Introduction

Different groups of planktonic organisms, such as phytoplankton and zooplankton, have been cultivated in semi-intensive and intensive systems for ecological and biological studies as live food in aquaculture, as well as for toxicological tests (Gilles et al. 2013; Edwards et al. 2016; Lopes et al. 2018). Unicellular marine algae are commonly grown as food for several commercially valuable marine organisms (e.g., fishes, crustaceans, mollusks) (Riquelme and Avendaño-Herrera 2003; Gilles et al. 2013). Copepods and barnacle larvae are used as live food in aquaculture for fishes (Støttrup 2000; López et al. 2010). Microalgae, copepods, and barnacles are also increasingly used as bioindicator species in ecotoxicology (Raisuddin et al. 2007; Buttino et al. 2012; Cid et al. 2012; Piazza et al. 2012; Lopes et al. 2018) and as models to predict the impact of chemicals on communities. These organisms are also cultivated to investigate species and population characteristics (growth, development, interactions, production, settlement, and life cycles) (Thiyagarajan et al. 2003a, b; Lee et al. 2013; Freckelton et al. 2017).

The presence of small amounts of organic matter in cultures can lead to a 10- to 1000-fold increase in bacteria in less than 24 h (D’Agostino 1975). Such rapid microbial growth is often accompanied by marked decreases in pH, dissolved oxygen, and concentrations of dissolved nutrients, as well as the release of potentially toxic bacterial metabolites (Fitt et al. 1992; Carvalho and Fernandes 2010). Bacteria associated with diatoms are usually very tenaciously attached, and some may penetrate the gelatinous sheaths and harm phytoplankton (Subhash et al. 2004; Hamdan and Jonas 2007). Bacteria have the potential to negatively affect the growth and reproduction of crustaceans (Wang 2011; Edlund et al. 2012) and are associated with failures in the development of zooplankton cultures (Agostini et al. 2016).

Pathogenic microbes (i.e., virus, bacteria, fungi) generally invade cultures by three principal routes: culture medium, broodstock, and food. Numerous studies have investigated chemical (antibiotics, hypochlorite, iodine, hormones, phenol and detergents) and physical (membrane filtration, autoclave, ultraviolet light, and sonication) treatments aiming to inhibit these microorganisms without negatively affecting the cultivation of eukaryotic organisms (i.e., algae, mollusks, crustaceans) (Spencer 1952; McCracken 1989; Skjermo and Vadstein 1999; Hamdan and Jonas 2007; Liu et al. 2012; Scholz 2014; Agostini et al. 2016). Most studies used different chemicals (mainly the antibiotics penicillin, streptomycin and neomycin) to inhibit bacterial activity without preliminary tests to ensure that the chemical would not affect experimental organisms (Spencer 1952; Droop 1967; McCracken 1989; Skjermo and Vadstein 1999). Most of these studies also used one antibiotic alone and assumed inhibition of the target microorganisms without previously testing the efficiency of these chemicals.

The importance of antibiotics in controlling bacterial growth in marine cultures has been recognized since the 1940s, when they were used in protozoan cultures (Ithaca and Mahmoud 1944). The advantage of antimicrobials over all other methods is their efficiency in microbial inhibition combined with the availability of several different types of antimicrobials. It remains uncertain, however, whether the experimental animals (non-target organisms) would be affected by exposure to these antibiotics. Antibiotics combinations have broad-spectrum bacterial inhibition, but some authors have emphasized that the niche provided by these prokaryotes may be occupied by filamentous fungi and/or yeasts. Application of antifungals would thus inhibit these fungi (Provasoli et al. 1951; Tang et al. 2006; Agostini 2014; Agostini et al. 2016).

