**2. Materials and methods**

### **2.1. Experiments with pure bacterial cultures**

Bacterial cultures used in experiments were always freshly prepared by incubation of pure culture on mineral‐modified glutamate broth (MMGB). Bacterial suspensions were taken in the exponential phase of growth and serially diluted in phosphate buffer solution. After determination of the concentration of *E. coli* cells by epifluorescence microscopy method, appropriate dilution, cca 103  CFU/mL, was kept at 4°C until the beginning of the experiments (cca 20 h).

Experiments in the dark were performed in two sets. In the first set, experiments were carried out at six different experimental conditions corresponding to Adriatic sea natural range of temperatures (12°C—mean winter temperature, 18°C—mean spring and autumn temperature and 24°C—mean summer temperature) and salinities (30.0 psu—lower salinity corresponding to areas near the mouth of rivers or sewage outfalls and 36.5 psu—typical salinity in coastal seawater). Two different *E. coli* strains were used: *E. coli* ATCC 35218 originating from canine feces and *E. coli* ATCC 8739, originating from human feces. Bacterial suspension was added to 500 mL borosilicate glass bottles containing autoclaved seawater of appropriate salinity (30.0 and 36.5 psu). Bacterial concentration in bottles was targeted to 103 CFU *E. coli*/100 mL. The bottles were then stirred and placed in a temperature controlled air incubators at three different temperatures (12, 18 and 24°C).

In the second set, additional experiments were performed at lower temperature (6°C) and lower salinity (15 psu). Two new bacterial strains were also introduced—wild *E. coli* strain isolated from sewage and wild *E. coli* strain isolated from seagull feces.

Experiments in the presence of solar radiation were performed with two ATCC *E. coli* strains, also combining the same temperatures (12, 18 and 24°C) and salinities (30.0 and 36.5 psu). Bacterial suspension was added to 500 mL ultraviolet B (UVB)—blocking borosilicate glass bottles containing autoclaved seawater of appropriate salinity. Bacterial concentration in bottles was also targeted to 103 CFU *E. coli*/100 mL. The bottles were then stirred and left in dark for 15 min to homogenize. Bacterial cells in bottles were then exposed to different intensities of the natural range and spectrum of solar radiation. The exposure to solar light was carried out in two ways: by placing the bottles in 50 L (60 × 30 × 30 cm) transparent plastic containers, filled with water of appropriate and controlled temperature, and placed on the open space in front of the laboratory;—by hanging the bottles from research vessel, vertically on rope to different depths (0.2, 5, 15 and 30 m) of the water column, to expose them to different intensities of solar light (for *E. coli* ATCC 8739 only). In both cases, the irradiance, Ed (μWcm−2nm−1), of ultraviolet A (UVA) and photosynthetically active radiation (PAR) was measured by optical radiometer PRR800 (Biospherical Inc.).

All experiments were performed in triplicates. The initial concentrations of culturable *E. coli* cells in all experiments were determined by taking 10 mL subsamples 15 min after adding bacterial suspension. After exposing the bottles to desired conditions, the concentrations of *E. coli* were monitored by taking 10 mL subsamples every 24 h in experiments performed in the dark and every 10 or 20 min in experiments performed in solar light.

### **2.2. Experiments with natural samples**

microorganisms that can enter the sea, time‐consuming and complex procedures of their determination, the assessment of microbiological quality of seawater has traditionally been based on the determination of indicator microorganisms, bacteria that suggest the presence of pathogens [1]. In the last decade, European Union member states have accepted *Escherichia coli* as the indicator for assessment of microbiological quality of bathing and shellfish waters [2, 3]. Apart from its human and warm‐blooded animal feces origin, selection of *E. coli* as indicator organisms is based on scientific understanding and reliable research of one of the most important criteria which good indicator have to meet: resistance to environmental conditions. When released into the sea, *E. coli* cells are exposed to the impact of a very hostile environment. That impact is reflected in the negative effect of the complex array of biotic and abiotic factors of the marine environment. Among abiotic factors, the effect of temperature, salinity, and particularly solar radiation is most pronounced [4–7]. The distribution and abundance of indicator bacteria in seawater depend mostly on their input, but also on the intensity of aforementioned environmental factors [8] and bacterial cell adaptation capacity. Due to their minor adaptation capacity to marine conditions, *E. coli* cells suffer a sublethal injury and enter a dormant, viable but nonculturable state, in which they can still maintain some metabolic activity [9–11], infective capacity and potential for pathogenicity [12, 13]. The "viable but nonculturable" state concept was introduced to describe cells that remain metabolically active but are unable to divide in or on nutritional media that normally support their growth. The capacity to form colonies on a solid medium is the first ability that enteric bacteria lose in seawater [14]. That means that they cannot be detected during standard routine monitoring by culturing method unless resuscitation methods on liquid media were used. With the prolonged exposure, particularly in the presence of solar radiation, *E. coli* cells are irreversibly inactivated and they die [15]. For how long they will maintain culturability on standard media depends on many factors, such as the intensity of environmental factors and the characteristics of the bacterial cell, where besides the origin and previous growth history, the bacterial strain plays an important role [7, 16, 17]. Although there are many studies in which the effects of the aforementioned factors have been addressed, their simultaneous effects have been poorly investigated. In this chapter, the separate and combined effect of temperature, salinity and solar radiation, as well as the growth history, strain and origin of *E. coli* cells on their survival in seawater were presented. Since microbial pollution of bathing waters is routinely monitored by culturing methods on standard solid media, and viable but nonculturable cells usually cannot be detected, for the purposes of this chapter, the term "survive" means to maintain

