Acute effects of the antibiotic oxytetracycline on the bacterial community of the grass shrimp, Palaemonetes pugio


  • Miguel Uyaguari,

    Corresponding author
    1. Department of Environmental Health Sciences, Arnold School of Public Health, University of South Carolina, 921 Assembly Street, Office 502 PHRC, Columbia, South Carolina 29208, USA
    • Department of Environmental Health Sciences, Arnold School of Public Health, University of South Carolina, 921 Assembly Street, Office 502 PHRC, Columbia, South Carolina 29208, USA
    Search for more papers by this author
  • Peter Key,

    1. U.S. Department of Commerce, National Oceanic and Atmospheric Administration, National Ocean Service, Center for Coastal Environmental Health and Biomolecular Research, 219 Fort Johnson Road, Charleston, South Carolina 29412
    Search for more papers by this author
  • Janet Moore,

    1. U.S. Department of Commerce, National Oceanic and Atmospheric Administration, National Ocean Service, Center for Coastal Environmental Health and Biomolecular Research, 219 Fort Johnson Road, Charleston, South Carolina 29412
    Search for more papers by this author
  • Kirby Jackson,

    1. Department of Epidemiology and Biostatistics, Arnold School of Public Health, University of South Carolina, Columbia, South Carolina 29208, USA
    Search for more papers by this author
  • Geoffrey Scott

    1. Department of Environmental Health Sciences, Arnold School of Public Health, University of South Carolina, 921 Assembly Street, Office 502 PHRC, Columbia, South Carolina 29208, USA
    2. U.S. Department of Commerce, National Oceanic and Atmospheric Administration, National Ocean Service, Center for Coastal Environmental Health and Biomolecular Research, 219 Fort Johnson Road, Charleston, South Carolina 29412
    Search for more papers by this author

  • Published on the Web 4/28/2009.


The toxicity of oxytetracycline (OTC) was evaluated in adult grass shrimp, Palaemonetes pugio. Initially, static acute (96 h) toxicity tests were conducted with shrimp exposed from 0 to 1,000 mg/L OTC. A calculated lethal concentration 50% value of 683.30 mg/L OTC (95% confidence interval 610.85–764.40 mg/L) was determined from these tests, along with a lowest-observable-effect concentration of 750 mg/L and no-observable-effect concentration of 500 mg/L. Moreover, chronic sublethal effects of OTC exposure on grass shrimp intestinal bacterial population were assessed using doses from 0 to 32 mg/L OTC. The total viable counts in digestive tract content had levels between 5.2 and 1 × 104 colony-forming units per gram of tissue at times 0 and 96 h, respectively. Aeromonas hydrophila were the most resistant isolates (27.78%) to OTC exposure. Vibrio alginolyticus showed significant positive growth following exposure to OTC, whereas other bacterial species abundance declined over time. A total of 268 bacterial isolates were screened using antibiotic resistance analysis from a library containing 459 isolates. Among the tested isolates from the OTC treatments, 15.4% were resistant to OTC and 84.6% were OTC sensitive. Oxytetracycline was generally not consistently quantifiable with liquid chromatography-mass spectroscopy technique in shrimp homogenates. The only peak detected was at the 32 mg/L dose of OTC at 96 h. Nevertheless, OTC had a significant biological effect on the bacterial population. Antibiotic resistance to five other antibiotics (penicillin G, sulfathiazole, trimethoprim, trimethoprim and sulfamethoxazole, and tetracycline) was strongly associated with OTC exposures. The present study indicates that OTC toxicity effects in P. pugio and changes in the shrimp microbial community would only be expected under special circumstances.


The management of pharmaceutically active compounds in the aquatic environment has been recognized as one of the emerging issues in environmental chemistry [1]. Discharges from numerous sources such as wastewater treatment plants and confined animal feeding operations may result in their input into the environment. In the United States, more than 50 million pounds of antimicrobials are produced every year, with approximately 40% destined for veterinary use, including antibiotics for nontherapeutic purposes, such as to promote growth [2]. The ability of these compounds to alter microbial community structure facilitates the development of antibiotic-resistant human pathogens, as well as their potential to serve as indicators for the presence of resistant pathogens. Most antibiotics are poorly absorbed by humans and animals after intake, with approximately 25 to 75% of added compounds leaving the organisms as unaltered via feces or urine [3]. As aquaculture and agricultural industries expand, so do concerns of drug overuse and potential risks to humans and the environment due to residues in seafood, as well as the ecological impact of antibiotic residues in the environment and the development of antibiotic resistant bacterial strains in aquatic species [4].

Research has demonstrated that many compounds may enter the environment, disperse, and persist over a long period of time [5]. Antibiotics are developed to perform certain biological actions in humans, pets, and domestic animals, which may generate collateral effects on the bacterial population and other organisms in the environment, depending on the residual quantity. Antibacterial or antimicrobial resistance is a threat to the efficacy of antibiotics used to treat human infections [6]. The development of resistance to chemical agents by many bacterial pathogens has compromised traditional therapeutic regimens, potentially making human treatments more difficult [7]. In a study conducted in 30 states during 1999 and 2000, Kolpin et al. [5] found levels of oxytetracycline (OTC) in streams at concentrations ranging from 0.10 to 0.34 μL. Moreover, in 2003, Thurman et al. [8] reported antibiotics in effluents from fish hatcheries, where OTC values in the water column ranged from 0.17 to 10 μL. Systems using recirculating water supplies may observe substantially higher OTC concentrations than in adjoining surface waters [9]. The task to detect antibiotics in the environment becomes even more daunting when dealing with complex mixtures of different chemicals that may be in various states of degradation or transformation, as well as the drug affinity to bind sediment, water, or tissue [10].