The application of antibiotics and antifungals (i.e., antimicrobials) in planktonic cultures can prevent disease and competition (Agostini et al. 2016). This allows ecological studies of the role of bacteria and fungi in aquatic communities (Trottet et al. 2011) and biological (genetic, taxonomy, development, and growth) studies without bacterial or fungal interference (Howes et al. 2014). The goal of this study was thus to evaluate the effect of a combination of penicillin, streptomycin and neomycin with an antifungal agent to select the most effective concentration for preventing bacterial and fungal growth without negatively affecting phytoplankton or zooplankton organisms. The novelty of the present study is the evaluation of the effect of antimicrobial (prokaryote and eukaryote inhibitors) combinations on target and experimental organisms.

Materials and methods

This study consisted of experiments to evaluate the effects of different antibiotics (penicillin, streptomycin and neomycin) at different concentrations (Agostini et al. 2016; DeLorenzo et al. 2001) administered in combination with nystatin (Finley 2012) (Table 1) to plankton cultures. The experimental organisms evaluated were Acartia tonsa Dana 1849 (Copepoda: Calanoida), Amphibalanus improvisus (Darwin 1854) (Crustacea: Cirripedia), and Conticribra weissflogii (Grunow) Stachura-Suchoples & Williams (Ochrophyta: Bacillariophyceae). These organisms represent zooplankton (holoplankton and meroplankton) and phytoplankton, respectively.

Table 1 Summary of the treatments tested on experimental (copepod, barnacle, microalgae) organisms
Full size table

The animals were obtained from zooplankton samples collected by means of horizontal tows performed at the surf zone of Cassino Beach, Rio Grande, RS, Brazil. All samples were collected with a conventional 200-µm plankton net with a 30-cm mouth diameter. The microalgae C. weissflogii were obtained from cultures kept at the Marine Phytoplankton and Microorganisms Laboratory of the Federal University of Rio Grande in F/2 medium (Guillard 1975) at salinity 30, temperature 25 ± 1 °C and a 12hL:12hD (Light:Dark) photoperiod under 70 mol photons s−1 m−2 artificial illumination inside BOD incubators (Marconi 403).

Copepods and barnacles were identified in the laboratory (Sabatini 1990; Jones and Crisp 1954, respectively). Healthy specimens of A. tonsa and A. improvisus were manually picked using Pasteur pipettes under a stereoscopic microscope (Olympus SZ40). All specimens were classified according to the development stage (i.e., copepodites VI males and females of A. tonsa and nauplius VI of A. improvisus to obtain cyprids of the same age after metamorphosis). Before the experiments were performed, the organisms were submitted to a 30-h acclimation under similar field conditions (e.g., salinity 30, temperature 25 °C and a 14hL:10hD photoperiod) to remove any organism that had been injured or debilitated by stress inherent in collection and sorting. Adults of Acartia tonsa and nauplius VI of A. improvisus were acclimated separately.

After acclimation, eight replicates containing two copepods, one male and one female, were placed in 50-mL experimental units (EUs) (density of 1 org 25 mL−1). Each EU was fitted with a 140-µm mesh to separate the copepod eggs and fecal pellets produced (Runge and Roff 2000). Copepods were exposed to different antimicrobial treatments (Table 1) following different application routes: antimicrobial applications only in the copepod culture medium (M), only in the culture medium of their food (F), and in the culture media of both the copepods and their food (M/F). The experiment was static; in this way, the treatments were applied once per culture (time 0). Recently molted cyprid larvae were deposited in each EU individually (1 org 20 mL−1) under the same acclimation conditions, with six replicates per treatment for barnacles. Cyprids were then exposed to different antimicrobial treatments, as indicated in Table 1, and the experiment was static.

The copepod and barnacle cultures were observed every 24 h for 96 h. Eggs and fecal pellets produced by copepods were counted only at the end of the experiment (96 h) to evaluate the effects of the treatments on reproduction and feeding, respectively. Copepods were fed daily with the diatom C. weissflogii at a concentration of 20,000 cells mL−1 (Teixeira et al. 1896). Because cyprids do not feed at this stage, no food was added to their EU.

