for cortisol fish densities, but this only occurred at densities two orders of levels to first increase from prestress levels, then acclimate and magnitude higher than the densities used in the present study. It return to prestress levels (Wendelaar Bonga 1997; Barton 2002), is worth noting that for laboratory work, stocking densities of but the changes in in-water cortisol observed in the present study Zebrafish are typically much greater (2–20 fish/L; Goolish et al. actually represent a decline in cortisol released relative to other 1998; Ramsay et al. 2009; Parker et al. 2012) than normally experiments as well. Few data are available for HSP responses in used in commercial or home aquaria. Fish welfare concerns at Zebrafishes. Drew et al. (2008) found little evidence of changes Downloaded by [Department Of Fisheries] at 23:33 28 February 2013 high density typically focus on water quality, but lower densities in Zebrafish transcriptomes (including HSPs) after 21 d of star- may also constitute social stressors (Larson et al. 2006; Parker vation. In other species, there is little consensus: Cara et al. et al. 2012). (2005) found that HSP70 and HSP90 expression increased after We hypothesized that there would be no differences in re- 7 d in response to food deprivation in Rainbow Trout and Gilt- sponses between fish in aquaria lighted with typical “cool head Seabream Sparus aurata, but Weber and Bosworth (2005) white” spectrum fluorescent indoor bulbs (color temperature, studied food deprivation in Channel Catfish Ictalurus punctatus 4,100 K) and enhanced daylight-spectrum lamps (color temper- and saw no changes in HSPs despite reduced growth. It is clear ature, 6,500 K), but that UV-enhanced bulbs (color temperature, that some responses may take place over a longer period than 7,100 K) would result in changes in behaviors and in elevated what was allowed in the present study. Oxidative responses in cortisol and HSPs. In fact, no evidence was found to support any Brown Trout Salmo trutta did not become detectible until after effect of lighting. De Oliveira Mesquita et al. (2008) reported 49 d of starvation (Bayir et al. 2011). Importantly, Lee et al. that stroboscopic light had a deterrent effect on Zebrafish behav- (2011), working with cichlids, pointed out that there is high ior, but also noted that fish acclimated to the treatment within an variability among fishes in their response to feeding and food hour. Ultraviolet radiation can elicit stress responses in devel- deprivation due to their relative dominance; nondominant fish oping Zebrafish (Behrendt et al. 2010), but such effects occur at are likely to get less than their share of food and thus store less, 62 GRONQUIST AND BERGES

and so incur an additional “cost of stress” (see below). However, uals. Thus, potentially important information was being lost. such issues may not actually arise in a case of food deprivation; This limitation of this type of observation has been noted for Brann¨ as¨ et al. (2003), working with Brown Trout, found that some time (Altmann 1974). Video-tracking systems for behav- more aggressive behaviors occurred with limited food rather ior are now available (e.g., Fontaine et al. 2008), and processor than no food. capability is up to the task of distinguishing multiple targets. We hypothesized that fish experiencing tapping on aquarium Moreover, such tracking can identify more subtle or aggregate glass would exhibit changes in behavior and elevated cortisol behaviors (e.g., “freezing” or “reduced exploration,” see Cachat and HSPs relative to those that did not. This hypothesis was et al. 2010) that are impossible to quantify using the methods of rejected, although significantly higher HSP40 was seen in fish the present study. subjected to the highest tapping rate. It seems odd there were no In-water cortisol measurements proved relatively straightfor- differences detected in behaviors because during observations ward; herein, we report another species, Zebrafish, to which this when a tank tapper began a tapping cycle, the fish displayed a method has been successfully applied (cf., Scott et al. 2008). marked startle response, typically a “jump” followed by acceler- Cortisol recoveries are in the range previously reported (Scott ated swimming for approximately 3–5 s. This response did not and Ellis 2007), and though the average release rates of the decrease noticeably over the course of the experiment. The tap- present study are higher than the range quoted in Scott et al. ping treatment differed from the other treatments in that it was (2008), we are using fish that are substantially smaller than applied in clear “pulses” versus continuously. We considered the those used in most previous work; there is a recognized inverse irregular and unpredictable tapping as a “chronic background” relationship between release rates and fish size (Scott et al. condition in aquaria, but this is clearly distinct from stimuli like 2008). pump noise (e.g., Anderson et al. 2011). In this sense, tapping Measurement of HSPs was terminal and labor intensive, and might be described as a “repeated acute stressor” (cf., handling results were highly variable among individuals in the present stress, Barton 2002) and it might have been more reasonable study. It was, however, the only truly individual measurement to monitor fish immediately before and after tapping. However, made. Though widely used, concerns have been expressed about such an approach would have had two weaknesses. Firstly, fish whether HSPs can actually serve as stress indicators in fish, are capable of cumulative responses to repeated stressors (see considering their variation with developmental stage, specific discussion in Barton 2002) and so a summative measurement tissue, and the lack of information of differences between acute is useful, and secondly, the random nature of the tapping ef- and chronic stressors (see Iwama et al. 2004). As we gain better fectively prevented making observations and measurements at understanding of the identity and roles of HSPs, some of these fixed points after the tapping occurred. Purser et al. (2011) re- issues may become clearer. For example, Marvin et al. (2008) ported that Threespined Sticklebacks Gasterosteus aculeatus demonstrated that of 13 small HSPs found in Zebrafish, seven exposed to acoustic noise showed declines in foraging efficiency were developmentally regulated in a tissue-specific manner, but and alterations in specific behaviors. Anderson et al. (2011) five appeared to show clear heat-shock responses. Having the demonstrated elevated cortisol in seahorses in tanks exposed to ability to target specific genes or protein isoforms may im- high levels of noise. Smith et al. (2004) found Goldfish Caras- prove interpretation substantially. Evolution of transcriptomic sius auratus exposed to noise did not have a long-term phys- and proteomic microarrays may offer hope for this approach in iological stress response, but there was a short-term increase Zebrafish (e.g., Ton et al. 2002; Bosworth et al. 2005). in plasma cortisol, suggesting rapid acclimation. Wysocki et al. (2006) found elevated cortisol in aquarium water of Common Complicating Issues Carp, Gudgeon Gobio gobio, and Eurasian Perch Perca fluvi- In the present study, we chose to work with fish in small Downloaded by [Department Of Fisheries] at 23:33 28 February 2013 atilis exposed to the underwater sound recorded in the presence groups rather than singly as is done in some cases (e.g., toxi- of a ship, relative to control fish. Other studies have failed to cological testing). Small groups represent the situation in most find effects. Leong et al. (2009) exposed Mozambique Tilapia aquaria, and there is good evidence of positive effects of main- Oreochromis mossambicus to “knocking” on tank walls over 7 tenance of fish in groups on welfare in aquaculture settings d and observed effects on locomotory activity and oxygen con- (e.g., Saxby et al. 2010). However, such an approach introduces sumption, but not on cortisol. Davidson et al. (2009) found little additional social aspects to the experiments. difference in growth and condition between Rainbow Trout ex- We randomized selection of fish for experiments to mini- posed to noises typical in an aquaculture setting and control fish mize handling stress and did not attempt to control aspects like after 5 months. sex ratio. It is clear that this is an issue for behavior; sex ratio clearly affects male territoriality and specific behaviors such as Usefulness of Specific Methods and Potential Refinements aggression and courtship (Darrow and Harris 2004; Spence and Behavioral monitoring of Zebrafish was conceptually sim- Smith 2005; Moretz et al. 2007). The relatively small number ple, though labor intensive. A problem noted from the start was of fish used may also have prevented some social behaviors that, while total numbers of behaviors of the six fish could be such as shoaling; Zebrafish show a preference for shoaling in tallied, these behaviors could not be associated with individ- larger numbers, which is more likely to occur in groups where AQUARIUM-RELATED STRESSORS ON THE ZEBRAFISH 63

females predominate (Spence et al. 2008). Zebrafish develop enzymes and elevates mRNA levels of heat shock protein 70 in Rainbow dominance hierarchies (Gerlach and Lysiak 2006; Gerlach et al. Trout. Livestock Science 141:69–75. 2007), and so stress in the environment is not simply a function Altmann, J. 1974. Observational study of behavior: sampling methods. Be- haviour 49:227–267. of physical and chemical factors, but can also be due to the social Anderson, P. A., I. K. Berzins, F. Fogarty, H. J. Hamlin, and L. J. Guillette Jr. environment. Indeed, while there are positive aspects to social 2011. Sound, stress, and seahorses: the consequences of a noisy environment groupings (e.g., improvement of growth and stress reduction in to animal health. Aquaculture 311:129–138. the presence of “kin”; Gerlach et al. 2007), it is also clear that Ashley, P. J. 2007. Fish welfare: current issues in aquaculture. Applied Animal subordinate animals experience measureable stress (including Behaviour Science 104:199–235. Bartelme, T. D. 2004. Short take: stress in fish, part II: why you should care elevated plasma cortisol) as a result of aggression experiences about stress in fish. Advanced Aquarist [online serial] 3(9):article 4. Available: because of their lower status (e.g., Filby et al. 2010). Moreover, www.advancedaquarist.com/issues/sept2004/fish2.htm. (April 2006). Parker et al. (2012) demonstrated that housing Zebrafish alone or Barton, B. A. 1997. Stress in finfish: past, present and future—a historical in groups could modify their behavioral responses to stressors. perspective. Pages 1–33 in G. K. Iwama, A. D. Pickering, J. P. Sumpter, It is unclear the extent to which social hierarchies established and C. B. Schreck, editors. Fish stress and health in aquaculture. Society for Experimental Biology Seminar Series 62, Cambridge University Press, themselves in our experiments, but according to Larson et al. Cambridge, UK. (2006) subordinate Zebrafish are paler in color, and color in- Barton, B. A. 2002. Stress in fishes: a diversity of responses with particular ref- tensity variations were noted among tanks in the present study. erence to changes in circulating corticosteroids. Integrative and Comparative Interestingly, in cichlids, Fox et al. (1997) observed that social Biology 42:517–525. and reproductive statuses interact to affect stress responses. Fi- Basquill, S. P., and J. W. A. Grant. 1998. An increase in habitat complexity reduces aggression and monopolization of food by Zebra Fish (Danio rerio). nally, environmental complexity may play a role. In order to Canadian Journal of Zoology 76:770–772. maximize observation ability and minimize interfering chemi- Basu, N., A. E. Todgham, P. A. Ackerman, M. R. Bibeau, K. Nakano, P. M. cals, we chose the simplest possible tank architecture; tanks did Schulte, and G. K. Iwama. 2002. Heat shock protein genes and their functional not contain shelters or plants, for example. Aggressive behaviors significance in fish. Gene 295:173–183. in social fishes, including Zebrafish, are reduced in vegetated Bayir, A., A. N. Sirkecioglu, M. Bayir, H. I. Haliloglu, E. M. Kocaman, and N. M. Aras. 2011. Metabolic responses to prolonged starvation, food restriction, versus nonvegetative habitats, which are more typical of display and refeeding in the Brown Trout, Salmo trutta: ozxidative stress and antiox- aquaria (Basquill and Grant 1998; Fox et al. 1997). The fail- idant defenses. Comparative Biochemistry and Physiology 159B:191–196. ure to control these aspects in the present study is most likely Behrendt, L., M. E. Jonsson,¨ J. V.Goldstone, and J. J. Stegeman. 2010. Induction to result in high variability among individuals; this was indeed of cytochrome P450 1 genes and stress response genes in developing Zebrafish observed in the HSP data. exposed to ultraviolet radiation. Aquatic Toxicology 98:74–82. Bosworth, C. A., IV, C. W. Chou, R. B. Cole, and B. B. Rees. 2005. Protein In summary, despite concerns about aquarium-related stres- expression patterns in Zebrafish skeletal muscle: initial characterization and sors such as crowding, food deprivation, lighting spectra, and the effects of hypoxic exposure. Proteomics 5:1362–1371. tapping on tanks, these appeared to have relatively little de- Brann¨ as,¨ E., S. Jonsson, and H. Lundqvist. 2003. Influence of food abundance tectible effect on Zebrafish. Part of the explanation may have on individual behaviour strategy and growth rate in juvenile Brown Trout to do with the overriding effects of intrinsic factors like so- (Salmo trutta). Canadian Journal of Zoology 81:684–692. Brett, J. R. 1958. Implications and assessments of environmental stress. Pages cial stressors, but it also appears that Zebrafish are hardy and 69–83 in P. A. Larkin, editor. The investigation of fish-power problems: a adaptable to aquarium conditions. symposium held at the University of British Columbia. Institute of Fisheries, University of British Columbia, Vancouver. Cachat, J., A. Stewart, L. Grossman, S. Gaikwad, F. Kadri, K. M. Chung, N. ACKNOWLEDGMENTS Wu, K. Wong, S. Roy, C. Suciu, J. Goodspeed, M. Elegante, B. Bartels, S. Elkhayat, D. Tien, J. Tan, A. Denmark, T. Gilder, E. Kyzar, J. DiLeo, K. We thank Fred Binkowski for advice and assistance in set- Frank, K. Chang, E. Utterback, P. Hart, and A. V. Kalueff. 2010. Measuring Downloaded by [Department Of Fisheries] at 23:33 28 February 2013 ting up and maintaining aquaria, Tom Consi and Kory Verhein behavioral and endocrine responses to novelty stress in adult Zebrafish. Nature for designing and building the tank tappers, and Mike Carvan, Protocols 5:1786–1799. Harvey Bootsma, and Rick Goetz for valuable discussions of Cara, J. B., N. Aluru, F. J. Moyano, and M. M. Vijayan. 2005. Food-deprivation experimental design and results. induces HSP70 and HSP90 protein expression in larval Gilthead Sea Bream and Rainbow Trout. Comparative Biochemistry and Physiology 142B: 426–431. Castranova, D., A. Lawton, C. Lawrence, D. P. Baumann, J. Best, J. Coscolla, REFERENCES A. Doherty, J. Ramos, J. Hakkesteeg, C. M. Wang, C. Wilson, J. Malley, Ackerman, P. A., R. B. Forsyth, C. F. Mazur, and G. K. Iwama. 2000. Stress and B. M. Weinstein. 2011. The effect of stocking densities on reproductive hormones and the cellular stress response in salmonids. Fish Physiology and performance in laboratory Zebrafish (Danio rerio). Zebrafish 8:141–146. Biochemistry 23:327–336. Chervova, L. S. 1997. Pain sensitivity and behavior of fishes. Journal of Ichthy- Airaksinen, S., C. M. I. Rabergh,˚ A. Lahti, A. Kaatrasalo, L. Sistonen, and ology 37:98–102. M. Nikinmaa. 2003. Stressor-dependent regulation of the heat shock re- Chizinski, C. J., B. Sharma, K. L. Pope, and R. Patino.˜ 2008. A bioenergetic sponse in Zebrafish, Danio rerio. Comparative Biochemistry and Physiology model for Zebrafish Danio rerio (Hamilton). Journal of Fish Biology 73: 134A:839–846. 35–43. Aksakal, E., D. Ekinci, O. Erdogan,˘ S. Beydemir, Z. Alım, and S. B. Ceyhun. Colwill, R. M., and R. Creton. 2011. Imaging escape and avoidance behavior in 2011. Increasing stocking density causes inhibition of metabolic–antioxidant Zebrafish larvae. Reviews in the Neurosciences 22:63–73. 64 GRONQUIST AND BERGES