Aquaculture effluents in the United States are regulated under the National Pollutant Discharge Elimination System as part of the Federal Water Pollution Control Act of 1972 and its subsequent amendments. However, the remarkable growth of the aquaculture industry since the 1980s has caused increasing numbers of states to consider developing regulatory statutes in U.S. waters [11]. Furthermore, the recent discovery of pharmaceuticals in the environment across the United States has raised the visibility and need for monitoring of antibiotics, as well as determining their potential impact on sensitive species [5,8].

The U.S. Food and Drug Administration and the Center for Veterinary Medicine currently have six drugs registered for legal use in U.S. aquaculture. These include only three antibiotics: OTC, florfenicol, and a combination drug containing sulfadimethoxine and ormetoprim. Among these drugs, OTC is the most widely used in aquaculture activities.

Various studies in both laboratory and field conditions have examined the fate and effects of OTC in the environment. Most of these studies have focused on the sediments, since 95% of the antibiotic in the feed can be released to the environment as the parent compound [12,13]. Most studies have generally concluded that deposition in the sediments is one of the primary fates of antibiotics in aquatic ecosystems. Researchers have measured OTC concentrations ranging from 0.4 to 495 μg of sediment [14,15]. Oxytetracycline problems may include persistence in sediment, chronic sediment exposure, and impacts on nontarget organisms, while the acute toxicity of OTC to aquatic animals appears to be low [16]. However, therapeutic doses of OTC cause sublethal effects including altered serum immunoglobulin levels in carp and suppression of the phagocytic response of fish macrophages in rainbow trout [17].

A few studies have provided evidence of the presence of antibiotics, specifically OTC, affecting nontarget organisms such as fish, oysters, clams, and crustaceans following discharge in the vicinity of aquaculture operations. Capone et al. [14] found concentrations less than 0.1 μg of OTC in oysters (Crassostrea gigas) and Dungeness crab, Cancer magister, while levels of 0.8 to 3.8 μg of OTC were detected in red rock crabs, Cancer productus. Other studies have determined OTC residues in mollusks up to 10.2 μg in soft tissue [18,19]. However, these studies evaluated only bioconcentration endpoints rather than chronic effects on aquatic organisms, such as determination of physiological and immunological effects on the microbial community present in these nontarget organisms. Clearly, researchers need to assess the impacts of antibiotics on chronic sublethal endpoints in aquatic organisms. This is of particular importance in addressing the toxicological hazards posed by pharmaceutical use of antibiotics in marine aquaculture, as well as environmental discharges from wastewater treatment plants and confined animal feeding operations.

The objective of this research was to study the ecotoxicological hazards of OTC in marine invertebrates, addressing both lethal and sublethal effects. The use of OTC may adversely affect bacterial populations in aquatic organisms and the environment. For the present study, the toxicity and potential effects of OTC were assessed using standardized ecotoxicological tests in crustaceans using the grass shrimp, Palaemonetes pugio, as a model organism. Key et al. [20] discussed the importance of grass shrimp in coastal ecosystems, which has led to the use of grass shrimp as possible sentinel organisms for toxicity testing, field population assessments, and biomarkers of exposure related to human activities on coastal systems.


Grass shrimp

Nongravid adult grass shrimp 2 to 3 cm long were caught by pushnet (net openings 25 cm high by 40 cm wide—1,009 cm2—with a mesh size of 4 mm). Shrimp were collected at Leadenwah Creek (32°38′51.00″N; 80°13′18.05″W), a pristine tidal tributary of the North Edisto River Estuary in Charleston, South Carolina, USA. This site has been used as a reference site in many ecotoxicological studies, and the physical-chemical parameters are well characterized [21]. For all testing described here, shrimp were acclimated 7 to 14 d before beginning the experiments in 76-L tanks at 25°C, with natural seawater adjusted with deionized water to 20 practical salinity units (psu) and a 16:8-h light:dark cycle. During this time, shrimp were fed Tetramin® Fish Flakes (Tetra).

OTC stocks

Analytical-grade OTC hydrochloride (≥95% purity, Sigma-Aldrich) was used for all grass shrimp exposures. To prepare the different concentrations used for the toxicity tests, a 10,000 mg/L solution was prepared daily in deionized water and then homogenized by sonication. Aluminum foil was wrapped around the volumetric flask containing the 100-ml stock solution to avoid photodegradation.