In the experiments with the microalgae C. weissflogii, the initial concentration was 31 cells mL−1 in 200 mL of medium, with six replicates per treatment. The same growing conditions were maintained, with the exception of the photoperiod, which was changed to 14hL:10hD to follow the same conditions of the experiments with the zooplankton. Due to the absence of aerators, the EUs were shaken every 6 h to prevent sedimentation of the cells. The experiment was static. To follow the growth of C. weissflogii, 2 mL of each culture was sampled at 24, 96, and 168 h at the same time and after homogenization of the EU. This subsample was deposited in Eppendorf tubes and preserved with lugol (1%). Cell counts were performed in Neubauer chambers under a microscope (40 × – Olympus BH-2) to quantify growth (density per exposure time) and yield (final cell density − initial cell density).

After the first set of experiments, the antifungal nystatin was added at different concentrations (based on Lopes 2014) to the antibiotic combinations that produced the highest survival and production rates of A. tonsa compared to the control (Table 1). This new set of treatments (Table 2) was performed in two steps. In the first step, the survival of the calanoid copepod A. tonsa under the treatments was assessed (Table 2). In the second step, the efficacy on the inhibition of biofilm bacteria and adhered fungi in marine cultures and the half-life of the best antimicrobial combinations obtained in the first step were investigated.

Table 2 Summary of the treatments tested on target (bacteria and fungi) and experimental (copepod, barnacle, microalgae) organisms
Full size table

Individuals of A. tonsa were collected from zooplankton tows and acclimated under the same conditions as described above. After acclimation, adult organisms were separated without sex distinction (ISO 14669 1999) and placed in pairs in each EU with 40 mL of culture medium (1 org 20 mL−1) (Lopes et al. 2018), representing seven treatments with ten replicates each (Table 2). The cultures were kept under the same condition as in the previous experiments, with a 14hL:10hD photoperiod. C. weissflogii at 20,000 cells mL−1 (Teixeira et al. 1896) was added daily to each EU as food. The experiment was semi-static, with the culture medium renewed daily. To evaluate the impact of treatments on copepod survival, each EU was observed every 24 h for 96 h.

To evaluate the antimicrobial efficacy and half-life, eight substrates of wood (20 × 10 mm) that had been previously cleaned with 70% ethyl alcohol (Caixeta et al. 2012; Agostini et al. 2017) were placed in each EU (120 mL), with three replicates per treatment. To prevent biofilm colonization on only one side of the wooden substrates, the EUs were shaken five times a day. All EUs were kept under the same conditions as in the previous experiments inside a BOD incubator (Marconi 403). To assess the initial effects of the antibiotics as well as the half-life in artificial seawater, the experiment was static, and antimicrobials were applied only once (time 0).

To estimate the biofilm bacterial density, three aliquots from three substrates were taken after 3, 6, 9, 12, 15, 18, 21, 24 and 168 h of exposure (based on Lopes 2014) and placed in sterile saline solution (20 mL). The microorganisms had to be detached from the wooden substrates using three pulses of 20 kHz for 15 s on each side of the substrate (Oliveira et al. 2006) using a Cole-Parmer® ultrasound (series 4710). The detached biofilm solution was fixed with 4% sterile formaldehyde, placed inside Eppendorf tubes, and stored in the dark at 8 °C until analysis. Biofilm bacterial density (org cm−2) was estimated in a flow cytometer (BD FACSVerse™) at the Faculty of Pharmacy of the Federal University of Rio Grande do Sul (UFRGS) (Agostini et al. 2016, 2017).

To evaluate the microbial community, biological material (of water and substrate) obtained from the EUs was filtered on polycarbonate filters (darkened with irgalan black), stained with acridine orange (1%) and viewed under an epifluorescence microscope (Zeiss Axioplan) at 1000× magnification. The presence or absence of fungi was also recorded, and bacterial morphotypes were classified (Zaritski 1975). The size (µm) and cell complexity were evaluated using the mode Forward Light Scatter (FSC-A) and Light Side Scatter (SSC-A) parameters using 6-µm latex beads (Molecular Probes®) as the standard (Herzenberg et al. 2006; Bouvier et al. 2011; Picot et al. 2012). The cell size (µm) estimated by flow cytometry was compared with measurements performed over epifluorescence photos (100 bacteria manually measured). The bacterial cell biovolume (μm3) estimation (Sun and Liu 2003) and conversion to bacterial cell biomass (pg C cell−1) were made using the allometric conversion factor (0.09*biovolume0.09) (Norland 1993).