Craven, A., A. G. Carton, C. R. McPherson, and G. McPherson. 2009. Determin- Iwama, G. K., M. M. Vijayan, R. B. Forsyth, and P. A. Ackerman. 1999. ing and quantifying components of an aquaculture soundscape. Aquacultural Heat shock proteins and physiological stress in fish. American Zoologist 39: Engineering 41:158–165. 901–909. Darrow, K. O., and W. A. Harris. 2004. Characterization and development of Jobling, M. 1981. The influences of feeding on the metabolic rate of fishes: a courtship in Zebrafish, Danio rerio. Zebrafish 1:40–45. short review. Journal of Fish Biology 18:385–400. Davidson, J., J. Bebak, and P. Mazik. 2009. The effects of aquacul- Karakatsouli, N., S. E. Papoutsoglou, G. Panopoulos, E. S. Papoutsoglou, S. ture production noise on the growth, condition factor, feed conversion, Chadio, and D. Kalogiannis. 2008. Effects of light spectrum on growth and and survival of Rainbow Trout, Oncorhynchus mykiss. Aquaculture 288: stress response of Rainbow Trout Oncorhynchus mykiss reared under recir- 337–343. culating system conditions. Aquacultural Engineering 38:36–42. Davis, K. B., and B. C. Small. 2006. Rates of cortisol increase and decrease in Karakatsouli, N., E. S. Papoutsoglou, N. Sotiropoulos, D. Mourtikas, T. Stigen- Channel Catfish and Sunshine Bass exposed to an acute confinement stressor. Martinsen, and S. E. Papoutsoglou. 2010. Effects of light spectrum, rearing Comparative Biochemistry and Physiology 143C:134–139. density and light intensity on growth performance of scaled and mirror Com- de Oliveira Mesquita, F., H. Pereira Godinho, P. Guimaraes˜ de Azevedo, and mon Carp Cyprinus carpio reared under recirculating system conditions. R. J. Young. 2008. A preliminary study into the effectiveness of stroboscopic Aquacultural Engineering 42:121–127. light as an aversive stimulus for fish. Applied Animal Behaviour Science Larson, E. T., D. M. O’Malley, and R. H. Melloni Jr. 2006. Aggression and va- 111:402–407. sotocin are associated with dominant–subordinate relationships in Zebrafish. Drew, R. E., K. J. Rodnick, M. Settles, J. Wacyk, E. Churchill, M. S. Powell, Behavioural Brain Research 167:94–102. R. W. Hardy, G. K. Murdoch, R. A. Hill, and B. D. Robison. 2008. Effect Lawrence, C. 2007. The husbandry of Zebrafish (Danio rerio): a review. Aqua- of starvation on transcriptomes of brain and liver in adult female Zebrafish culture 269:1–20. (Danio rerio). Physiological Genomics 35:283–295. Lee, G., J. W. A. Grant, and P. Comolli. 2011. Dominant convict cichlids Ellis, T., B. North, A. P. Scott, N. R. Bromage, M. Porter, and D. Gadd. 2002. (Amatitlania nigrofasciata) grow faster than subordinates when fed an equal The relationships between stocking density and welfare in farmed Rainbow ration. Behaviour 148:877–887. Trout. Journal of Fish Biology 61:493–531. Leong, H., A. F. H. Ros, and R. F. Oliveira. 2009. Effects of putative stressors Filby, A. L., G. C. Paull, E. J. Bartlett, K. J. W. Van Look, and C. R. Tyler. 2010. in public aquaria on locomotor activity, metabolic rate and cortisol levels in Physiological and health consequences of social status in Zebrafish (Danio the Mozambique Tilapia Oreochromis mossambicus. Journal of Fish Biology rerio). Physiology and Behavior 101:576–587. 74:1549–1561. Fontaine, E., D. Lentink, S. Kranenbarg, U. K. Muller,¨ J. L. van Leeuwen, Marvin, M., D. O’Rourke, T. Kurihara, C. E. Juliano, K. L. Harrison, and L. D. A. H. Barr, and J. W. Burdick. 2008. Automated visual tracking for study- Hutson. 2008. Developmental expression patterns of the Zebrafish small heat ing the ontogeny of Zebrafish swimming. Journal of Experimental Biology shock proteins. Developmental Dynamics 237:454–463. 211:1305–1316. Migaud, H., M. Cowan, J. Taylor, and H. W. Ferguson. 2007. The effect of spec- Fox, H. E., S. A. White, M. H. F. Kao, and R. D. Fernald. 1997. tral composition and light intensity on melatonin, stress and retinal damage Stress and dominance in a social fish. Journal of Neuroscience 17: in post-smolt Atlantic Salmon, Salmo salar. Aquaculture 270:390–404. 6463–6469. Moretz, J. A., E. P. Martins, and B. D. Robison. 2007. The effects of early and Francis-Floyd, R. 1996. Use of formalin to control fish parasites. Institute of adult social environment on Zebrafish (Danio rerio) behavior. Environmental Food and Agricultural Sciences Extension, University of Florida, Publication Biology of Fishes 80:91–101. VM-77, Gainesville. Available: edis.ifas.ufl.edu/VM061. (April 2006). Novak, C. M., X. Jiang, C. Wang, J. A. Teske, C. M. Kotz, and J. A. Levine. Gerlach, G., A. Hodgins-Davis, B. MacDonald, and R. C. Hannah. 2007. Bene- 2005. Caloric restriction and physical activity in Zebrafish (Danio rerio). fits of kin association: related and familiar Zebrafish larvae (Danio rerio)show Neuroscience Letters 383:99–104. improved growth. Behavioral Ecology and Sociobiology 61:1765–1770. Olsen, R. E., K. Sundell, T. M. Mayhew, R. Myklebust, and E. Ringø. 2005. Gerlach, G., and N. Lysiak. 2006. Kin recognition and inbreeding avoidance in Acute stress alters intestinal function of Rainbow Trout, Oncorhynchus mykiss Zebrafish, Danio rerio, is based on phenotype matching. Animal Behaviour (Walbaum). Aquaculture 250:480–495. 71:1371–1377. Orlando, E. F., and L. J. Guillette Jr. 2001. A re-examination of variation Goolish, E. M., R. Evans, K. Okutake, and R. Max. 1998. Chamber volume associated with environmentally stressed organisms. Human Reproduction requirements for reproduction of the Zebrafish Danio rerio. Progressive Fish- Update 7:265–272. Culturist 60:127–132. Owen, M. A. G., S. J. Davies, and K. A. Sloman. 2010. Light colour influences Gornati, R., S. Gualdoni, R. Cavaliere, G. Terova, M. Saroglia, and G. the behaviour and stress physiology of captive Tench (Tinca tinca). Reviews Bernardini. 2005. Molecular biology and fish welfare: a winning combi- in Fish Biology and Fisheries 20:375–380. Downloaded by [Department Of Fisheries] at 23:33 28 February 2013 nation. Aquaculture International 13:51–55. Papoutsoglou, S. E., N. Karakatsouli, E. S. Papoutsoglou, and G. Vasilikos. Greene, R. M., R. J. Geider, and P. G. Falkowski. 1991. Effect of iron limi- 2010. Common Carp (Cyprinus carpio) response to two pieces of music tation on photosynthesis in a marine diatom. Limnology and Oceanography (“Eine Kleine Nachtmusik” and “Romanza”) combined with light inten- 36:1772–1782. sity, using recirculating water system. Fish Physiology and Biochemistry 36: Haukenes, A. H., and B. A. Barton. 2004. Characterization of the cortisol 539–554. response following an acute challenge with lipopolysaccharide in Yellow Parker, M. O., M. E. Millington, F. J. Combe, and C. H. Brennan. 2012. Perch and the influence of rearing density. Journal of Fish Biology 64: Housing conditions differentially affect physiological and behavioural stress 851–862. responses of Zebrafish, as well as the response to anxiolytics. PLoS (Pub- Head, A. B., and J. A. Malison. 2000. Effects of lighting spectrum and dis- lic Library of Science) ONEE [online serial] 7(4):e34992. DOI: 10.1371/ turbance level on the growth and stress responses of Yellow Perch Perca journal.pone.0034992. flavescens. Journal of the World Aquaculture Society 31:73–80. Purser, J., and A. N. Radford. 2011. Acoustic noise induces attention shifts Iwama, G. K., L. O. B. Afonso, A. Todgham, P. Ackerman, and K. Nakano. and reduces foraging performance in Three-Spined Sticklebacks (Gasteros- 2004. Are hsps suitable for indicating stressed states in fish? Journal of teus aculeatus). PLoS (Public Library of Science) ONE [online serial] Experimental Biology 207:15–19. 6(2):e17478. DOI: 10.1371/journal.pone.0017478. Iwama, G. K., L. O. B. Afonso, and M. M. Vijayan. 2006. Stress in fishes. Ramsay, J. M., G. W. Feist, Z. M. Varga, M. Westerfield, M. L. Kent, and C. Pages 319–342 in D. H. Evans and J. B. Claiborne, editors. The physiology B. Schreck. 2006. Whole-body cortisol is an indicator of crowding stress in of fishes, 3rd edition. CRC Press, Boca Raton, Florida. adult Zebrafish, Danio rerio. Aquaculture 258:565–574. AQUARIUM-RELATED STRESSORS ON THE ZEBRAFISH 65