Acute toxicity tests

To determine grass shrimp toxicity to OTC, a total of five 96-h static renewal toxicity tests were performed (Table 1). All experiments were conducted in a Revco® environmental chamber at 25°C and a 16:8-h light:dark cycle. Acclimated animals were placed in 4-L glass jars filled with 2 L of seawater at 20 psu with supplemented aeration being maintained at or above 4 mg/L (60% saturation) for each aquarium. Grass shrimp were not fed during the tests. Water quality parameters including dissolved oxygen (mg/L), pH, temperature (°C), and salinity (psu) were measured at time 0 h and every 24 h in the control aquaria. Mortality was recorded every 24 h. Range finder test concentrations were conducted using a logarithmic scale (control, 1, 10, 100, and 1,000 mg/L). Each concentration in the definitive test was in geometric series and overlapped at least 50% of the next higher dose. The definitive tests had 10 shrimp per aquarium with three aquaria per concentration for a total of 30 shrimp per concentration [22]. For the microbiological and body burden analysis tests, a density of 15 animals/L was employed to provide tissue biomass for the samples needed.

Microbiological analysis of intestinal bacterial flora

Chronic sublethal effects of OTC exposure on grass shrimp intestinal bacterial population were evaluated in two separate tests. In the first one, P. pugio were exposed to higher concentrations of OTC: 0, 250, 500, and 750 mg/L. Lower concentrations were used in the second experiment: 0, 1, 16, and 32 mg/L OTC. These experiments were based on OTC toxicity tests reported for crustaceans and OTC minimum inhibitory concentrations (MICs) for environmental isolates [23–25]. All exposures were 96 h in duration (acute) for grass shrimp (e.g., a fraction of the life cycle) but are chronic (e.g., multigeneration) for grass shrimp microbial population.

Sampling was conducted at 24-h intervals. One aquarium per concentration was selected randomly every 24 h. Grass shrimp in the control and in the randomly selected exposures were rinsed five times with 10 ml of phosphate buffered saline (PBS) [26]. The last PBS rinse in each pool per concentration per day was used as negative control for microbiological analysis to assure that shrimp had been rinsed free of bacteria.

Then, the animals were ground up and homogenized at 8,000 rpm and 4°C with a PBS solution volume, proportional to sample weight (between 3 and 4 g). Remaining ground grass shrimp were stored at −80°C for subsequent chromatographic analysis.

Table Table 1.. Experimental design used in grass shrimp 96-h exposure tests
TestOxytetracycline concentration (mg/L)n
Acute toxicity tests  
First range finder0, 1, 10, 100, 1,0003
Second range finder0, 1, 10, 100, 1,0003
Definitive0, 62.5, 125, 250, 500, 750, 1,0003
Chronic tests  
First chronic sublethal tests0, 250, 500, 7507
Second chronic sublethal tests0, 1, 16, 327

Media preparation

Dilutions ranging from 10−1 to 10−3 using PBS were made for each homogenate in the different tests [27]. Samples were spread plated onto different types of bacterial media plates, including a more general T1N3 agar and five more selective agars for Vibrio vulnificus: vibrio vulnificus agar, modified cellobiose-polymyxin B-colistin agar, CHROMagar (DRG International), thiosulfate citrate bile sucrose agar (Difco™, Becton, Dickinson and Company), cellobiose-polymixin B-colistin agar. Unless otherwise indicated, all media and reagents were prepared according to various protocols [27–29] and the manufacturer media instructions.

Bacterial counts and isolate selection

General heterotrophic total bacterial numbers were determined by counting colonies on the T1N3 and selective media spread plates after incubation at 36°C for 24 h. Then, most probable numbers of bacteria for the homogenates were calculated according to the method reported by Johnson and Case and using bacterial counts in T1N3 media [30]. Typical Vibrio bacteria classified by color, shape, and abundance were picked from the selective agars using sterile toothpicks. Each isolate was transferred to an individual 2-ml sterile Biostor™ (National Scientific) vial with skirted base, containing 0.75 ml of tryptic soy broth with an extra 2% sodium chloride and 20% glycerol. The isolates were labeled, placed in plastic boxes, and stored at −80°C until further analyses could be performed.

Bacteria identification

Isolate identifications were based on analytical profile index (API) strips, which are standardized strips for identification of non-Enterobacteriaceae gram-negative rods (API® 20NE developed by bioMerieux). Bacteria isolated as described earlier were grown on T1N3 plates containing 3% NaCl and then inoculated in tubes containing 3 ml of 0.85% PBS.

The following reactions were tested: reduction of nitrates to nitrites, reduction of nitrates to nitrogen, indole production, glucose fermentation, arginine dihydrolase, urease, esculin and gelatin hydrolysis, β-galactosidase, carbohydrate assimilation, and cytochrome oxidase.

Strips were incubated for 24 and 48 h at 29 ± 1°C. Preliminary reactions were recorded after 24 h incubation, and definitive readings were made at 48 h using the API database (ver 6.0 for API 20NE). Pseudomonas aeruginosa American Type Culture Collection 27853 was used as a positive control for each batch of strips.

Antibiotic resistance analysis

Antibiotic resistance analysis (ARA), one of the most commonly employed nonmolecular microbial source tracking methods to identify and classify microbial resistance [31], was used to evaluate and characterize antibiotic resistance of bacteria isolated from the different OTC doses at different exposure times (0, 24, 48, 72, and 96 h).