General linear model (GLM) analysis was performed for A. tonsa survival and production, A. improvisus survival and settlement, and C. weissflogii and bacterial growth at each exposure time using a statistical computing package (R Development Core Team 2016). The model used was adapted to the data with a binomial distribution (survival and settlement) with a “logit” link function and a Poisson distribution (growth and production) with a “log” link function. Post hoc Tukey’s test was performed after the analyses. After analysis, antimicrobials were stored in sealed containers and collected by the Engineering and Environmental Services, SANIPLAN©.

Results

Significant survival rates of Acartia tonsa among treatments were observed. Significant differences occurred among the antimicrobial treatments applied in M after 48 h of exposure. At 48 h, 100, 100, 38, 88 and 25% survival rates were observed in the control, TA, TA+nystatin, TD and TD+nystatin, respectively. TD+nystatin was significantly different from the control (p < 0.028) and TA (p < 0.028). At 72 h of exposure, TA+nystatin and TD+nystatin showed lower survival rates (13 and 0%, respectively), being different from all treatments (p < 0.029). At 96 h of exposure, no significant difference was observed between the control and the TA and TD treatments (p > 0.985); copepod survival rates were 88, 88 and 75%, respectively (Fig. 1a). When antimicrobials were applied only in F, no significant differences were observed (p > 0.441) (Fig. 1b). At 96 h of exposure, the average survival was 88, 75, 50, 63 and 63% for the control, TA, TA+nystatin, TD and TD+nystatin, respectively. When antimicrobials were applied in M/F, differences were observed from 48 h of exposure, where 100, 88, 0, 75 and 38% survival were recorded in the control, TA, TA+nystatin, TD and TD+nystatin, respectively. TA+nystatin (p < 0.002) was significantly different from the control, as well as from TA (p < 0.007) and TD (p = 0.027). At 72 h, TA+nystatin and TD+nystatin showed no survivors, differing from all treatments (p < 0.028). At the end of the experiment (96 h), no significant difference was observed between the control (88%) and the TA (88% survival) and TD (75% survival) treatments (Fig. 1c).

Fig. 1
figure 1

Acartia tonsa copepod survival and egg and fecal pellet production under different treatments and application routes. a Survival under antimicrobial exposure in the culture medium; (M) b survival under antimicrobial exposure through food (F); c survival under antimicrobial exposure in the culture medium and through food (M/F); d egg production (female−1 day−1 at 96 h); e fecal pellet production (copepod−1 day−1 at 96 h). The vertical lines denote 95% confidence intervals (standard error*1.96), and lowercase letters indicate statistical similarities or differences between treatments for each time or situation evaluated

Full size image

Significant differences in the egg production of A. tonsa (female−1 day−1 at 96 h) between treatments (p < 0.001) were observed (Fig. 1d). TA showed higher average egg production than the control (20 ± 3), with significant differences when antimicrobials were applied in M (p < 0.001), F (p < 0.048), and M/F (p < 0.001) at 49 ± 6, 27 ± 7, and 63 ± 24, eggs female−1 day−1, respectively. In M/F and M, TA+nystatin and TD+nystatin also presented differences from the control (p < 0.001), as the copepods under those antimicrobial treatments did not produce any eggs. TA+nystatin showed egg production only in F (16 ± 8), which was statistically equivalent to the control (p = 0.088). No differences in TD between M (13 ± 8) and F (8 ± 5) and the control (p = 0.088) were observed. In M/F, 36 ± 21 eggs were produced, which was significantly higher than the control (p < 0.001). TD+nystatin only presented egg production in F (7 ± 1), with significant differences from the control (p < 0.044). TA showed higher average egg production (46 ± 11 eggs female−1 day−1). Among the different antimicrobial application routes, M/F presented the best average results (25 ± 11 eggs female−1 day−1).