Ramsay, J. M., G. W. Feist, Z. M. Varga, M. Westerfield, M. L. Kent, and C. Spence, R., G. Gerlach, C. Lawrence, and C. Smith. 2008. The behaviour and B. Schreck. 2009. Whole-body cortisol response of Zebrafish to acute net ecology of the Zebrafish, Danio rerio. Biological Reviews of the Cambridge handling stress. Aquaculture 297:157–162. Philosophical Society 83:13–34. Robinson, J., E. A. Schmitt, F. I. Harosi,´ R. J. Reece, and J. E. Dowling. Spence, R., and C. Smith. 2005. Male territoriality mediates density and sex 1993. Zebrafish ultraviolet visual pigment: absorption spectrum, sequence, ratio effects on oviposition in the Zebrafish, Danio rerio. Animal Behaviour and localization. Proceedings of the National Academy of Sciences of the 69:1317–1323. USA 90:6009–6012. Szisch, V., A. L. van der Salm, S. E. Wendelaar Bonga, and M. Pavlidis. 2002. Saverino, C., and R. Gerlai. 2008. The social Zebrafish: behavioral responses to Physiological colour changes in the Red Porgy, Pagrus pagrus, following conspecific, heterospecific, and computer animated fish. Behavioural Brain adaptation to blue lighting spectrum. Fish Physiology and Biochemistry 27: Research 191:77–87. 1–8. Saxby, A., L. Adams, D. Snellgrove, R. W. Wilson, and K. A. Sloman. 2010. Ton, C., D. Stamatiou, V. J. Dzau, and C. C. Liew. 2002. Construction of a The effect of group size on the behaviour and welfare of four fish species com- Zebrafish cDNA microarray: gene expression profiling of the Zebrafish dur- monly kept in home aquaria. Applied Animal Behaviour Science 125:195– ing development. Biochemical and Biophysical Research Communications 205. 296:1134–1142. Scott, A. P., and T. Ellis. 2007. Measurement of fish steroids in water—a review. Towbin, H., T. Staehelin, and J. Gordon. 1979. Electrophoretic transfer of pro- General and Comparative Endocrinology 153:392–400. teins from polyacrylamide gels to nitrocellulose sheets: procedure and some Scott, A. P., K. Hirschenhauser, N. Bender, R. Oliveira, R. L. Earley, M. Sebire, applications. Proceedings of the National Academy of Sciences of the USA T. Ellis, M. Pavlidis, P. C. Hubbard, M. Huertas, and A. Canario. 2008. 76:4350–4354. Non-invasive measurement of steroids in fish-holding water: important con- Weber, T. E., and B. G. Bosworth. 2005. Effects of 28 day exposure to cold siderations when applying the procedure to behaviour studies. Behaviour temperature or feed restriction on growth, body composition, and expres- 145:1307–1328. sion of genes related to muscle growth and metabolism in Channel Catfish. Scott, A. P., M. Pinillos, and T. Ellis. 2001. Why measure steroids in fish plasma Aquaculture 246:483–492. when you can measure them in water? Pages 1291–1295 in H. J. T. Goos, Wendelaar Bonga, S. E. 1997. The stress response in fish. Physiological Reviews R. K. Rastogi, H. Vaudry, and R. Pierantoni, editors. Perspective in compar- 77:591–625. ative endocrinology: unity and diversity. Monduzzi Editore, Bologna, Italy. Westerfield, M. 1993. The Zebrafish book: a guide for the laboratory use of Scott, A. P., and P. W. Sorensen. 1994. Time course of release of pheromonally Zebrafish (Danio rerio). University of Oregon Press, Eugene. active gonadal steroids and their conjugates by ovulatory Goldfish. General Wysocki, L. E., J. P. Dittami, and F. Ladich. 2006. Ship noise and corti- and Comparative Endocrinology 96:309–323. sol secretion in European freshwater fishes. Biological Conservation 128: Shin, H. S., J. Lee, and C. Y.Choi. 2011. Effects of LED light spectra on oxidative 501–508. stress and the protective role of melatonin in relation to the daily rhythm of Wysocki, L. E., F. Ladich, and J. P. Dittami. 2004. Noise, stress, and cortisol the Yellowtail Clownfish, Amphiprion clarkii. Comparative Biochemistry and secretion in teleost fishes: society for behavioral neuroendocrinology annual Physiology 160A:221–228. meeting abstract. Hormones and Behavior 46:125. Smith, M. E., A. S. Kane, and A. N. Popper. 2004. Noise-induced stress response Zarate, J., and T. M. Bradley. 2003. Heat shock proteins are not and hearing loss in Goldfish (Carassius auratus). Journal of Experimental sensitive indicators of hatchery stress in salmon. Aquaculture 223: Biology 207:427–435. 175–187. Downloaded by [Department Of Fisheries] at 23:33 28 February 2013 This article was downloaded by: [Department Of Fisheries] On: 28 February 2013, At: 23:34 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Aquatic Animal Health Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/uahh20 Experimental Infection of Australian Freshwater Fish with Epizootic Haematopoietic Necrosis Virus (EHNV) Joy A. Becker a , Alison Tweedie a , Dean Gilligan b , Martin Asmus c & Richard J. Whittington a

a Faculty of Veterinary Science, University of Sydney, 425 Werombi Road, Camden, New South Wales, 2570, Australia b New South Wales Department of Primary Industries, Batemans Bay Fisheries Office, Beach Road and Orient Street, Batemans Bay, New South Wales, 2536, Australia c New South Wales Department of Primary Industries, Narrandera Fisheries Centre, Buckingbong Road, Narrandera, New South Wales, 2700, Australia Version of record first published: 22 Jan 2013.

To cite this article: Joy A. Becker , Alison Tweedie , Dean Gilligan , Martin Asmus & Richard J. Whittington (2013): Experimental Infection of Australian Freshwater Fish with Epizootic Haematopoietic Necrosis Virus (EHNV), Journal of Aquatic Animal Health, 25:1, 66-76 To link to this article: http://dx.doi.org/10.1080/08997659.2012.747451

PLEASE SCROLL DOWN FOR ARTICLE

The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material. Journal of Aquatic Animal Health 25:66–76, 2013 C American Fisheries Society 2013 ISSN: 0899-7659 print / 1548-8667 online DOI: 10.1080/08997659.2012.747451

ARTICLE

Experimental Infection of Australian Freshwater Fish with Epizootic Haematopoietic Necrosis Virus (EHNV)

Joy A. Becker* and Alison Tweedie Faculty of Veterinary Science, University of Sydney, 425 Werombi Road, Camden, New South Wales 2570, Australia Dean Gilligan New South Wales Department of Primary Industries, Batemans Bay Fisheries Ofﬁce, Beach Road and Orient Street, Batemans Bay, New South Wales 2536, Australia Martin Asmus New South Wales Department of Primary Industries, Narrandera Fisheries Centre, Buckingbong Road, Narrandera, New South Wales 2700, Australia Richard J. Whittington Faculty of Veterinary Science, University of Sydney, 425 Werombi Road, Camden, New South Wales 2570, Australia

Abstract The ranavirus, epizootic hematopoietic necrosis virus (EHNV), is endemic to southern Australia with natural outbreaks resulting in mass mortality events in wild Redfin Perch Perca fluviatilis (also known as Eurasian Perch) and less severe disease in farmed Rainbow Trout Oncorhynchus mykiss. To further investigate the host range for EHNV, 12 ecologically or economically important freshwater fish species from southeastern Australia were exposed experimentally to the virus. A bath-challenge model at 18 ± 3◦C was employed with limited use of intraperitoneal inoculation to determine if a species was likely to be susceptible to EHNV. Of the species tested, Murray–Darling Rainbowfish Melanotaenia fluviatilis and Dewfish Tandanus tandanus (also known as Freshwater Catfish) were considered to be potentially susceptible species. EHNV was isolated from approximately 7% of surviving Eastern Mosquitofish Gambusia holbrooki, indicating this widespread alien fish species is a potential carrier. The infection of Silver Perch Bidyanus bidyanus and Macquarie Perch Macquaria australasica and the lack of infection in Murray Cod Maccullochella peelii peelii and Golden Perch Macquaria ambigua ambigua after exposure to EHNV via water confirmed earlier data from

Downloaded by [Department Of Fisheries] at 23:34 28 February 2013 Langdon (1989). Five other species of native fish were potentially not susceptible to the virus or the fish were able to recover during the standard 35-d postchallenge observation period. Overall, it appeared that EHNV was less virulent in the present experimental model than in previous studies, but the reasons for this were not identified.

The ranavirus, epizootic hematopoietic necrosis virus (Langdon et al. 1986; Langdon and Humphrey 1987). Sub- (EHNV), is a virus listed by the Office international des sequently, disease outbreaks were documented in wild Redfin epizooties´ (OIE) and is found only in Australia. It first emerged Perch in many lakes, rivers, dams, and impoundments in south- in the mid-1980s during epidemic disease events in wild eastern Australia, where the virus is now considered endemic, Redfin Perch Perca fluviatilis (also known as Eurasian Perch) albeit in a discontinuous distribution (Whittington et al. 2010).

*Corresponding author: [email protected] Received May 21, 2012; accepted November 1, 2012 Published online January 23, 2013

66 INFECTION OF FRESHWATER FISH WITH EHNV 67

Additionally, natural outbreaks of EHNV are known to cause 1986) and Murray Cod (Langdon 1989). This led to speculation low rates of mortality at Rainbow Trout Oncorhynchus mykiss that EHNV may have contributed to the decline of native fish farms, most of which use surface water from a source with a local species, in particular the endangered Macquarie Perch (Langdon wild Redfin Perch population present (Whittington et al. 1994, 1989; Lintermans 2007). Laboratory studies have also revealed 1999). An infection with EHNV results in multifocal necrosis that a number of freshwater fish species from European wa- of the hematopoietic tissue found in liver, spleen, and kidney ters, such as the Northern Pike Esox lucius, Pikeperch Sander (Langdon and Humphrey 1987; Langdon et al. 1988; Whitting- lucioperca (also known as Zander), and Black Bullhead Ameiu- ton et al. 1994; Reddacliff and Whittington 1996). Based on rus melas are susceptible to EHNV infection (Bang Jensen et al. the results of experimental bath challenges with EHNV, adult 2009, 2011; Gobbo et al. 2010). It is well recognized that suscep- Redfin Perch succumb to the disease, epizootic hematopoietic tibility to disease is not a fixed parameter and is influenced by a necrosis (EHN), within 28 d at water temperatures greater than complex series of interactions involving the host, pathogen, and 12◦C (Whittington and Reddacliff 1995). Unlike Redfin Perch, environmental factors (Trust 1986; Hedrick 1998; Reno 1998). Rainbow Trout tend to develop a chronic, low-grade infection Thus, it is useful for challenge experiments to be repeated in a with few clinical signs and a low rate of mortality that may not range of laboratories where operating conditions may differ. rise above the expected background mortality at a given farm The objective of this study was to determine the susceptibil- (Whittington et al. 2010). Thus, in nature there is a wide range ity to EHNV of 12 native and nonnative fish species commonly of susceptibility to EHNV between species of finfish, a fact that found throughout the Murray–Darling Basin in southeastern has been confirmed experimentally (Langdon 1989). Australia. The study involved a bath-challenge model with lim- Transmission of EHNV between susceptible hosts within a ited use of IP injection and identified several previously untested population is possible via water or ingestion of tissues from species to be susceptible to EHNV. infected fish. Movement of infected trout fingerlings containing a low number of individuals with a subclinical infection was probably the most common means of spread of EHNV within METHODS the aquaculture industry (Whittington et al. 2010). Epidemio- Selection of fish species.—Fish species were selected based logical evidence supports this as several affected trout farms, on their current aquaculture status, conservation status, prior located in different catchment areas, have a history of trading knowledge of EHNV susceptibility, distribution across the fish and fish products (Langdon et al. 1988; Whittington et al. Murray–Darling Basin, juvenile size ranges, and likelihood of 1994, 1999). Translocation of fish in aquaculture enterprises is acclimatization to artificial feeds and holding tanks, and were a major risk factor for the spread of disease among farms as selected to maximize the representation of taxonomic families well as to wild populations (Reno 1998; Fenichel et al. 2008). present in the Murray–Darling Basin (Table 1). There was no However, the need to pursue and advance production techniques history of EHN from any of the hatcheries from which fish were for aquaculture species is increasing in response to the demand sourced. for fish for human consumption, the ornamental fish trade, and Fish husbandry.—Fish were acquired either as hatchery- to support restocking schemes for threatened species conser- reared fingerlings or from wild-caught stock (Table 1). The vation or sport. In Australia, fish species are emerging in the size and age of fish varied substantially among species. Prior to aquaculture sector that fall under several of these different uses. overnight shipment to the University of Sydney, Camden, New For example, Silver Perch Bidyanus bidyanus and Murray Cod South Wales, fingerlings from Narrandera Fisheries Centre were Maccullochella peelii peelii are both iconic endemic freshwater harvested from outdoor earthen ponds and quarantined for 48 h fishes with wild populations in decline (Barrett 2004) and are in concrete tanks. During this time they were given a prophy- Downloaded by [Department Of Fisheries] at 23:34 28 February 2013 reared for human consumption (Ingram 2009; Rowland 2009), lactic salt bath treatment (5 g/L) and were inspected for gill and for stock enhancement programs, and to support a sport fish- skin parasites. Southern Pygmy Perch Nannoperca australis and ery. For these reasons comprehensive data are required on the Unspecked Hardyhead Craterocephalus stercusmuscarum ful- potential susceptibility of finfish species to EHNV. vus were collected from the wild and transported to Narrandera Initial laboratory investigations revealed that EHNV was able Fisheries Centre where they were held and treated as above. to infect several species of fish, including both native and alien Trout Cod Maccullochella macquariensis and Macquarie Perch sympatric species in southeastern Australia (Langdon 1989). fingerlings were obtained from Snobs Creek Hatchery, Victo- The Australian native species Silver Perch, Murray Cod, Golden ria Department of Primary Industries, Alexandra, Victoria, and Perch Macquaria ambigua ambigua, Mountain Galaxias Galax- were hatchery-bred in 2009. Macquarie Perch captured from ias olidus, and Macquarie Perch Macquaria australasica were Cataract Dam (Hawkesbury–Nepean catchment, New South susceptible to EHNV when challenged in the laboratory via Wales, were transported to the university on the same day. immersion or intraperitoneal (IP) injection (Langdon 1989). Before experimentation, fish were housed in variably sized Also, evidence was provided for a potential carrier state in glass or Perspex aquaria (average habitable volume, 100 L), Redfin Perch (Langdon and Humphrey 1987), Rainbow Trout plastic tanks (320 L), or 1,000-L tanks, with one species (Langdon et al. 1988), Atlantic Salmon Salmo salar (Langdon per tank. Water quality was monitored and conditions were 68 BECKER ET AL.