Isolates were regrown from glycerol stocks on T1N3 agar and inoculated into 3 ml of 2.5% PBS solutions. For inoculation, 100 μl of emulsified saline solutions with adjusted turbidity to a 0.5 McFarland barium sulfate standard were diluted in 25 ml of 2.5% NaCl cation adjusted Mueller Hinton broth tubes (Difco). Panels were inoculated with the microbial suspension using a Renok® pipettor (Siemens Healthcare Diagnostics) and incubated 24 h at 36 ± 1°C.

The MIC for the isolates was determined for 26 antibiotics, including OTC, using MicroScan® Panels (Dade Behring) developed by National Oceanic and Atmospheric Administration (NOAA) researchers (J. Gooch, L.A. Reed, and the Ecotoxicology Modeling Group) and manufactured by Dade Behring specifically for NOAA. Protocols provided by the manufacturer were followed. The criterion to indicate whether an isolate was resistant was based on the MIC value for OTC greater or equal to 16 μml according to the Clinical Laboratory and Standards Institute (formerly National Committee on Clinical Laboratory Standards, MIC Testing Supplemental Tables, M100-S13 M7 in [23]).

After 24 h, the ARA panels were read using the Micro-Touch® system (Dade Behring) and results were recorded on spreadsheets (Excel® Microsoft Office XP). Five antibiotic resistant positive control strains were also tested for quality assurance or quality control: Escherichia coli 25922, Enterococcus faecalis 29212, GSTP 12, P. aeruginosa 27853, and Staphylococcus aureus 29213, selected from the NOAA isolate library.

Bioaccumulation of OTC in grass shrimp

Shrimp tissue levels of OTC were measured using a Thermo TSQ Quantum triple quad mass spectrometer (Thermo-Finnigan). The liquid chromatograph attached to this mass spectrometer was an Agilent 1100 system. This system consisted of a degasser (G1322A), a binary pump (G1312A), a thermostated autosampler (G1329A), and a column compartment (G1316A).

The Thermo-Finnigan Xcalibur software was used for running, collecting, and processing of data for this experiment. The mobile phase consisted of a gradient system with water (eluent A) and methanol (eluent B), each with 0.05% acetic acid for mass spectrometer ionization. The gradient system consisted of time 0 (100% A held for 3 min), from 3 to 15 min. A ramp from 100% A to 100% B was used. At 15.1 min, the liquid chromatograph was back to the initial conditions of 100% A held for 18 min. The flow was constant at 0.5 ml/min.

Analytical separations were performed using the Phenomenex Gemini C18 Column (100 × 2.0 mm, 5 μm particle size, 110 A pore size). The column was heated to 40°C within the Agilent column oven. Standards and samples were injected at a volume of 10 μl, and the sample tray temperature was set at 4°C. The TSQ Quantum was set to positive ion mode and signal ion monitoring at a mass of 462.20 units and a width of 1.00 units. The acquisition time for the mass spectrometer was 18 min. The limit of detection and quantification was 1 μg. The percent recovery was 65 ± 7.2%.

Figure Fig. 1..

Survival curves for grass shrimp exposed in the definitive toxicity tests for oxytetracycline (OTC). * = significantly different from controls (p < 0.05).

Data analysis

Toxicity tests. The 96-h lethal concentration 50% (LC50) determination and 95% confidence interval were calculated for the OTC results using Trimmed Spearman-Karber LC50 estimation [32]. In addition, survival curves, as well as no-observable-effect concentration and lowest-observable-effect concentration, were computed by the generalized linear model procedure in Statistical Analysis Software (SAS, ver 9.1.3 for Windows®). Dunnett's test was conducted to detect statistical differences between mortality in the different OTC concentrations and controls. A safety factor interval for OTC was calculated using no-observable-effect concentration and MIC values. This interval allowed comparison of the relative margin of safety for OTC in other species [25].

Bacterial analysis

The data results for API 20 NE® and ARA were entered in Microsoft Excel spreadsheets (Microsoft Office for Macintosh). Bacterial densities, measured in colony-forming units per gram, were transformed to common logarithms. Analysis of variance was applied to determine differences among the percentages of bacteria resistance observed (Tukey's test). The bacterial library obtained during the present study was analyzed by logistic regression analysis with a generalized logit procedure, Kruskal-Wallis test, and Pearson correlation coefficients using SAS statistical software for comparisons among treatments and antibiotics. A probability value (p) of 0.05 was assumed for all tests as a minimum level of significance.



Water quality toxicity test conditions. During collection and acclimatization, dissolved oxygen ranged from 3.44 to 6.70 mg/L, averaging 5.07 mg/L; during toxicity tests, dissolved oxygen ranged from 6.6 to 6.7 mg/L, averaging 6.63 mg/L (±0.4 mg/L). Temperature values for collection and acclimatization ranged from 22.1 to 25.7°C, and temperature in toxicity tests ranged from 23.8 to 24.2°C, averaging 24.1°C (±0.2°C). Salinities for collection and acclimatization ranged from 19.9 to 29.6 psu, while salinities in toxicity tests ranged from 19.7 to 21.2 psu, averaging 20.6 psu (±0.7 psu). The pH values ranged from 7.72 to 7.92, averaging 7.83 (±0.07). No significant variations were found in any of the water quality parameters during the OTC exposures, which would have had an effect on the shrimp survival since conditions were optimal for survival and growth.