Fecal pellet production by A. tonsa (copepod−1 day−1 at 96 h) in those routes presented significant differences between treatments (p < 0.001) (Fig. 1e). TA in M/F showed a higher average pellet count (60 ± 16), followed by M (60 ± 26) and F (22 ± 17); but these values did not statistically differ from the control (49 ± 18). TA+nystatin showed fecal pellet production only in F (28 ± 10), equivalent to the control. No differences in TD between the different application methods and the control were observed. When antimicrobials were administered in M, 38 ± 17 pellets were produced, followed by F (37 ± 12) and M/F (35 ± 10). TD+nystatin only presented fecal pellet production in F (16 ± 8), with significant differences from the control (p < 0.001). Among the antimicrobial application methods, TA applied in M/F and M showed the best average results.

No differences were observed in A. improvisus survival (p > 0.991) (Fig. 2a) or settlement (p > 0.692) (Fig. 2b) between treatments. TA+nystatin and TD showed a lower survival average (83%). TA+nystatin also presented a lower settlement percentage at 96 h of exposure (50%). A higher settlement percentage was observed in the control (83%), but TA+nystatin only settled at 48 h. Significant differences in C. weissflogii growth rates between treatments were also found at 96 h (p < 0.001) (Fig. 2c). At 24 h, no significant differences were observed between all treatments with antimicrobials and the control (109 ± 76 cells mL−1), although TA presented higher growth rates (198 ± 61 cells mL−1). At 96 h of exposure, TA+nystatin (109 ± 21 cells mL−1) and TD+nystatin (78 ± 42 cells mL−1) had a lower average growth rate than the control (266 ± 100 cells mL−1) (p < 0.001). At approximately 168 h of exposure, only TA (536 ± 134 cells mL−1) and TD (365 ± 83 cells mL−1) were different from the control (270 ± 52 cells mL−1) (p < 0.001), with a higher cell concentration. TA showed the highest yield (505 cells mL−1), followed by TD (333 cells mL−1), TA+nystatin (193 cells mL−1), the control (177 cells mL−1), and TD+nystatin (162 cells mL−1).

Fig. 2
figure 2

Amphibalanus improvisus cyprid a survival and b settlement at 24, 48, 72 and 96 h; c Conticribra weissflogii diatom growth (cells mL−1) at 24, 96, and 168 h of exposure. The vertical lines denote 95% confidence intervals (standard error*1.96), and lowercase letters indicate statistical similarities or differences between treatments for each time evaluated

Full size image

In the resistance tests with A. tonsa, A. improvisus and C. weissflogii, the treatments with the antifungal nystatin (0.05 g L−1) were the most harmful to non-target organisms, as was the antibiotic concentration TD. Subsequent tests (Table 2) thus applied antimicrobials only in the culture medium, the condition in which the copepod A. tonsa was the most sensitive in the current study. The results presented significant survival rates of A. tonsa between treatments after 48 h of exposure. TA + n4 and TA + n5 were lethal to copepods, different from the control (90% survival) and the other antimicrobial treatments (p < 0.003). The same situation occurred at 72 h of exposure; however, the survivorship in the control was 80% (p < 0.009). At 96 h, no significant differences (p = 0.960) between the control (80% survival) and TA (100% survival), TA + n1 (70% survival), and TA + n2 (65% survival) were observed. TA + n3 (20% survival) (p < 0.009), TA + n4 (0% survival) (p < 0.010) and TA + n5 (0% survival) (p < 0.034), in contrast, showed the lowest survival rates when compared to the control.