TABLE 1. Origin, size, age, and prior knowledge of epizootic hematopoietic necrosis virus (EHNV) susceptibility for ﬁsh species selected for the study. NFC = Narrandera Fisheries Centre, NSW = New South Wales, DPI = Department of Primary Industries.

Arrival date at Prior data on University of Mean TL EHNV Species Sydney Source Age (mm) ± SE susceptibility Silver Perch Bidyanus November 2007 NFC hatchery >1 year, juvenile 103.8 ± 1.4 Susceptible and bidyanus carriera Unspecked Hardyhead June 2009 Wild from the Juvenile, adult 33.3 ± 0.70 Unknown Craterocephalus Murray River, near stercusmuscarum fulvus Robinvale, NSW Eastern Mosquitofish February 2009 Wild from NFC pond Adult 31.8 ± 0.59 Susceptiblea Gambusia holbrooki Carp Gudgeon October 2007 Wild from the Adult 33.3 ± 0.65 Unknown Hypseleotris spp. Murray River, NSW (Locks 7 and 8) Trout Cod Maccullochella February 2010 Victoria DPI ∼5 months, 54.7 ± 0.61 Unknown macquariensis hatchery juvenile Murray Cod March 2008 NFC hatchery <1 year, juvenile 121.5 ± 1.4 Susceptible and Maccullochella peelii carriera peelii Golden Perch Macquaria May 2009 NFC hatchery ∼4 months, 34.5 ± 0.50 Susceptible and ambigua ambigua juvenile carriera January 2010 NFC hatchery 37.7 ± 0.35 Macquarie Perch May 2009 Wild from Cataract >1 year, adult 100.9 ± 3.1 Susceptiblea Macquaria australasica Dam, NSW February 2010 DPI Victoria <5 months, 55.4 ± 0.45 hatchery juvenile Murray–Darling February, April NFC hatchery Juvenile 46.5 ± 2.6 Unknown Rainbowfish 2009 Melanotaenia fluviatilis Southern Purple-spotted February, April NFC hatchery Adult 59.4 ± 0.66 Unknown Gudgeon Mogurnda 2009 adspersa Southern Pygmy Perch June 2009 Wild from Adult 50.2 ± 0.77 Unknown Nannoperca australis Coppabella Creek, NSW Redfin Perch Perca Quarterly, Wild from NSW Juvenile Susceptiblea

Downloaded by [Department Of Fisheries] at 23:34 28 February 2013 fluviatilis 2008–2010 dams and (<55 mm) and impoundments adult (>55 mm) from areas with and without a history of EHNV Dewfish (Freshwater June 2009 Private Juvenile 98.3 ± 1.3 Unknown Catfish) Tandanus hatchery-reared tandanus aFrom Langdon (1989).

maintained within acceptable ranges throughout the duration of nitrate (<40 mg/L). Water quality was maintained with high the holding period with daily to weekly records of temperature capacity biological ﬁltration units and there was a weekly to (Table 2), pH (7.0–7.2), dissolved oxygen (>6.5 mg/L), salinity fortnightly (or more often as required) exchange of 10–30% (1–3‰), ammonia (<0.50 mg/L), nitrite (<0.25 mg/L), and of total water volume. Artiﬁcial lighting was provided as 12 h INFECTION OF FRESHWATER FISH WITH EHNV 69

TABLE 2. Overview of the challenge trials to determine the susceptibility of freshwater ﬁshes commonly found in the Murray–Darling Basinto epizootic hematopoietic necrosis virus (EHNV). Viral dose was either ﬁnal viral concentration in bath or total dose for IP injection.

Water temperature Virus dose Trial Trial start date Species challenged Challenge type (◦C) (TCID50/mL) 1 November 2007 Carp Gudgeon (pilot trial) 1-h immersion bath in 19–22 835 12 L 2 January 2008 Carp Gudgeon, Redfin Perch 1-h immersion bath in 4 L 19–22 75 × 103 Batha 19–22 160 3 April 2008 Murray Cod, Redfin Perch Batha 18–21 40 4 August 2008 Silver Perch, Murray Cod, Redfin Batha 19–21 190 Perch 5 November 2008 Silver Perch, Murray Cod, Redfin Batha 20–24 40 Perch 6 March 2009 Eastern Mosquitofish, Southern Batha 21–22 60 Purple-spotted Gudgeon, Murray–Darling Rainbowfish, Redfin Perch 7 May 2009 Eastern Mosquitofish, Southern Batha 21–22 50 Purple-spotted Gudgeon, Murray–Darling Rainbowfish, Redfin Perch 8 July 2009 Golden Perch, Unspecked Batha 19–22 5 Hardyhead, Dewfish, Southern IP injection 106.5 Pygmy Perch, Macquarie Perch, Redfin Perch 9 February 2010 Golden Perch, Macquarie Perch, Batha 22–24 500 Trout Cod, Dewfish, Redfin Perch aTank volume was 100 L.

light (0700–1900 hours) : 12 h dark. Fish were fed to satiety culture infectious dose with 50% endpoint)/mL (Langdon 1989; daily or up to five times weekly with their preferred feed Whittington and Reddacliff 1995) (Table 2). Due to their being type selected from frozen bloodworms, commercial flakes, of sufficient size and having significant conservation value, a native fish commercial pellets, earthworms, and lamb liver. All random selection of surviving Macquarie Perch (n = 5) and laboratory-based animal experimentation was carried out with Dewfish Tandanus tandanus (also known as Freshwater Cat- the prior approval of the University of Sydney Animal Ethics fish) (n = 28) received 106.5 TCID50 of EHNV by IP injection Committee. or were given a sham injection. Up to 12 Redfin Perch were Experimental challenge model.—Experimental infection tri- exposed to EHNV as a positive control for the viral inoculum Downloaded by [Department Of Fisheries] at 23:34 28 February 2013 als were conducted in 100-L aquaria that contained aerated in each experiment. All surviving fish were monitored once water and had an established external biological filtration unit. to twice daily for 28–35 d after exposure for clinical signs of For each trial, water temperatures were maintained within 3◦C EHNV. This was based on a maximum observed incubation by controlling the air temperature of the room (Table 2). Wa- period of 28 d in Redfin Perch (Whittington and Reddacliff ter quality was monitored twice weekly and pH maintained at 1995). Depending on size, moribund and dead fish were sam- 7.0–7.2, dissolved oxygen at >6.5 mg/L, and nitrogenous wastes pled for virus isolation, histopathology, or both for evidence of + − (e.g., NH3 ,NO2 ) at 0 mg/L, and water changes were made as EHNV infection. All surviving fish from the tanks challenged needed. Fish were randomly allocated to the experimental tanks with EHNV were tested at the end of each trial. All mortalities and allowed to reacclimate to the new tank for 3–7 d before in the control tanks during the trial were tested for evidence exposure to the virus. A minimum of two replicates per species of EHNV infection, and approximately three to five individuals were exposed to EHNV or given a sham exposure (Table 2). from each control tank were selected for testing on the final ob- The EHNV inoculum for the bath challenge was grown in servation day. Also, at least three apparently healthy individuals cell culture (see below), harvested, and added to the water at from each consignment of fish in addition to any mortalities the inlet or delivered as an immersion challenge in a smaller arising from transportation or during the holding period prior to volume with the minimum dose equal to one TCID50 (tissue experimentation were tested for EHNV. 70 BECKER ET AL.

Virus.—The EHNV isolate 86/8774, originally isolated from Histopathology.—For larger fish, portions of posterior kid- clinically affected Rainbow Trout (Langdon et al. 1988), was ney, liver, and spleen were excised and fixed in 10% neutral used for all experimental trials. From a single stock batch, buffered formalin for at least 24 h. Small whole fish were fixed in EHNV was propagated in Bluegill Lepomis macrochirus fry 10% neutral buffered formalin, placed in a 12.5% (w/v) EDTA (BF-2) cells or Fathead Minnow Pimephales promelas (FHM) (pH 7) decalcification solution for 24 h, and then transferred cells grown in Eagle’s minimum essential medium (MEM) sup- back to formalin. Tissues were processed routinely through plemented with 0.43 g/L sodium bicarbonate, 4.75 g/L HEPES graded ethanol solutions, embedded in paraffin wax, sectioned at [4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid], and 10% 5 µm, and stained with hematoxylin and eosin (H&E). Sections fetal bovine serum (FBS), and incubated at 22◦C. After de- containing evidence of characteristic tissue degeneration and struction of the monolayer, the culture medium was collected, necrosis with or without viral inclusion bodies were recorded centrifuged at 1,000 × g for 10 min to remove cell debris, and as positive (Langdon and Humphrey 1987; Reddacliff and then used as inoculum as previously described (Whittington and Whittington 1996). Reddacliff 1995). In general this procedure took place on the Immunohistochemistry.—Immunoperoxidase stains were same day, except in initial trials where the inoculum was kept conducted on EHN histopathology-positive tissue sections using at 4◦C for a maximum of 10 d before use (Whittington and a commercial kit (DAKO) containing peroxidase-labeled strep- Reddacliff 1995). An endpoint titration assay was conducted in tavidin and a mixture of biotinylated anti-rabbit–anti-mouse– BF-2 cells grown in MEM as above to accurately determine the anti-goat immunoglobulins as link antibodies (Reddacliff and TCID50 applicable to each challenge trial, but the results were Whittington 1996). The primary antibody (affinity-purified rab- not known for 3 weeks after commencement of each trial. bit anti-EHNV Lot No. M708, OIE Reference Laboratory, Uni- Virus reisolation and confirmation by PCR.—Kidney, liver, versity of Sydney) and a negative control reagent (nonimmune and spleen were target organs, pooled either after dissecting rabbit serum) were used at a dilution of 1:1,500 (Whittington the viscera or by using whole fish with head and tail removed and Deece 2004). Positive results were characterized by the (for fish < 40 mm TL). Tissues were placed in sterile micro- deposition of a brown stain associated with areas of necrosis centrifuge tubes and stored at −80◦C if not processed imme- (Reddacliff and Whittington 1996). diately. Each sample was weighed, and a 9 × ratio of weight Classification of the infection status and susceptibility of in- to volume of homogenizing medium (HM; minimum essential dividual fish and host species.—The status of individual fish medium supplemented with 200 IU/mL penicillin, 200 µg/mL after challenge with EHNV was defined based on laboratory streptomycin, and 5 µg/mL Fungizone) was added. Tissues were test results (Table 3). Operational definitions were applied at the prepared by grinding in a chilled mortar and pestle with sterile species level for the infection model: (1) if one or more individ- sand and HM and then clarified by centrifuging at 900 × g for ual fish in a species was infected (e.g., positive virus isolation) 10 min in a microcentrifuge. A 200-µL aliquot of the clarified or diseased (e.g., positive histopathology), then the species was homogenate was removed for DNA extraction and a 500-µL classified as potentially susceptible; (2) if one or more individ- aliquot was prepared for virus isolation. It was further diluted ual fish in a trial was classified as a carrier, then the species 1:4 (v:v) in HM, passed through a 0.22-µm, low protein-binding, was classified as being a potential carrier host. A species could syringe-end filter, and used to inoculate BF-2 cells in suspension be both susceptible and a carrier host. If none of the individual in 24-well tissue culture plates. Briefly, the cells were prepared fish were classified as being susceptible or a carrier, then the by resuspending 80–90% confluent cell monolayers in MEM species was considered to be not susceptible or could recover supplemented with 10% FBS, 200 IU/mL penicillin, 200 µg/mL from infection during the observation period. Potentially sus- streptomycin, and 5 µg/mL Fungizone, to a concentration of 2 ceptible, carrier, and not susceptible statuses recognize the lim- Downloaded by [Department Of Fisheries] at 23:34 28 February 2013 × 105 cells/mL. In duplicate, 150 µL of each sample were inoc- itations of experimental infection models to reflect the situation ulated directly into a 1.5-mL cell suspension, and the cells were in nature. allowed to attach and form a monolayer. Cells were incubated at 22◦C and examined for development of a cytopathic effect (CPE). If CPE developed, the infected tissue culture supernatant RESULTS (TCSN) was harvested, with 150 µL passaged into a fresh cell A total of seven previously untested fish species and an addi- suspension, and a 200-µL aliquot was reserved for DNA extrac- tional six species with previously reported EHNV susceptibility tion to confirm the presence of EHNV by PCR. Any wells with (Langdon 1989) were included in a series of nine experimental cells that exhibited no CPE after 7–10 d were frozen at −20◦C challenge trials (Table 2). At least one positive control Redfin overnight, then thawed, and 150 µL of TCSN was passaged by Perch was determined to be infected with EHNV or diseased well-to-well transfer into fresh cells. This was repeated after a based on virus isolation or histopathology (Reddacliff and Whit- further 7–10-d incubation to confirm samples as negative. The tington 1996) for each of the challenge trials (except for the pilot DNA was extracted from 200 µL of TCSN using the HighPure trial where no Redfin Perch were available) (Figure 1; Table 4). Viral Nucleic Acid Extraction Kit (Roche) and examined by con- This indicated that all of the viral inoculae were viable. There ventional (OIE 2003) or quantitative PCR (Jaramillo et al. 2012). was no evidence of EHNV infection in any individuals of any INFECTION OF FRESHWATER FISH WITH EHNV 71

TABLE 3. Classiﬁcation of individual ﬁsh status following experimental challenge with epizootic hematopoietic necrosis virus (EHNV) based on virus isolation, histopathology, or a combination of both laboratory tests.