Acute toxicity tests. Grass shrimp survival curves are depicted in Figure 1. Significant mortality (100%) was observed after 24 h in the 1,000 mg/L OTC concentration. Animals exposed to the other concentrations (62.5, 125, 250, 500, and 750 mg/L) did not show significant mortality compared to controls until 96 h of exposure. At 96 h, animals exposed to 750 mg/L OTC also had significantly (p < 0.05) higher mortality (40%) compared to controls (0%).

Results from the present study indicate that OTC toxicity to P. pugio is very low, with an LC50 of 683.30 mg/L OTC (95% confidence interval 610.85–764.40 mg/L OTC). This confirms the criterion described by Alderman and Hastings [16], who explained the low acute toxicity of OTC to aquatic animals due to the poor bioavailability of OTC.

Ferreira et al. [24] conducted 48-h toxicity tests of Artemia parthenogenetica exposed to OTC. An LC50 of 806 mg/L OTC (95% confidence interval 650.71–1,129.81 mg/L) for nauplii was similar to the values found in the present study. Furthermore, the no- and lowest-observable-effect concentrations determined by these authors had values as high as our findings: 500 and 750 mg/L OTC, respectively. These slight differences observed in the LC50 and no- and lowest-observable-effect concentration values are likely due to species differences between the two studies. Adult grass shrimp are generally less sensitive to OTC than other are crustaceans. In other studies of OTC effects on crustacean species, Williams et al. [25] and Baticados et al. [33] reported tests of five penaeid shrimp larvae stages (Penaeus vannamei and P. monodon) exposed to OTC and reported overall LC50 levels for all five shrimp larval stages ranging from 100 to 238.4 mg/L OTC. Even though these penaeid shrimp species were more sensitive to OTC than the species used in the present study, the safety factor determined by these authors had a range of 12.5 to 53, which were comparable to the values calculated in the current study (15.62–125).

Figure Fig. 2..

Total bacterial species identified in grass shrimp digestive tract (n = 268).


OTC effects on bacteria microflora. Bacterial counts were estimated for the two microbiology exposure tests. Bacterial numbers in the shrimp exposed to higher OTC concentrations (250, 500, and 750 mg/L) were inconsistent and after 48 h did not yield sufficient bacteria to be recovered from the plates. This suggests that exposures greater than 250 mg/L OTC cause significant mortality in intestinal bacteria of the grass shrimp. Similar patterns using high doses of OTC have been reported by Hansen et al. [34], who evaluated bacterial population response to OTC levels of 400 μg wet sediment under salmon cages; after 2 d of exposure, they found significant decreases of 50% of the microbial population. This indicates that levels at or above the acute no-observable-effect concentrations for grass shrimp produce a significant effect on digestive tract bacteria survival.

Given the general lack of acute toxicity effects of OTC as described previously, the present study focused on chronic effects on bacterial populations in the grass shrimp digestive tract as an approach to further evaluate OTC effects. Results of the second sublethal test at lower OTC concentrations (1, 16, and 32 mg/L) did not show significant bacterial mortality; however, more demonstrable effects on antibiotic resistance were observed. No marked differences appeared in total bacterial population densities for shrimp exposed to 1, 16, and 32 mg/L OTC and control (p > 0.05). Bacteria coming from shrimp homogenates tended to decrease proportionally over time in the OTC exposures and the control. The density of total bacteria colonies on T1N3 over time were 5.2 × 104 and 1 × 104 colony-forming units per gram of tissue at exposure times of 0 and 96 h, respectively.

Bacterial identification. A total of 268 bacteria (58.4%) were picked at random and then selected and analyzed from a bacterial library containing 459 isolates. These homogenized grass shrimp isolates were isolated and identified as A. hydrophila (19%), A. sobria (0.7%), Aeromonas spp. (1.9%), Brevundimonas vesicularis (3%), Chryseobacterium indologenes (1.1%), Pasteurella multocida (1.1%), Pasteurella spp. (1.1%), Pseudomonas fluorescens (0.7%), V. alginolyticus (26.1%), V. cholerae (5.6%), V. parahaemolyticus (3.7%), V. vulnificus (13.8%), Vibrio spp. (17.5%), and Weeksella virosa (3%), and four isolates were classified as unknown, representing 1.5% of the isolates (Fig. 2). The most abundant species were V. alginolyticus (26.1%), A. hydrophila (19%), V. vulnificus (13.8%), V. cholerae (5.6%), and V. parahaemolyticus (3.7%), accounting for more than 68% of all isolates. The predominance of Aeromonas and Vibrio spp. from the digestive tract in other species of shrimp isolates related to environmental and aquatic sources have previously been characterized and reported [35,36].

Figure Fig. 3..

Frequencies of oxytetracycline (OTC)–resistant bacterial isolates over time in grass shrimp intestinal fauna. * = significantly different (p < 0.05).