The bacterial density (org cm−2) decreased in treatments with antimicrobials (TA, TA + n1 and TA + n2), as expected, compared to the control at 3 (p < 0.007), 6 (p < 0.001), 9 (p < 0.003), 12 (p < 0.001), 15 (p < 0.001), 18 (p < 0.001), and 21 (p < 0.001) h of exposure (Fig. 3a). In the first 3 h of the experiment, a reduction in the bacterial density (> 26%) was observed compared with the control. Between 9 and 15 h, an inhibition of up to 94% of the biofilm bacterial density was recorded compared to the control. After 15 h, bacterial densities increased in all antimicrobial treatments. The bacteria found in the control and antimicrobial treatments were in the coccus form with similar sizes.

Fig. 3
figure 3

a Biofilm bacteria density from 3 to 168 h of exposure. The vertical lines denote 95% confidence intervals (standard error*1.96). The lowercase letters indicate statistical similarities or differences between treatments at each time evaluated. The numbers above the columns denote the bacteria inhibition percentage (%) compared to the control. b biofilm bacteria populations by flow cytometer (BD FACSVerse™) after 10,000 acquired events at 12 h. Lighter colors are related to higher density cells; c microbial community under epifluorescence microscope (×1000) stained with acridine orange with emphasis on fungi presence/absence at 168 h

Full size image

All treatments with antimicrobials had the same cell size (FSC-A), biovolume and biomass values as the control at 12 h of exposure: 1.11 µm, 0.72 µm3 and 0.09 pg C cell−1, respectively. All treatments showed just one population and the same cell complexity (SSC-A) (Fig. 3b). The following size, biovolume and biomass were verified independent of treatment using an epifluorescence microscope at 12 h: 0.61 µm, 0.12 µm3, and 0.07 pg C cell−1, respectively. Overall, all antimicrobial treatments were effective in reducing bacterial density, but only TA + n2 prevented fungal colonization. No fungi were observed in the control. The antimicrobial treatment without nystatin (TA) showed filamentous fungi and Fusarium sp., while TA + n1 showed only a filamentous form (Fig. 3c).

Discussion

Our findings indicated that the application of TD (0.025 g L−1 penicillin G potassium + 0.04 g L−1 streptomycin sulfate + 0.08 g L−1 neomycin sulfate) proposed by DeLorenzo et al. (2001) resulted in a lower survival rate, with lower egg and fecal pellet production for the copepod Acartia tonsa when compared with TA (0.025 g L−1 penicillin G potassium + 0.08 g L−1 streptomycin sulfate + 0.04 g L−1 neomycin sulfate) proposed by Agostini et al. (2016). The differences between these two treatments are the streptomycin and neomycin concentrations. The higher concentration of neomycin sulfate (0.08 g L−1) proposed by DeLorenzo et al. (2001) caused a greater mortality of A. tonsa than did the same concentration of streptomycin proposed by Agostini et al. (2016).

Antibiotics such as penicillin, streptomycin and neomycin have been applied in cultures to inhibit bacterial contamination without negative effects in microalgae (Molina-Cárdenas et al. 2016), protozoans (Divan and Schnoes 1982), fish (Forberg et al. 2011), mollusks (Howes et al. 2014), crustaceans (Agostini et al. 2016) and ecological assays with the aquatic community (Middelburg and Nieuwenhuize 2000a, b; DeLorenzo et al. 2001; Veuger et al. 2004; Cozzi and Cantoni 2006; Trottet et al. 2011). The concentration of these antibiotics, however, needs to be evaluated before application in cultures. In this study, TA and TD, composed only of antibiotics, showed different responses in non-target organisms due to differences in the concentrations of some compounds.

Acartia tonsa, A. improvisus and C. weissflogii are possibly more sensitive to high concentrations of neomycin than to streptomycin. Droop (1967), for example, used 0.04 g L−1 neomycin in combination with other antibiotics in diatom cultures. Jones et al. (1973) applied only 0.0000001 to 0.000016 g mL−1 mixed with other prokaryotic inhibitors in cyanobacteria cultures, while Green et al. (1967) purified seaweed cultures with 0.0002 g L−1 neomycin mixed with other antibiotics. In addition to the lower resistance to TD compared to TA, TA+N and TD+N had lower survival of non-target organisms, probably due to the nystatin concentration used (0.05 g L−1).