STATUS OF INDIVIDUAL FISH

Alive on final sample day Moribund or died prior to final sample

T Not infected or recovered from Not infected and presumed to have Both tests negative a E infection died from other causes S T Positive virus isolation, Infected and presumed diseased histopathology not done R E Carrier or S Positive histopathology, incubation time exceeds U virus isolation not done observation period Infected and diseased L T Both tests positive a S

aEach ﬁsh was examined for evidence of an EHNV infection by viral isolation, histopathology, or both of these diagnostic tests.

species prior to experimentation or in any of the individuals the two challenge trials so that the lowest infective dose tested sampled from the sham-exposed control tanks. was 40 TCID50/mL (Table 2). Therefore, the Silver Perch was classified as a species potentially susceptible to infection with Species Susceptible to EHNV EHNV because the virus was isolated from affected fish, and it Silver Perch.—Mortalities in the juvenile Silver Perch over was also considered a potential carrier species as the virus was the duration of the study were 40% of 107 in the challenged isolated from surviving fish at day 35 (Table 3). Finally, EHNV group compared with 9.2% of 163 in the sham-exposed tanks was not isolated from any dead or surviving fish in the control (Table 4). EHNV was isolated from two (5.1%) Silver Perch group. that died on day 16–18 postexposure and from an additional 2 Eastern Mosquitofish.—Mortalities in the adult Eastern of 64 (3.1%) of the surviving fish sampled on day 28. The in- Mosquitofish over the duration of the experiment were 28% dividual fish that tested positive were equally distributed across of 176 in the challenged group compared with 7.1% of 98 in the Downloaded by [Department Of Fisheries] at 23:34 28 February 2013

FIGURE 1. Liver of a Redﬁn Perch that was exposed to EHNV and developed multifocal necrosis. (A) Control without anti-EHNV antibody; (B)thesame section as (A), but with the anti-EHNV antibody included. L = zone of liquefaction, D = zone of hepatocellular degeneration and necrosis (immunoperoxidase staining). [Figure available online in color.] 72 BECKER ET AL.

TABLE 4. The susceptibility of freshwater fishes from the Murray—Darling Basin to epizootic hematopoietic necrosis virus (EHNV) following experimental challenge. Fish species in this study were classified as potentially susceptible (S), carrier (C), or not susceptible (NS); see text for definitions.

Proportion of ﬁsh with an EHNV positive result

Challenged Challenged ﬁsh that died survivorsa Control ﬁshd

Total Deaths/total Virus Dead/ Positive/ Challenge replicate number Virus isola- total total Species Species method tanks exposed Histology isolationb Histologyc tion observed tested susceptibility

Silver Perch Bath 4 43 / 107 0 / 4 2 / 39 2 / 64 15 / 163 0 / 63 S, C Unspecked Bath 2 9 / 96 1 0 / 8 Not donef 0 / 80 3 / 32 0 / 13 NS Hardyhead insufficient tissuee Eastern Bath 5 47/167 0/5 + 3 17 / 39 0 / 3 7 / 105 7 / 98 0 / 37 S, C Mosquitofish insufficient tissuee Carp Gudgeon Bath 4 50 / 95 0 / 4 0 / 25 0 / 4 0 / 26 21 / 59 0 / 59 NS Trout Cod Bath 4 4 / 115 0 / 1 0 / 3 Not donef 0/103 4/34 0/4 NS Murray Cod Bath 6 9 / 61 0 / 9 0 / 52 5 / 55 0 / 38 NS Golden Perch Bath 5 34 / 109 2 0 / 32 Not donef 0/67 1/27 0/7 NS insufficient tissuee Macquarie Bath 5 26/135 1/1 + 6 0 / 21 Not donef 0 / 92 32 / 132 0 / 32 S Perch insufficient tissuee IP injection 1 3 / 3 1 / 1 + 2 1/3 1/2g 0/2g S insufficient tissuee Murray– Bath 5 27 / 166 1 / 3 5 / 24 0 / 3 0 / 112 19 / 74 0 / 23 S Darling Rainbowfish Southern Bath 4 3/71 0/2 1 0/3 0/56 3/40 0/10 NS Purple- insufficient spotted tissuee Gudgeon Southern Bath 2 2 / 68 0 / 2 Not donef 0/60 10 / 48 0 / 10 NS Pygmy Perch Redfin Perch Bath 15 59 / 93 43 / 49 49 / 52 1 / 8 4 / 29 S, C Dewfish Bath 3 0 / 65 Not donef 0/35 0/80 0/3 S (Freshwater Catfish) IP injection 2 13 / 15 4 / 7 2 / 12 Not donef 0/2 1/13g 0/13g S

aSurvivors were euthanized at least 28 d after exposure. Downloaded by [Department Of Fisheries] at 23:34 28 February 2013 bPositive virus culture was confirmed using PCR. cPositive histology sections were confirmed with immunohistochemistry for EHNV. dConspecifics were given a sham exposure and the EHNV status confirmed by virus isolation. eInsufficient amount of tissue was available for processing for a valid laboratory testing result. fHistology samples were not processed as there was no evidence of infection with EHNV based on the more sensitive diagnostic test, virus isolation. gControl fish were survivors of the bath challenge that received a sham injection.

control group (Table 4). During the observation period, EHNV (Table 3). The virus was not isolated from dead or surviving fish was isolated from 17 of 39 (44%) mosquitofish that died, the in the control group. earliest virus-associated mortality occurring on day 7 and the Macquarie Perch.—Overall mortalities in the juvenile Mac- latest on day 27 postexposure. From the surviving mosquitofish, quarie Perch sourced from Cataract Dam, New South Wales, the virus was isolated from 7 of 105 (7%) individuals and the and fingerlings obtained from a hatchery were 28% of 135 for lowest infective dose tested was 50 TCID50/mL. Therefore, the bath-challenged tanks but were similar in the control tanks Eastern Mosquitofish was classified as being both potentially (Table 4). One Macquarie Perch (exposed at 5 TCID50/mL) susceptible and a potential carrier species for EHNV infection that died on day 35 showed microscopic lesions characteristic INFECTION OF FRESHWATER FISH WITH EHNV 73

FIGURE 2. Ranavirus infection in a Macquarie Perch liver. (A)Therewas FIGURE 3. (A) The spleen of a Murray–Darling Rainbowfish infected with a large irregular zone of hepatic necrosis (H&E staining); (B) there was lo- EHNV demonstrating the locally extensive granular staining of splenic tissue cally extensive granular staining of numerous hepatic lobules consistent with mainly associated with sinusoids (immunoperoxidase staining); (B)afocus widespread viral replication (immunoperoxidase staining). [Figure available of hepatocellular necrosis with granular staining of hepatocytes and associated online in color.] cells was observed in the liver of Dewfish (immunoperoxidase staining). [Figure available online in color.] of an EHNV infection; other tests were not conducted on this Downloaded by [Department Of Fisheries] at 23:34 28 February 2013 individual (Figure 2). However, the virus was not isolated from replicate tanks in the first trial (trial 6, Table 2) were EHNV any fish that died in the challenged group or from the surviving positive. However, in a subsequent trial in three replicate tanks, challenged fish, or any of the sham-exposed control group. On the virus was isolated from 23% of 22 fish that died and an- day 35 postexposure, three surviving fish were injected IP with other individual that died had lesions consistent with EHNV 100 µL of virus culture and two were given a sham injection. infection (Figure 3A). The lowest infective dose tested was One injected Macquarie Perch that died was infected and dis- 50 TCID50/mL. None of the exposed survivors showed any played microscopic lesions indicative of an EHNV infection. evidence of EHNV infection and were considered not to be in- Macquarie Perch was therefore defined to be a species that is fected or to have recovered from infection within the 35-d period potentially susceptible to EHNV, but there was no evidence of (Table 3). Therefore, Murray–Darling Rainbowfish was classi- a carrier state. fied as a potentially susceptible species and there was no evi- Murray–Darling Rainbowfish.—Overall mortalities in two dence of a carrier state. separate challenge trials on juveniles were 16% of 166 in the Dewfish (Freshwater Catfish).—A total of 65 juvenile Dew- group challenged by bath exposure and 26% in the controls fish were bath challenged with EHNV, and no mortality was (Table 4). However, there was variation in outcome between observed during the observation period (Table 4). There was trials. None of the Rainbowfish that were challenged in two no evidence of EHNV based on virus isolation at day 35 74 BECKER ET AL.

postexposure in 35 surviving fish or in any of the sham-exposed water temperature, poor water quality, and skin lesions due fish. Of the remaining surviving Dewfish, 15 were subjected to parasites are known risk factors for outbreaks of disease to an IP injection with 100 µL of virus culture and 13 were (Whittington et al. 1994, 1999; Whittington and Reddacliff given a sham injection. Four of the IP-injected fish that died 1995). The species tested in this study are all considered to had microscopic lesions consistent with EHNV infection (Fig- be temperate freshwater fish and, for most of them, they were ure 3B; Table 3). Positive virus isolation results were obtained once considered widespread throughout the Murray–Darling from one additional Dewfish that was exposed to the virus via Basin (Lintermans 2007). Although there is limited knowledge IP injection. Therefore, Dewfish were classified as a species that regarding the preferred thermal range for each species, several is potentially susceptible to EHNV infection, but there was no of the tested species (e.g., Murray Cod, Unspecked Hardyhead, evidence of a carrier state. Carp Gudgeon, Golden Perch) are known to spawn within the range of water temperatures used in this study. The objective of Species Not Susceptible to EHNV this study was to expand our knowledge of the host range for Individuals from seven species of freshwater fish did not EHNV, but it was not possible to explore all factors in species show any clinical signs of EHN and EHNV could not be iso- susceptibility. For example, it would be useful in future studies lated from any of the exposed fish. Therefore, Murray Cod, Un- to test age-related differences in susceptibility. Furthermore, specked Hardyhead, Carp Gudgeon, Trout Cod, Golden Perch, there would be merit in determining the duration of infection Southern Purple-spotted Gudgeon, and Southern Pygmy Perch and the possibility that the potential carriers identified in this were classified as species that are potentially not susceptible study may eventually succumb clinically, recover completely to EHNV following a bath challenge, or the fish were able to from infection, or develop an intermittent viral shedding recover during the challenge period under these experimental state. Like EHNV, the ranavirus, Ambystoma tigrinum virus conditions (Table 4). (ATV) is not universally virulent in its tiger salamander A. tigrinum host (Brunner et al. 2005). Different clutches of tiger DISCUSSION salamanders varied in susceptibility to experimental infection Ranaviruses are recognized as major pathogens of econom- with ATV, suggesting that host genetics and life history stage ically and ecologically important species as they are gener- were important risk factors (Brunner et al. 2005). ally not host-specific and may result in outbreaks associated Recognizing that experimental laboratory challenge models with high levels of mortality (Chinchar 2002; Whittington et al. may overestimate or underestimate susceptibility relative to the 2010). The movement of animals out of their original geograph- situation in nature, conservative conclusions were made about ical range brings together new hosts and potential pathogens, potential resistance or susceptibility and the carrier state. A rel- facilitates host-switching, and continues the cycle for disease atively uniform challenge model was used, but it is difficult to emergence for a pathogen (Peeler et al. 2011). Evidence sug- compare it with those used by others on all criteria. However, the gests that ranaviruses have undergone at least three events of dose of virus is probably a vital factor and can be measured. An host switches followed by speciation in the new host during inoculum with a greater viral concentration will attain greater evolutionarily recent times with jumps occurring from fish to densities before the host can mount an adequate immune re- frogs, fish to salamanders, and frogs to reptiles (Jancovich et al. sponse, thereby reducing the chance of clearance (Brunner et al. 2010). The origin of EHNV is uncertain, but to date it has been 2005). Conversely, it has been suggested that an inocululm with isolated only from finfish. In this study a range of finfish species a lower starting virus density is more easily cleared or modulated were challenged with EHNV to determine their susceptibility by the hosts’ immune system, leading to chronic or sublethal and the results confirmed an earlier report (Langdon 1989) that infections (Brunner et al. 2005). Previously, it was shown that Downloaded by [Department Of Fisheries] at 23:34 28 February 2013 Silver Perch, Macquarie Perch and Eastern Mosquitofish were an EHNV infection would result in adult Redfin Perch exposed susceptible to EHNV following an experimental bath challenge. at a dose of <0.1 TCID50/mL (Langdon 1989; Whittington Two additional species, Dewfish and Murray–Darling Rainbow- and Reddacliff 1995), whereas Rainbow Trout required at least fish were identified as being potentially susceptible to EHNV. 103 TCID50/mL (Langdon et al. 1988). The lowest dose in the Unspecked Hardyhead, Carp Gudgeon, Trout Cod, Southern present study was a bath challenge of 5 TCID50/mL, and the Purple-spotted Gudgeon, and Southern Pygmy Perch were con- range was generally between 40 and 50 TCID50/mL. Recent sidered to be potentially not susceptible to EHNV following the European models applied doses of 104 TCID50/mL to test the bath challenge. Northern Pike, Pikeperch, and Black Bullhead for EHNV sus- It is well recognized that susceptibility to viral infection is ceptibility (Bang Jensen et al. 2009, 2011; Gobbo et al. 2010). not always a fixed trait for a species and is influenced by host, Even within a given experimental model, variation between tri- pathogen, and environmental factors. For example, embryonic als and between replicates within trials can be problematical. For and larval life stages of the northern leopard frog Rana pipiens example, a total of five tanks of Murray–Darling Rainbowfish, were susceptible to infections with frog virus 3, the type all sourced from the same wild population, were bath-challenged species for the genus Ranavirus, whereas the adult frogs did in two separate trials with individuals from only two tanks show- not develop disease (Tweedell and Granoff 1968). For EHNV, ing evidence of EHNV infection, for unknown reasons. INFECTION OF FRESHWATER FISH WITH EHNV 75