Bacterial acquired antibiotic resistance. Initially, no differences were observed between controls and OTC concentrations 1, 16, and 32 mg/L (p > 0.05) at time 0. For controls, bacteria isolated from shrimp had detectable 0% resistance during the entire 96-h exposure period; thus, the data suggest that observed resistance in bacteria was the direct result of exposure to OTC. After 24 h, grass shrimp isolates exposed to 1,16, and 32 mg/L OTC had a average resistant value (number of isolates that showed resistance per total number of isolates per time point and concentration) of 13%. Percentages of resistance results for 48 h of exposure showed values of 6.67% for 16 mg/L, while no isolates from the 1 mg/L exposure showed any resistance. At 48 h of exposure, the highest resistance was observed for the 32 mg/L OTC, with 73% of the bacteria exhibiting microbial resistance (e.g., MIC ≥ 16 mg/L OTC) at this time. Compared to the time 0 values, clear evidence in the 32 mg/L dose of OTC showed that bacteria had acquired resistance to OTC. After 72 h of exposure, a decrease is observed in the overall resistance (53.33%) at the 32 mg/L OTC dose, while the 1 and 16 mg/L OTC doses had isolates with a resistance mean value of 6.67%. Furthermore, a trend of decreased resistance over time in isolates exposed to 32 mg/L OTC was again observed at 96 h of exposure; at the end of the experiment, 50% of the isolates were resistant. After 96 h, no resistant isolates were observed in controls and grass shrimp exposed to 1 and 16 mg/L OTC. The overall frequency of resistance at 32 mg/L OTC concentration was significantly (p < 0.0056) higher than the frequencies determined in the lower OTC exposures (1 and 16 mg/L) and controls (Fig. 3).

This temporal trend of having an initial rapid increase in antibiotic resistance followed by a decreasing resistance over time during OTC exposure has been reported by Samuelsen [17] in sediments under controlled conditions. A higher resistance rate was reported (13.5%) after 7 d of exposure to OTC; it was followed by a lower rate at 80 d of exposure (8.2%). Furthermore, Herwig and Gray [37] determined similar microbial responses to OTC in marine microcosms; in a 10 d antimicrobial dosing period, 75 to 80% of the resistance was observed during the first 3 d, which then tended to decrease and stabilize at levels of 50 to 60% of resistance (posttreatment). The general pattern for microbial resistance in our study was an initial increase followed by a sudden decline observed during the first 48 h for all treatments. Variability may be related in part to feces interaction among the grass shrimp exposed to OTC. Even though grass shrimp were not fed during these exposures, defecation did occur. Shrimp in each treatment were not excluded from feces exposure from other shrimp.

Bacterial population change in response to OTC. Cluster analysis of these data indicated that as exposure time to OTC increased, there was an effect of time and OTC concentrations to reduce the growth of V. vulnificus, V parahaemolyticus, V cholerae, and Vibrio spp. when compared to the control group (data not shown), while a clear trend of increased growth over time was observed for V. alginolyticus. This suggests that V. alginolyticus is more resistant to OTC than were other Vibrio spp. (e.g., V. cholerae, V. parahaemolyticus, and V. vulnificus). The bacterial composition interactions and changes affected by OTC over time among bacterial groups in shrimp, water, and sediment have been well documented [35,36,38,39]. The findings indicate that the dose effect was greater than the temporal effect on the microbial community in the present study, due to the short period used for this experiment (96 h). A rapid decrease in species richness (bacterial composition over time) was observed at higher OTC concentrations.

Table Table 2.. Kruskal-Wallis pairwise comparisons among resistant bacterial species and oxytetracycline (OTC) treatmentsa
Bacterial species (% OTC-resistant isolates)1–16 mg/L1–32 mg/L16–32 mg/L
  1. a * = p values significant at the 0.05 level.

Aeromonas hydrophila (27.78)0.55160.38070.1367
Aeromonas spp. (8.33)10.0114*0.0114*
Brevundimonas vesicularis (8.33)10.0126*0.0126*
Chryseobacterium indologenes (5.56)10.0114*0.0114*
Unknown (2.78)10.0114*0.0114*
Vibrio alginolyticus (19.44)10.0139*0.0139*
V. cholerae (2.78)0.0114*0.0114*0.1292
V. parahaemolyticus (5.56)10.0114*0.0114*
V. vulnificus (8.33)10.0114*0.0114*
Vibrio spp. (11.11)0.20590.12920.0114*

A multiple comparison procedure using weighted resistance values among treatments was performed using the Kruskal-Wallis test. This allowed a determination of whether OTC had a significant biological impact in at least one of the treatments (Table 2). Individual Kruskal-Wallis tests were performed on 10 out of 15 (67%) different bacterial species (A. hydrophila, Aeromonas spp., B. vesicularis, C. indologenes, V. alginolyticus, V. cholerae, V. parahaemolyticus, V. vulnificus, Vibrio spp., and one unknown group), which demonstrated significant biological effects in terms of multiple bacteria response in at least one OTC exposure concentration. Multiple pairwise comparisons were also performed to establish differences among treatments. Significant differences were detected among groups of bacteria isolated in controls versus bacteria isolated from treatments (1, 16, and 32 mg/L OTC).

The majority of the OTC-resistant isolates came from grass shrimp exposed to 32 mg/L OTC (75%). Overall, A. hydrophila were the most resistant isolates (27.78%) to OTC. High frequencies of resistance for the Aeromonas group of bacteria caused by selective pressure applied through OTC treatment have been reported [40,41]. These authors attributed the resistance for this group to OTC regulating conjugative transfer of a specific plasmid present in the Aeromonas group bacteria, together with a pronounced promiscuity of the plasmid, which may result in a wide variety of plasmid-carrying bacteria acting as potential donors for further conjugation.