Kaminski and Montú (2005) reported A. tonsa egg production of 14–34 eggs female−1 day−1 when fed with Nannochloropsis oculata (Droop) Hibberd and Chaetoceros calcitrans (Paulsen) Takano in excess. Teixeira et al. (1896) observed an average of 28 eggs female−1 day−1 when fed with C. weissflogii at the same concentrations as used this study. The data presented here indicate a beneficial effect of those antibiotics in copepod cultures; our results recorded an average of 63 eggs female−1 day−1 in TA in M/F and 49 eggs female−1 day−1 in TA in M. In the control, we found 20 eggs female−1 day−1. The number of fecal pellets produced by copepods can be used as an additional tool to evaluate the nutritional quality of the food (Ianora et al. 1995), although food quality, concentration and size must be considered. Shaw et al. (1994) observed that the culture medium of the copepod Tigriopus californicus (Baker, 1912) without antibiotics had a production rate of 1.6 pellets h−1, while cultures with antibiotics had a production rate of 1.9 pellets h−1 with the same amount of food. The current study confirms the results of Shaw et al. (1994) because TA presented higher production of pellets in M/F and M (2.5 pellets h−1) than the control (2.0 pellets h−1). The effect of the antimicrobials on fecal pellets and egg production should be analyzed carefully because production was observed during short periods and the antimicrobials were not replaced, decreasing their effect over the duration of the experiment. The benefit of using antibiotics in cultures was also reported by Tighe-Ford et al. (1970), who used 100 µL L−1 Crystamycin® (0.3 g of penicillin G sodium and 0.5 g of streptomycin sulfate in 2 mL of distilled water) in barnacle cultures of Austrominius modestus (Darwin 1854) to virtually eliminate bacterial growth and increase the rate of survival of larvae (Tighe-Ford et al. 1970; Harms 1987). The survival of Amphibalanus improvisus in all treatments indicates that these antimicrobials can be applied in cultures of this species without causing lethal damage. Thus, we assume that barnacles are less susceptible than copepods and microalgae. Laboratory cultures allow the study of the growth, development, metamorphosis, fertility and nutritional needs of species as a means of research into the role of organisms in nature (D’Agostino 1975). Thus, the application of TA + n2 in marine plankton cultures can have different objectives.

We expected that the antimicrobial treatments would lead to lower A. improvisus settlement rates, because the presence of bacterial biofilms has been reported to condition larval settlement of a number of fouling benthic species (Unabia and Hadfield 1999; Lau et al. 2002; Hung et al. 2005; Bao et al. 2010; Agostini et al. 2017), such as barnacles. This relationship was only observed from 72 h of exposure, when the control reached the highest settlement rate. Because the antibiotics were applied only at the beginning of the cultures, we believe that at 24 h, the effect of the treatments had decreased, resulting again in an increase in the bacterial load.

The use of TA+nystatin, TD and TD+nystatin in cultures of the diatom C. weissflogii did not produce differences when compared to the control. This indicates the possibility of applying these treatments for purification of microalgae cultures. TA presented better results than the control, being the most recommended treatment for intensification of microalgal growth. The use of antibiotics in microalgae culture has been reported to be beneficial because bacteria actively compete for the same resources necessary for the survival of phytoplankton (Spencer 1952; Subhash et al. 2004; Hamdan and Jonas 2007). It should be considered, however, that this study was performed only with C. weissflogii and that the half-life of antimicrobials reduces the effect over time. In addition, bacteria can mediate a variety of harmful or beneficial interactions with eukaryotic organisms. Beneficial interactions would be the acquisition of vitamin B12 through symbiotic relationships with bacteria (Croft et al. 2005). In contrast, the disruption of the microflora would likely delay the development of some juvenile crustaceans (Edlund et al. 2012). Antimicrobial substances can thus affect symbiont-mediated interactions in hosts, negatively so if this inhibition continues over time. Lethal effects on microalgae are the lowest when using penicillin compared to other antibiotics (Youn and Hur 2007), although the efficiency of the antibiotics and their concentrations for axenic cultures vary with microalgal species (Lai et al. 2009). The establishment of Bacillariophyceae and Dinophyceae axenic cultures is more difficult than that of Chlorophyceae and Haptophyceae because of their complicated external morphology (Youn and Hur 2007).