In recent reports, histopathology was not always per- in older fish less than 1% (1 of 135) were affected. The same formed during disease outbreaks or in experimental trials with trend was observed for Silver Perch with 30% of 3-month-old ranaviruses (Bigarre´ et al. 2008; Ariel and Bang Jensen 2009; fish dying with EHNV (Langdon 1989) compared with <4% Ariel et al. 2009; Gobbo et al. 2010). The current study classified (4 of 107) of older fish (>100 mm TL) becoming infected in the susceptibility of species based on the results from individ- the present study. In the current study Eastern Mosquitofish had ual fish from one or both diagnostic tests (histopathology and a lower susceptibility to EHNV with 14% (24 of 167) of fish virus isolation). This was because it was not biologically (due succumbing to infection or developing a carrier state, compared to fish size) or logistically (degradation of tissue from dead fish) with 90% percent mortality observed by (Langdon 1989). Per- possible to have paired histopathology and virology samples for haps this was associated with differences in source of fish (New each individual fish. Although virus isolation is considered to South Wales versus Victoria) between the studies and the asso- have a higher sensitivity compared with histopathology, it was ciated genetic variation (Smith et al. 1989; Congdon 1995), or useful to confirm, through microscopic examination, that fish age of fish (which could not be determined accurately, but East- were affected by the disease, EHN. ern Mosquitofish have a relatively short life cycle; Lintermans Many of the species used in the experimental infections were 2007). sourced from the wild as artificial propagation techniques were There are 10 different families represented among the 14 not available, and these fish had unknown history and health fish hosts identified to date as being potentially susceptible to status. Mortality that was not associated with EHNV was ob- EHNV (Langdon 1989; Bang Jensen et al. 2009, 2011; Gobbo served in some species, particularly the wild-sourced fish such et al. 2010; the present study), but susceptibility does not appear as Eastern Mosquitofish, Carp Gudgeon, and Macquarie Perch to correlate with taxonomic relationships. The families Perci- collected from Cataract Dam. It is important to note that all in- chthyidae and Percidae contain the greatest number of known dividual fish that died during the observation period in all trials, potential hosts (Golden Perch, Murray Cod, Macquarie Perch, regardless of exposure to EHNV or apparent cause of death, and Australian Bass Macquaria novemaculeata, and Redfin were sampled for evidence of EHNV infection. Perch and Pikeperch, respectively). A species from one fam- Although Dewfish were unaffected by EHNV following a ily may be susceptible while another species from the same bath challenge, different life stages (i.e., younger juveniles) or related families within the same order is not. For example, or fish under higher levels of environmental stress may have within the order Atheriniformes, the Murray–Darling Rainbow- increased susceptibility because one-third (5 of 15) of the IP- fish (family Melanotaeniidae) and the Unspecked Hardyhead injected catfish succumbed to EHNV. Dewfish are farmed and (family Atherinidae) differed in susceptibility following bath translocated widely across southeastern Australia and, due to challenge with the virus. Even within a genus, closely related declines in wild populations, a conservation stocking program species differed in susceptibility depending on the challenge for natural waterways is being considered, which could increase model employed. Within the genus Macquaria, only the Mac- the risk of spread of EHNV. The Murray–Darling Rainbowfish quarie Perch was susceptible following bath challenge, while is a subtropical omnivorous species that was once widespread Australian Bass and Golden Perch were susceptible only after across the Murray–Darling Basin, but it has declined in some an IP challenge (Langdon 1989). Based on these observations areas (Lintermans 2007). It has high conservation value and policy development for prevention or control of EHNV infection in the southern parts of its range its main threats include pre- should not rely solely on taxonomic relatedness as a predictor dation from two other EHNV hosts, Redfin Perch and Eastern of susceptibility. Mosquitofish (Lintermans 2007). Also in agreement with Lang- Anecdotal and documented outbreaks of EHN in wild Redfin don (1989), Murray Cod and Golden Perch were not susceptible Perch occur most years in impoundments and dams in New Downloaded by [Department Of Fisheries] at 23:34 28 February 2013 to the virus following a bath challenge. However, both species South Wales and the Australian Capital Territory (Whittington were previously shown to be highly susceptible (100% mortal- et al. 2010). The current study clearly demonstrated the poten- ity) after IP injection (Langdon 1989) and should be considered tial of the virus to infect several freshwater fish species found potentially susceptible to EHNV. across the Murray–Darling Basin. The widespread dispersal of Overall, it appeared that EHNV was less virulent in the hatchery-produced fish into adjacent or more distant catchments present experimental model than in previous studies, but the for sport fishery enhancement and conservation stocking pro- reasons for this are uncertain. There were differences in ages of grams may be a risk for the translocation of EHNV. Of par- the fish between the study of Langdon (1989) and the present ticular interest are Silver Perch, Murray Cod, and Dewfish as study. As most outbreaks in wild Redfin Perch affect age-0 fish they are considered to be emerging aquaculture species. As the (Langdon and Humphrey 1987), older fish may be less suscep- aquaculture industries grow and new facilities are built on new tible to the virus. Similar observations have been made in tiger waterways, the potential exists to expand the current distribution salamanders infected with ATV with the case mortality vary- of EHNV. It is important to consider the implementation of pre- ing between 40% and 100% (Brunner et al. 2005). Langdon screening of susceptible species before moving them outside the (1989) reported that all 20 3-month-old Macquarie Perch died EHNV endemic zone. It is also prudent to continue to assume with EHNV following bath exposure, while in the present study that infection with EHNV is a factor potentially detrimental to 76 BECKER ET AL.

both farmed and wild native fish populations in southeastern Langdon, J. S. 1989. Experimental transmission and pathogenicity of epizootic Australia (Lintermans 2007). haematopoietic necrosis virus (EHNV) in Redfin Perch, Perca fluviatilis L., and 11 other teleosts. Journal of Fish Diseases 12:295–310. Langdon, J. S., and J. D. Humphrey. 1987. Epizootic haematopoietic necrosis, ACKNOWLEDGMENTS a new viral disease in Redfin Perch, Perca fluviatilis L., in Australia. Journal Funding for this project was provided by the Murray–Darling of Fish Diseases 10:289–297. Basin Authority. The authors gratefully acknowledge the tech- Langdon, J. S., J. D. Humphrey, and L. M. Williams. 1988. Outbreaks of an nical support provided by the staff at the New South Wales EHNV-like iridovirus in cultured Rainbow Trout, Salmo gairdneri Richard- son, in Australia. Journal of Fish Diseases 11:93–96. Department of Primary Industries (S. Thurston, M. McLennan), Langdon, J. S., J. D. Humphrey, L. M. Williams, A. D. Hyatt, and H. A. and the University of Sydney (C. Kristo, R. Mauer, K. Williams, Westbury. 1986. First virus isolation from Australian fish: an iridovirus-like J. McIvor). Trout Cod and Macquarie Perch were kindly donated pathogen from Redfin Perch, Perca fluviatilis L. Journal of Fish Diseases by B. Ingram (Department of Primary Industries, Victoria). 9:263–268. Lintermans, M. 2007. Fishes of the Murray-Darling basin: an introductory guide. Murray-Darling Basin Authority, Canberra, Australia. REFERENCES OIE (World Organization for Animal Health). 2003. Manual of diagnostic tests Ariel, E., and B. Bang Jensen. 2009. Challenge studies of European stocks of for aquatic animals 2003, 4th edition. OIE, Paris. Redfin Perch, Perca fluviatilis L., and Rainbow Trout, Oncorhynchus mykiss Peeler, E. J., B. C. Oidtmann, P.J. Midtlyng, L. Miossec, and R. E. Gozlan. 2011. (Walbaum), with epizootic haematopoietic necrosis virus. Journal of Fish Non-native aquatic animals introductions have driven disease emergence in Diseases 32:1017–1025. Europe. Biological Invasions 13:1291–1303. Ariel, E., J. Kielgast, H. E. Svart, K. Larsen, H. Tapiovaara, B. Bang Jensen, and Reddacliff, L. A., and R. J. Whittington. 1996. Pathology of epizootic R. Holopainen. 2009. Ranavirus in wild edible frogs Pelophylax kl. esculentus haematopoietic necrosis virus (EHNV) infection in Rainbow Trout (On- in Denmark. Diseases of Aquatic Organisms 85:7–14. corhynchus mykiss Walbaum) and Redfin Perch (Perca fluviatilis L.). Journal Bang Jensen, B., A. K. Ersbøll, and E. Ariel. 2009. Susceptibility of pike of Comparative Pathology 115:103–115. Esox lucius to a panel of ranavirus isolates. Diseases of Aquatic Organisms Reno, P. W. 1998. Factors involved in the dissemination of disease in fish 83:169–179. populations. Journal of Aquatic Animal Health 10:160–171. Bang Jensen, B., R. Holopainen, H. Tapiovaara, and E. Ariel. 2011. Suscep- Rowland, S. J. 2009. Review of aquaculture research and development of the tibility of pike–perch Sander lucioperca to a panel of ranavirus isolates. Australian freshwater fish Silver Perch, Bidyanus bidyanus. Journal of the Aquaculture 313:24–30. World Aquaculture Society 40:291–324. Barrett, J. 2004. Introducing the Murray-Darling basin native fish strategy and Smith, M. H., K. T. Scribner, J. D. Hernandez, and M. C. Wooten. 1989. initial steps towards demonstration reaches. Ecological Management and Demographic, spatial, and temporal genetic variation in Gambusia. Pages Restoration 5:15–23. 235–257 in G. K. Meffe and F. F. Snelson Jr., editors. Ecology and evolution Bigarre,´ L., J. Cabon, M. Baud, F. Pozet, and J. Castric. 2008. Ranaviruses of live-bearing fishes (Poeciliidae). Prentice-Hall, Englewood Cliffs, New associated with high mortalities in catfish in France. Bulletin of the European Jersey. Association of Fish Pathologists 28:163–168. Trust, T. J. 1986. Pathogenesis of infectious diseases of fish. Annual Review of Brunner, J. L., K. Richards, and J. P. Collins. 2005. Dose and host characteristics Microbiology 40:479–502. influence virulence of ranavirus infections. Oecologia 144:399–406. Tweedell, K., and A. Granoff. 1968. Viruses and renal carcinoma of Rana Chinchar, V.G. 2002. Ranaviruses (family Iridoviridae): emerging cold-blooded pipiens: V. effect of frog virus 3 on developing frog embryos and larvae. killers—brief review. Archives of Virology 147:447–470. Journal of the National Cancer Institute 40:407–410. Congdon, B. C. 1995. Unidirectional gene flow and maintenance of genetic Whittington, R. J., J. A. Becker, and M. M. Dennis. 2010. Iridovirus infections in diversity in mosquitofish Gambusia holbrooki (Teleostei: Poeciliidae). Copeia finfish: critical review with emphasis on ranaviruses. Journal of Fish Diseases 1995:162–172. 33:95–122. Fenichel, E. P., J. I. Tsao, M. Jones, and G. J. Hickling. 2008. Fish pathogen Whittington, R. J., and K. Deece. 2004. Aquatic animal health subprogram: de- screening and its influence on the likelihood of accidental pathogen introduc- velopment of diagnostic and reference reagents for epizootic haematopoietic tion during fish translocations. Journal of Aquatic Animal Health 20:19–28. necrosis virus of finfish. Fisheries Research and Development Corporation, Gobbo, F., E. Cappellozza, M. R. Pastore, and G. Bovo. 2010. Susceptibility of University of Sydney, Sydney, Australia. Black Bullhead Ameiurus melas to a panel of ranavirus isolates. Diseases of Whittington, R. J., A. Philbey, G. L. Reddacliff, and A. R. Macgown. 1994. Downloaded by [Department Of Fisheries] at 23:34 28 February 2013 Aquatic Organisms 90:167–174. Epidemiology of epizootic haematopoietic necrosis virus (EHNV) infection Hedrick, R. P. 1998. Relationships of the host, pathogen, and environment: in farmed Rainbow Trout, Oncorhynchus mykiss (Walbaum): findings based implications for diseases of cultured and wild fish populations. Journal of on virus isolation, antigen capture ELISA and serology. Journal of Fish Aquatic Animal Health 10:107–111. Diseases 17:205–218. Ingram, B. A. 2009. Culture of juvenile Murray Cod, Trout Cod and Macquarie Whittington, R. J., and G. L. Reddacliff. 1995. Influence of environmental tem- Perch (Percichthyidae) in fertilised earthen ponds. Aquaculture 287:98–106. perature on experimental infection of Redfin Perch (Perca fluviatilis)and Jancovich, J. K., M. Bremont, J. W. Touchman, and B. L. Jacobs. 2010. Evi- Rainbow Trout (Oncorhynchus mykiss) with epizootic haematopoietic necro- dence for multiple recent host species shifts among the ranaviruses (family sis virus, an Australian iridovirus. Australian Veterinary Journal 72:421–424. Iridoviridae). Journal of Virology 84:2636–2647. Whittington, R. J., L. A. Reddacliff, I. Marsh, C. Kearns, Z. Zupanovic, and R. Jaramillo, D., A. Tweedie, J. A. Becker, A. Hyatt, S. Crameri, and R. J. Whitting- B. Callinan. 1999. Further observations on the epidemiology and spread of ton. 2012. A validated quantitative polymerase chain reaction assay for the epizootic haematopoietic necrosis virus (EHNV) in farmed Rainbow Trout detection of ranaviruses (Family Iridoviridae) in fish tissue and cell cultures, Oncorhynchus mykiss in southeastern Australia and a recommended sampling using EHNV as a model. Aquaculture 356/357:186–192. strategy for surveillance. Diseases of Aquatic Organisms 35:125–130. This article was downloaded by: [Department Of Fisheries] On: 28 February 2013, At: 23:36 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Aquatic Animal Health Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/uahh20 Guide for Authors Version of record first published: 07 Feb 2013.