In the present study, V. alginolyticus (19.44%) and Vibrio spp. (11.11%) showed high levels of resistance at 32 mg/L OTC for shrimp homogenates isolates (Table 2). Alterations in Vibrionaceae family resistance associated with OTC exposure have been explained by a resistance determinant, Tet 34 determinant, which encodes a new Mg2+-dependent OTC mechanism [42]. Moreover, Vibrio strains could acquire resistance through bacterial conjugation and plasmid stability (depending on the plasmid size, 9 to 123 kb) and consequently increase their antibiotic resistance in the presence of one or more antibiotics [43]. The relevance of the present research was to identify and test bacteria isolated from a cornerstone biomonitoring species, P. pugio. The Vibrionaceae species (V. alginolyticus, V. cholerae, V. parahaemolyticus, V. vulnificus, and Vibrio spp.) accounted for 47.2% of the resistant shrimp isolates following exposure to OTC. These species are capable of causing public health problems such as cholera and human intestinal diseases through eating contaminated seafood or having an open wound that is exposed to seawater or drinking contaminated water. Their potential impact in the environment is obvious, considering the water-sediment-microorganism system as a transmission vehicle. Moreover, other OTC-resistant bacterial species like B. vesicularis and C. indologenes, which showed significant response to OTC (Table 2), have been reported primarily in soil and water associated to hospital-acquired virulent strains [44,45].

Table Table 3.. General antibiotic resistance measured in grass shrimp bacteria from oxytetracycline (OTC) treatment and control isolates
 Intestinal bacteria (n = 268)
Antibiotic tested (range of concentration, μml)No. strains% IsolatesPearson correlation coefficienta
  1. a – = zero value; * = significant difference from the controls at p < 0.05; + = positive linear association detected for OTC.

Amikacin (8-64)145.2−0.2349
Amoxicillin (4-32)3513.1−0.1961
Ampicillin (4-32)2810.4−0.2777
Apramycin (8-32)24591.40.0170
Azithromycin (2-8)114.1−0.1713
Cefoxitin (8-32)155.6−0.4442
Ceftriaxone (8-64)00
Cephalexin (16-128)11342.2−0.5392
Cephalothin (16-128)5620.9−0.2204
Chloramphenicol (8-32)00
Ciprofloxacin (1-4)00
Erythromycin (16-128)134.9−0.1719
Gentamicin (2-16)93.40.0541
Imipenem (2, 8-16)00
Meropenem (2, 8-16)00
Moxifloxacin (0.25-4)00
Nalidixic acid (4-32)00
Nitrofurantoin (16-128)124.50.0482
Ofloxacin (1-8)00
OTC (4-32)3613.41
Penicillin (16-128)4115.30.8476* (+)
Streptomycin (16-128)10639.6−0.4856
Sulfathiazole (250-500)4115.30.9526* (+)
Tetracycline (4-32)3312.30.9835* (+)
Trimethoprim (2-16)3111.60.9592* (+)
Trimethoprim and2910.80.9802* (+)
sulfamethoxazole (2/38-4/76)   

Antibiotic resistance analysis. The 268 isolates from grass shrimp digestive tract were tested against 26 other antibiotics, including OTC. No resistance to OTC was detected in the isolates from controls. Among the tested isolates from the OTC treatments, 15.4% were resistant to OTC and 84.6% were OTC sensitive. In general, bacteria showed a natural resistance for apramycin (MIC ≥ 32 μml), where levels of 91.4% were observed for the intestinal bacteria. Furthermore, isolates had a resistance value of 42.2% for cephalexin (MIC ≥ 32 μml). All isolates were sensitive to eight antibiotics: chloramphenicol, ceftriaxone, ciprofloxacin, imipenem, meropenem, moxifloxacin, nalidixic acid, and ofloxacin. Results of the ARA panels are shown in Table 3.

The Pearson correlation analysis of the relationship between OTC-resistant bacterial isolates and other antibiotics was performed. Statistical analysis of resistance indicated positive linear relationships (p < 0.05) between bacteria exposed to OTC treatments and enhanced antibiotic resistance for five other antibiotics, each with a different mechanism of action than OTC: penicillin G (cell wall synthesis inhibitor), sulfathiazole, trimethoprim, trimethoprim and sulfamethoxazole (interference of DNA synthesis), and tetracycline (protein synthesis inhibitor). Isolates resistant to OTC were two times the MIC value for the same isolates exposed to tetracycline. For instance, 33 intestinal bacterial isolates out of 36 (92%) were resistant to OTC and tetracycline (data not shown). This close relationship showed in the ARA between OTC and tetracycline indicates a strong association in terms of resistance for the tetracycline group, as well as their mechanism of action. Similar findings have been reported from a previous study carried out by DePaola et al. [46]. This ARA pattern observed among OTC and five other related and unrelated antibiotics may be caused by a variety of mechanisms, such as the AcrAB efflux pump, or acquisition of multiple transposons and plasmids [40,47,48]. These findings suggest the need for mixture studies of antibiotics to further evaluate multiple resistance induction through genes that might encode the production of nonselective proteins. This provides clear evidence that the observed multiple resistance for the rest of the antibiotics was preexisting and was acquired by bacteria prior to exposure.