Our results indicate that the copepod A. tonsa is more sensitive than A. improvisus and C. weissflogii, especially when treatments are applied only in the culture medium. This result was somewhat expected because this copepod species is very sensitive and is indicated for acute ecotoxicological tests by the International Standardization Organization (ISO 14669 1999). A. tonsa was therefore selected to test the other treatments with different concentrations of nystatin. The results obtained in the settlement of A. improvisus prompted us to re-apply the antimicrobial every 24 h to maintain the culture medium throughout the experiment. Testing different concentrations of nystatin in combination with TA revealed that TA, TA + n1 and TA + n2 could be applied to A. tonsa cultures. According to Lopes (2014), it is possible to apply up to 0.02 g L−1 nystatin in combination with antibiotics in cultures without lethal effects on A. tonsa for up to 48 h. There was a higher mortality of copepods in the treatments that showed the antifungal nystatin, probably because nystatin is a eukaryotic inhibitory substance (Groll et al. 1999). The use of an antifungal, however, is essential for avoiding fungal contamination due to the ecological niche provided by bacteria (Agostini et al. 2016). Diseases in cultures of invertebrates are caused by fungi of the genera Aspergillus, Penicillium, and Fusarium (Santos et al. 2016).

TA, TA + n1 and TA + n2 showed inhibition in biofilm bacterial density of up to 94% compared to the control at 9–15 h of exposure. In this time interval, the microbial community started to restore, and a new dosage of antibiotics was necessary (Agostini et al. 2016). Lopes (2014) obtained adherent bacterial reduction of up to 95% using the same combination of antimicrobials. Trottet et al. (2011) tested six different antibiotics under laboratory conditions. These authors found that penicillin and streptomycin showed the greatest effectiveness in inhibiting bacterial density. Penicillin targets Gram-positive bacteria; streptomycin targets Gram-negative and Gram-positive bacteria, and neomycin targets Gram-negative bacteria and some Gram-positive bacteria. This combination thus inhibits various bacterial species (i.e., Pseudomonas, Vibrio, Agrobacterium, Xanthomonas, Achromobacter, Flavobacterium, Micrococcus, and Staphylococcus) (Berland and Maestrini 1969).

Fungal proliferation could only be avoided by using TA + n2. To eliminate bacterial and fungal contamination in marine cultures without causing lethal damage to non-target organisms, we suggest the application of TA + n2. To maintain the culture medium with a minimum bacterial load for more time, antibiotics should be replaced every 12 h. Further studies are therefore necessary to evaluate the effects of these antimicrobials on non-target organisms and to examine alternatives to these substances in culture medium. The antifungal nystatin may be reapplied over a longer time interval, considering that in 7 days (168 h) of observation, the presence of adherent fungi was not recorded in TA + n2.

Conclusion

We found an absence of negative effects on experimental organisms and efficiency in the inhibition of biofilm bacteria using a combination of 0.025 g L−1 penicillin G potassium + 0.08 g L−1 streptomycin sulfate + 0.04 g L−1 neomycin sulfate diluted in seawater. Nevertheless, the absence of adherent fungi was only observed using 0.005 g L−1 nystatin antifungal together with the antibiotics. This concentration of nystatin did not show significant differences from the control regarding A. tonsa survival, but this substance should be avoided if fungal contamination is not a concern because this substance is a eukaryote inhibitor. Species of other planktonic groups should also be tested at this concentration to improve our understanding of the effectiveness of the treatments proposed.