To cite this article: (2013): Guide for Authors, Journal of Aquatic Animal Health, 25:1, 77-81 To link to this article: http://dx.doi.org/10.1080/08997659.2013.769836

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material. Journal of Aquatic Animal Health 25:77–81, 2013 c American Fisheries Society 2013 ISSN: 0899-7659 print / 1548-8667 online DOI: 10.1080/08997659.2013.769836

JOURNAL OF AQUATIC ANIMAL HEALTH Guide for Authors Review Process Submitted papers will be critically reviewed by at least two experts in the relevant discipline(s) and evaluated by one of the Editorial Policy journal’s associate editors. A manuscript may be returned to its author without review if it is judged to be of poor quality or The Journal of Aquatic Animal Health serves the North inappropriate for this journal. American and international communities of scientists concerned In submitting a paper, you are stipulating that, except where with the health of aquatic organisms. We encourage the sub- explicitly indicated otherwise, all of the statements, data, and mission of papers dealing with the causes, effects, treatments, other elements reflect your own work and not that of others. and prevention of diseases of marine and freshwater organisms, References to the work of others should be properly cited; exact particularly fish and shellfish. We also welcome manuscripts de- quotations from other sources should be in quotation marks. scribing biochemical and physiological investigations into fish Failure to follow these requirements may result in rejection of health that relate to assessing the impacts of both environmental the paper and, in extreme cases, restrictions on publishing in and pathogenic factors. this journal. Authors have the option of anonymity; if they wish to exercise it, they should prepare their manuscripts accordingly. Manuscript Submission and Review Review of manuscripts relies on volunteers and can be a fairly lengthy process. However, we strive to get decisions to authors Manuscript Categories in 9–12 weeks. If revisions are requested, authors should make them promptly, normally within 30 days of receiving the editor’s Manuscripts may be submitted in any of the following cate- decision (short extensions will be allowed if there are justifiable gories: (1) Articles are reports of substantial, controlled research delays). If a revision is not received within the allowed time, that will be judged on their scientific merit. Critical reviews of the paper will be considered withdrawn; late revisions will be timely topics will also be considered in this category. Arti- treated as new submissions and may have to go through the cles should ordinarily not exceed 5,000 words, excluding refer- review process again. ences and tables (about 20 double-spaced manuscript pages), but longer ones will be considered. (2) Communications are shorter Publication Charges papers based on more restricted study objectives, sometimes without extensive statistical data, but with sound biological ob- Publication charges are US$100 per printed page and will servations; promising work that may lead to additional in-depth be billed when the paper is in proof. Full and partial subsidies studies; thorough testing of a technique; or case histories. Such are available to voting members of the American Fisheries So- papers will be evaluated as much for their practical utility as for ciety who certify that grant or agency funds are not available. their scientific quality. Communications should generally not Manuscript reviews are not affected by requests for subsidies; Downloaded by [Department Of Fisheries] at 23:36 28 February 2013 exceed 3,000 words (about 12 double-spaced manuscript pages). however, at least one author must be (or become) an AFS mem- (3) Comments are critiques of papers published by this journal, ber by the time that a paper is published. Every paper published responses to which will be invited from the original authors; in the journal is subject to a $30 fee to offset handling costs. brief presentations of experiences or additional data related to Authors will receive a free PDF of the published article and may previously published papers; or short discussions of technical purchase reprints of their paper when they receive their proofs. issues pertinent to the aquatic animal health community. Manuscript Preparation Submission Procedures

Manuscripts and associated correspondence should be sub- Components mitted at the journal’s online submission and tracking site, http://mc.manuscriptcentral.com/jaah (this site may also be ac- A typical manuscript will have the following components: cessed through the Publications section at the American Fish- Title page.—The title page should give the title of the pa- eries Society’s Web site, www.fisheries.org). Detailed instruc- per and the name(s) and complete mailing address(es) of the tions, including acceptable file formats, are available at the site. author(s). In addition to accurately reflecting the content of the

77 78 GUIDE FOR AUTHORS

paper, the title should be short (preferably no more than 12 Discussion.—A good discussion provides broad syntheses words) and to the point. See a recent issue of the journal for the and stresses the relevance of the paper. In it authors should indi- format to use for authors’ names and addresses. A suggested cate the significance of their research, how it relates to current running head (shortened version of the title) may also be in- knowledge, and any avenues that it suggests for further research. cluded on the title page. Keywords are not used in this journal, Informed speculation is acceptable as long as it is clearly identi- however, and so should not be included. fied as such. Authors should avoid merely restating their results Abstract.—Articles and communications require abstracts; and/or (re)summarizing the literature. comments do not. The abstract should consist of one paragraph Acknowledgments.—In this section authors may acknowl- (up to 300 words for an article and up to 200 words for a com- edge the sources of their funding and thank those who con- munication) that concisely states why and (generally) how the tributed directly to the project or the preparation of the study was done as well as what the results were and what they manuscript. Dedications and acknowledgment of emotional sup- mean. It should not simply outline the contents (e.g., avoid state- port from family and friends are not appropriate. If all authors ments to the effect that such-and-such is presented) or present are employees of the U.S. Government, this section should state the methods in detail. Citations and footnotes are not allowed in that the mention of specific products does not constitute en- abstracts, and abbreviations should be used sparingly. Because dorsement by their agency. abstracts tend to be more widely read than complete papers, au- References.—References should be selected with a view thors should take care to make them comprehensive, clear, and to relevance and availability, with preference given to peer- interesting. reviewed publications that are widely available. Internal reports, Introduction.—The introduction should provide a context for papers presented at conferences, articles in preparation, and so the work to be reported. In doing so, it should present at least a forth should be treated as unpublished and cited like personal general overview of previous literature on the subject, guiding communications (i.e., parenthetically in the text alone). Authors the reader to the paper’s purpose and importance. should obtain written permission to cite such material. Common Methods.—Descriptions of the methods employed in the reference formats are given below; a more complete list is given study should be detailed enough to enable readers to repeat in chapter 8 of the AFS style guide, which is available at the it. Previously published descriptions may be cited in lieu of pre- AFS Web site as well as the manuscript submission site. senting complete new ones provided that the sources are readily Footnotes.—Text footnotes should be kept to a minimum. available (in general, avoid citations to theses, dissertations, Typically, they are used to report changes of address for au- agency reports, and similar sources in this instance). If more thors, identify additional sources of data, or explain technical than one method was used or a particular method entails a series nomenclature (e.g., the structures of fatty acids). of major steps, present each method or step in a separate sub- Tables.—In general, tables should be designed to present section. Appropriate tables and figures can reduce the need for related information as simply and directly as possible. A good detailed verbal descriptions of methods. Papers focusing entirely rule of thumb is to establish the point(s) that the table is on techniques do not require a separate section on methods. intended to make, then to select the information required to do Results.—As a rule, it is preferable to present detailed re- that and determine the most logical order in which to present it. sults in tables and/or figures and to devote the text to summary Detailed guidelines for the preparation of tables may be found statements and analyses. Display data in tables if numerical pre- in chapter 12 of the AFS style guide, but a few of the more cision is important, in figures if trends are paramount. Although important ones may be mentioned here: the presentation of a large amount of raw data is generally not meaningful, data should not be refined to the point that the reader Downloaded by [Department Of Fisheries] at 23:36 28 February 2013 cannot verify the analyses or use the information for other pur- 1. We prefer to print tables in “portrait” orientation but will poses. In presenting the results of statistical tests, report the type allow ones in “landscape” orientation as long as they take of test, the test statistic, the degrees of freedom, and the signif- up no more than one page. icance level (P -value). Although the value 0.05 is commonly 2. Tables that are too long or too wide to fit on one page can be used as the threshold in hypothesis testing, we have no specific carried over to a facing page, but authors should try to avoid requirements in this area; in the interest of providing useful creating tables that span more than two pages. In general, information, authors should report all P -values. It is very im- very large tables should appear as supplementary tables in portant that statistical designs and models be appropriate for the the online version of the article only. studies in which they are used; we encourage authors to have a 3. Tables should contain only three horizontal rules (lines)— statistician review their work before submitting a paper for pub- one before the column headings, one after those headings, lication. Lastly, statistical results should be presented in biolog- and one at the bottom of the table—and no vertical rules. ically meaningful terms rather than in purely statistical jargon. 4. As a rule, captions should be detailed enough that the table Authors who describe hybridomas or cell lines should state can be understood apart from the text (if there is more than that they are willing to share their material with other interested one table with the same general structure, only the first needs investigators. to have a detailed caption). Captions should be written so GUIDE FOR AUTHORS 79