Figure Fig. 4..

Preliminary environmental risk assessment for oxytetracycline (OTC) based upon results obtained in the present study, other published studies [5,7,13,14,17–19,34,49,50,52,53], and combined studies (present study vs. other studies) on OTC effects.

Determination of OTC

Analytical method conditions. The calibration curve was linear over the range of 0.1 to 1,000 μml OTC. The equation from the curve for grass shrimp homogenates was y = 10,001,600 + 1,327,350 x, where y was the area found through the liquid chromatography-mass spectrometry analysis and x was the OTC concentration for each sample. Calibration curves had a correlation coefficient of 0.9993. A methanol blank, which provides the background of the system, was applied as a correction factor (333,510) in the equation. The lower limit of detection and quantification was 1 μg.

Bioaccumulation of OTC. Detectable OTC concentrations were generally not found in grass shrimp. The only detectable peak where OTC could be quantified was at 96 h in the 32 mg/L OTC exposure for shrimp homogenates. Tissue levels of 12.5 μg OTC were measured, which resulted in an estimated bioconcentration factor of 0.39 for OTC. This indicates that the bioaccumulation potential for OTC is minimal. Similar results have been reported for crustaceans and marine sediments [4,49,50].

The analytical method used in the present study was not able to detect concentration lower than 1 μg. No previous references or studies were published on OTC quantification in grass shrimp. The quantification method in the present research had less sensitivity than other studies, which have determined more sensitive quantification limits in water and shrimp matrixes; these values range from 0.0014 to 0.25 parts per million OTC [4,49–51].

The peaks that were detected could not be quantified due to interference among peaks within the chromatograph. Even though OTC uptake was not consistently quantifiable, what little bioconcentration that did occur had an obvious qualifiable effect on the bacterial population within the grass shrimp digestive microflora. Results of the present study indicate that OTC concentrations in the P. pugio habitat required to cause a decrease in the grass shrimp population do not pose significant concern in the majority of the cases and would occur under extremely special conditions based upon current biomonitoring results. For instance, a recent study in China has reported concentrations of OTC up to 712 μL and 262 mg/kg OTC in surface water and sediments, respectively [51], indicating that antibiotic resistance of bacteria could be induced and maintained due to levels of OTC present in the environment.


The use of grass shrimp digestive tract microbial flora as a biomonitoring tool demonstrated that antibiotics like OTC do affect the digestive tract microflora and may be useful to assess the likely impact and risks of pharmaceuticals in the environment. Given the trophic importance of the bacterial microflora in crustaceans, this has the potential to affect feeding or assimilation efficiencies and may potentially affect growth and reproduction. The levels of significance of predicted impacts in current risk assessment models for aquatic animals are based on population mortality rather than the effects of pharmaceuticals in terms of antibiotic resistance or alteration in digestive tract microbial flora.

The present research demonstrated the existence of sublethal responses in digestive tract bacterial flora (e.g., Vibrio and Aeromonas spp.) that are mainly recognized as causative associated agents of human diseases. The present study indicated that pharmaceuticals in the environment may have potential effects to increase microbial resistance for these public health pathogens to both humans and aquatic organisms.

A preliminary environmental risk assessment using results obtained in the present research and other related studies indicates that acute toxicity effects, bioaccumulation in grass shrimp, alterations, and antibiotic resistance in the grass shrimp microbial community would only be expected under special circumstances. These concentrations have been measured in downstream of effluents from major pharmaceutical manufacturers and sediments associated with aquaculture and confined animal feeding operations but not at levels measured in sewage treatment plant effluents and surface waters of the United States in general (Fig. 4). The present study suggests that OTC exposure from untreated aquaculture effluents, confined animal feeding operations, and discharges from pharmaceutical manufacturers may pose greater risks than the risks associated with OTC exposure from municipal wastewater treatment plant effluents. This preliminary risk assessment requires additional field monitoring and assessment to confirm the risk estimates suggested here.

Future studies should also consider using lower OTC subchronic exposures or different P. pugio life stages, as well as other crustacean and vertebrate species. More research is needed to fully measure the environmental impacts on nontarget organisms. If the final fate of pharmaceuticals is indeed the sediments, then the use of benthic species like clams, amphipods, isopods, polychaetes, and copepods will be required to fully assess the most vulnerable organisms in aquatic environments, particularly where aquaculture facilities are located.


The authors would like to acknowledge B. Thompson, L. Webster, J. Hoguet, J. Venturella, K. Chung, H. Harper, A. Chandler, J. Jutzi, and B. West (NOAA). We also thank G.T. Chandler, J. Ferry, and R. Vesselinov (University of South Carolina), and P. Pennington, P. Moeller, and K. Huncik (NOAA). Academic support and scholarship for M. Uyaguari was granted by the Fulbright Commission. The National Ocean Service (NOS) does not approve, recommend, or endorse any proprietary product or material mentioned in this publication. No reference shall be made to NOS, or to this publication furnished by NOS, in any advertising or sales promotion that would indicate or imply that NOS approves, recommends, or endorses any proprietary product or proprietary material mentioned herein or that has as its purpose any intent to cause directly or indirectly the advertised product to be used or purchased because of NOS publication.