as to stress the purpose of the table and not merely list its to be easily readable (300 dpi or more); and be consistent in contents in a mechanical way. the use of lettering, line widths, and other graphic elements. In 5. There should be only one set of column headings. If the addition, they need to conform to AFS style. It is particularly information to be presented seems to require more than important to remember that most figures will be reduced by up that, the table should be redesigned (e.g., by switching the to 50% when printed and thus need to be designed with this in rows and columns) or split into two or more tables. mind. We recommend that authors use a copier to reduce each 6. Bold, centered headings may be used within the body of the figure to the width of one or two printed columns (3.50 and table to distinguish different types of data as long as they 7.25 inches, respectively), depending on the dimensions of the do not conflict with the column headings. particular figure, and verify that all elements are still legible. The 7. Only the first letter of a row or column heading should be following are particularly problematical: bold type (which tends capitalized (along with words or symbols that would be to blur), italic type (which tends to become less visible), dashed capitalized in ordinary text). lines (which tend to appear continuous), dotted lines (which 8. The data within the body of the table should not be crowded; tend to disappear entirely), and shading in which the different if need be, blank rows can be inserted to separate data into shades are not distinct enough. Additional guidelines for the logical groups or provide guides for the eye. preparation of figures may be found in the AFS style guide. 9. Significant differences among multiple means should be There is no additional charge for the reproduction of color indicated by lowercase letters, beginning with the letter “z” figures in the online version of the paper. However, all figures (“z” may mark either the highest or the lowest value[s], but in the print version will be in black and white unless specific subsequent letters have to follow suit); there should be no arrangements have been made with the Journals Department omissions in the sequence of the letters. The letters should to cover the additional costs of color printing. Because color be set on the same lines as the values to which they pertain printing is expensive, authors are advised not to use color to (not as superscripts) and be separated from those values by distinguish phenomena when other means (different shading, single spaces. symbols, and so forth) are adequate. 10. Values less than 1.00 should be preceded by zeroes (e.g., Digital files in EPS, TIFF, and PSD formats are preferred; 0.78). figures should be submitted as separate files rather than being 11. Values need not be reported to all significant digits if a lesser imbedded in text files. number of digits conveys the information in a meaningful Mathematical and statistical expressions.—Chapter 4 of the way. AFS style guide covers the treatment of these expressions in 12. Footnotes should be indicated by superscripted lowercase detail, but a few general points may be mentioned here: letters, beginning with the letter “a”; the letters may appear in the row and column headings as well as the body of the 1. Symbols representing variables and parameters should be table but not in the caption. The footnotes per se should be italicized if they consist of single letters in the Latin alphabet K F listed on separate lines at the bottom of the table. (e.g., and ). All other symbols except Greek letters may be italicized or not, provided that the treatment is consistent Figure captions.—Figure captions should appear together (e.g., CPUE or CPUE); Greek letters should not be italicized. in a list rather than separately with each figure (however, the 2. Natural logarithms may be expressed as loge or ln; logarithms number of the figure and the name of the corresponding au- with other bases should identify the base (e.g., log10). thor should be given outside the image area of each figure for 3. Long equations should be “broken” at logical points, nor- purposes of identification). Like table captions, figure captions mally after an operator such as a plus or minus sign. Downloaded by [Department Of Fisheries] at 23:36 28 February 2013 should generally be detailed enough that the figure can be under- 4. Definitions of variables and parameters may be run into the stood apart from the text. To the extent possible, however, panel text if only a few such terms are involved. If there are a descriptions, (full) variable names, units of measure, legends, number of them or they are used in more than one equation, and so forth should be included in the figure itself rather than a list is preferable (see section 4.8 of the style guide). ± in the caption; in no case should they be given in both places. 5. Avoid the expressions “the mean length was 45.2 3.84 ± ± Different panels may be designated “A,” “B,” and so forth, but it mm” and “the mean ( SD) length was 45.2 3.84 mm” is preferable to give them substantive labels (e.g., “Treatment” because they are at best awkward and at worst inaccurate. ± ± and “Control”). Use the expressions “the mean SD length was 45.2 3.84 Figures.—Figures include visual materials such as graphs, mm” or “the mean length was 45.2 mm (SD, 3.84)” instead. maps, diagrams, and photographs. Figures have proved to be one of the most troublesome aspects of the publishing process. As the Style and Format Journals Department has only limited ability to modify figures, they frequently have to be sent back to the authors for correction. Published articles represent the culmination of research ef- At the most fundamental level, figure design should follow forts, often lengthy and highly sophisticated ones. To do those certain commonsense principles: figures should be as simple efforts justice, however, the articles must be well written; and straightforward as possible; have a high enough resolution poorly written articles not only place an unnecessary burden on 80 GUIDE FOR AUTHORS

readers, they also cast doubt on the quality of the research itself. parentheses. See chapter 9 of the AFS style guide for additional Although some people naturally write better than others, most information on the treatment of species’ names; the accepted can develop the ability to write well through practice and atten- plurals of fish names are given in Appendix C of the guide. tion to detail. The introduction to the AFS style guide should be Names of parasitic diseases should follow “Standardized a particularly valuable resource in this regard; in a few pages, it Nomenclature of Animal Parasitic Diseases (SNOAPAD)” by identifies the errors in composition mostly commonly encoun- Kassaietal.(Veterinary Parasitology 29:299–326, 1988). Our tered in the papers submitted to AFS journals and shows how to standard sources for chemical and enzyme names are the current correct them. We also encourage authors to have other fisheries editions of the Merck Index (Merck & Co., Rahway, New Jersey) professionals critique their initial drafts with respect to presen- and Enzyme Nomenclature (Academic Press, San Diego, tation as well as substance. Authors whose native language is California), respectively. The preferred treatment of allozymes not English should make a point of having English speakers is noted in the article “Gene Nomenclature for Protein-Coding review their manuscripts before submission. Loci in Fish” by J. B. Shaklee et al. (Transactions of the In writing for AFS journals, authors are also expected American Fisheries Society 119:2–15, 1990). Additional infor- to follow certain style conventions pertaining to capitaliza- mation on the treatment of these and other technical matters tion, spelling, punctuation, mathematical expressions, technical may be found in chapter 11 of the AFS style guide. terms, and so forth. For instance, we require that the letter P Manuscript format.—As an aid to reviewers and editors, au- (indicating the degree of statistical significance) be capitalized thors should as well as italicized, whereas some journals require that it be lowercased. Although some of the more important style con- 1. use a line spacing of at least space and a half for all com- ventions are noted below, all of them are discussed in detail in ponents of the paper, including the title page, footnotes, and the AFS style guide. Authors would be well advised to become tables; familiar with the main elements of AFS style and to consult the 2. number all pages sequentially and provide continuous line guide frequently in preparing their manuscripts. numbering beginning with the title page; Resources for authors.—As suggested above, the principal 3. use a 12-point font throughout; resource on matters of style is the AFS style guide. Authors 4. use three levels of headings, as follows: for the major sec- may also find it helpful to consult the Chicago Manual of Style tions of the paper (Methods, Results, Discussion, Acknowl- (University of Chicago Press, Chicago) and Scientific Style and edgments, and References), type them flush left with initial Format (Council of Science Editors, Chicago), though the AFS letters capitalized (except for prepositions and conjunctions) style guide always takes precedence. in ordinary type, preceded by “” (e.g., Methods); The standard resource for word usage and spelling is Web- for subsections in Results and Discussion, type them flush ster’s Third New International Dictionary, as updated by the left with initial letters capitalized in ordinary type, preceded latest edition of Merriam-Webster’s Collegiate Dictionary.Ap- by “” (e.g., Treatment 1); and for subsections in pendix A of the AFS style guide shows the proper way to spell Methods and sub-subsections in Results and Discussion, run many of the terms used in fisheries writing (some of which are them into the text with only the initial letter of the first word not in the dictionary), including terms for which our preferred capitalized, all words italicized, and followed by a period spelling differs from that in the dictionary. and a long dash (e.g., Sampling design.—); and The standard resource for the common and scientific names 5. turn off automatic hyphenation and justification. of North American fish species is the current edition of Common and Scientific Names of Fishes from the United States, Canada, General style conventions.—A detailed presentation of AFS Downloaded by [Department Of Fisheries] at 23:36 28 February 2013 and Mexico (American Fisheries Society, Bethesda, Maryland). style is beyond the scope of these guidelines. The following For other aquatic species, authors should follow the companion conventions, however, are so general as to apply to virtually publications World Fishes Important to North Americans and every paper: Common and Scientific Names of Aquatic Invertebrates from the United States and Canada (the volumes Mollusks, Deca- 1. Only symbols and abbreviations included in Webster’s dic- pod Crustaceans, and Cnidaria and Ctenophora are currently tionaries or listed at the end of these guidelines (as well as available in the latter series). at the back of each printed issue of the journal) may be used In most cases, scientific names should be included only at without definition. All others should be defined at first use first mention in the abstract and text; full common names (e.g., (e.g., index of biotic integrity [IBI]). Abbreviations should “Coho Salmon” rather than simply “Coho”) should be used not be introduced unless they are used at least two more elsewhere. The format for the first mention is times. 2. As a rule, only metric units may be used. The only exceptions Coho Salmon Oncorhynchus kisutch, are a few quantities that are typically expressed only one way (e.g., g [of medication]/lb [of feed]). in which all parts of the common name are capitalized and the 3. Single-digit numbers should be spelled out unless they are scientific name follows the common name but is not given in used with units of measure or in conjunction with larger GUIDE FOR AUTHORS 81

values (e.g., 8 Walleyes and 16 Saugers). Numbers with four Books or more digits should contain commas; those less than 1.00 should be preceded by zeroes. Krebs, C. J. 1989. Ecological methodology. Harper and Row, New York. 4. Ratios involving two values or units of measure should be indicated by forward slashes (e.g., 0.30 g/d); ratios involving Chapters in books three such terms should be indicated by negative exponents Omernik, J. M. 1995. Ecoregions: a spatial framework for environmental man- −1 −1 (e.g., 0.01 g·g ·d ). agement. Pages 49–62 in W. S. Davis and T. P. Simon, editors. Biological 5. Ages of ﬁsh should be expressed by Arabic numerals and assessment and criteria: tools for water resource planning and decision not contain plus signs (e.g., a ﬁsh is age 1 [not age 1+]from making. Lewis Publishers, Boca Raton, Florida. the January 1 after it hatches to the following December 31). Use “age 0” or “young of the year” instead of YOY. Government reports 6. Dates may be expressed as either day–month–year (e.g., 11 Reports that are issued on a regular basis are treated much January 2011) or month–day–year (e.g., January 11, 2011), like articles in journals (the principal difference being that page provided that the same format is used throughout the paper. numbers should not be given); other reports are treated like Note that the term “Julian day” does not mean day of the books: year and should not be used in that context. 7. Time should be expressed in terms of the 24-hour clock Everest, F. H., C. E. McLemore, and J. F. Ward. 1980. An improved tri-tube followed by the word “hours” (e.g., 1435 hours rather than cryogenic gravel sampler. U.S. Forest Service Research Note PNW-350. 2:35 p.m.). [journal format] USEPA (U.S. Environmental Protection Agency). 1998. Water quality crite- Reference formats.—Text citations should conform to the ria and standards plan: priorities for the future. USEPA, 822-R-98-003, author–year system. Examples of common types are as follows: Washington, D.C. [book format]

(Johnson 1995) (Johnson and Smith 1996) Electronic publications (Johnson et al. 1997, 1998) [three or more authors] (Johnson et al. 1999, 2001; Smith 2000) Formats for references to electronic publications are still (Johnson 2000a, 2000b) (Johnson, in press) evolving. The important thing is to give the reader enough in- (E. M. Johnson, National Marine Fisheries Service, personal communication) formation to be able to locate the reference easily. If a book or report is only available online or is available Note that with one exception citations should be listed in in print form but was accessed online, the reference should be chronological order; the exception is that all citations to the same formatted as follows: author(s) should be grouped together (see the fourth example above). Baldwin, N. A., R. W. Saalfield, M. R. Dochoda, H. J. Buettner, and R. L. In reference lists, references should be in strict alphabetical Eshenroder. 2000. Commercial fish production in the Great Lakes, 1867– order by authors’ last names; if there are two or more references 1996. Great Lakes Fishery Commission, Ann Arbor, Michigan. Available: www.glfc.org/databases/. (September 2000). with the same authors, those references should then be listed chronologically. All authors must be named in references. The month and year in parentheses indicate when the site Detailed information on reference formats may be found in was last accessed. chapter 8 of the AFS style guide. The more common types are If a journal is available in print form, authors should use as follows: the standard reference format even if they accessed the article Downloaded by [Department Of Fisheries] at 23:36 28 February 2013 Articles in journals online. If a journal is only available electronically, the format depends on the way(s) in which articles are designated. Two Pace, M. L., and J. D. Orcutt. 1981. The relative importance of protozoans, ro- possible formats are as follows: tifers, and crustaceans in a freshwater zooplankton community. Limnology and Oceanography 26:822–830. Gallagher, M. B., and S. S. Heppell. 2010. Essential habitat information for age- 0 rockfish along the central Oregon coast. Marine and Coastal Fisheries: Note that (1) except for the first author, authors’ initials come Dynamics, Management, and Ecosystem Science [online serial] 2:60–72. before their last names; (2) only the first word of the title of the DOI: 10.1577/C09-032.1 article is capitalized (along with any other words that would be Kimmerer, W. J. 2004. Open-water processes of the San Francisco Estuary: capitalized in ordinary text); and (3) the name of the journal is from physical forcing to biological responses. San Francisco Estuary and given in full. Watershed Science [online serial] 2(1):article 1.