170 *Escherichia coli* Escherichia coli - Recent Advances on Physiology, Pathogenesis and Biotechnological Applications - Recent Advances on Physiology, Pathogenesis and Biotechnological Applications

culturability.

**2. Materials and methods**

**2.1. Experiments with pure bacterial cultures**

Bacterial cultures used in experiments were always freshly prepared by incubation of pure culture on mineral‐modified glutamate broth (MMGB). Bacterial suspensions were taken in the exponential phase of growth and serially diluted in phosphate buffer solution. After Natural samples of moderately polluted seawater were collected near sewage outfalls and were stored at 4°C. Sampling was performed in the morning (9 o'clock AM) and at noon. The number of culturable *E. coli* was determined after sampling and every 24 h until their reduction to zero. After isolation and counting on selective agar plates, one colony of *E. coli* cells that survived T90 (the time required to reduce culturability by 90%) or longest was randomly selected from each plate of samples collected in the morning. Cultures were incubated on MMGB and processed in the same way as pure cultures of ATCC strains. A bacterial suspension was added to 500 mL borosilicate bottles, containing autoclaved, 36.5 psu seawater, and stored in dark at 4°C. The number of culturable *E. coli* was determined every 24 h until their reduction to zero.

### **2.3.** *E. coli* **determination**

The concentration of *E. coli* in suspensions was determined by the direct method of epifluo‐ rescence microscopy [18]. The number of culturable *E. coli* cells in suspension and seawater was determined by the modified ISO/TS 16649‐1 method [19] using membrane filtration technique and expressed as CFU *E. coli*/100 mL.

### **2.4. Statistical analysis**

Data were processed using statistical package Statistica 8.0 (StatSoft Inc., 2007). High and significant (*R*<sup>2</sup>  > 0.9, *p* < 0.01) fittings of raw die‐off data of *E. coli* to exponential function, which can be expressed by Eq. (1), were found in the dark as well as in the presence of solar radiation. Consequently, T90 was derived from Eq. (1) and was calculated by Eq. (2),

$$N\_t = N\_0 e^{\lambda t} \tag{1}$$

The results of two‐way ANOVA revealed that there were no statistically significant (*p* > 0.05) separate and/or interactive effects of the variations in temperature and salinity in their natural range, on changes in T90 values of ATCC *E. coli* strains. Unlike the T90 values, these results were not consistent with those of previous studies. In general, most studies showed enhanced stability of indicator bacteria at lower temperatures [4, 5], although Anderson et al. [21] found a lower stability and negative effect. The increased inactivation of indicator microorganisms at higher temperatures was mostly attributed to increased metabolic activities in terms of reduced nutrient concentration [22], increasing predation, grazing and the deleterious effect of solar radiation [23]. One of the possible explanations for the absence of clear effect of temperature in our experiments might be the absence of grazing and predation, since seawater was sterilized by autoclaving. Anderson et al. [21] attributed a positive effect of temperature in their study mostly to sublethal stress that was induced by laboratory manipulation and related to the pre‐exposure history of the used isolates. In our additional experiments with *E. coli* ATCC 8739 at 6°C, significantly higher T90 values were recorded (**Figure 1**), indicating more significant negative effect of temperature in its natural range than at lower values. A positive effect of temperature in whole measured range

Effect of Environmental Conditions on *Escherichia coli* Survival in Seawater

http://dx.doi.org/10.5772/67912

173

was observed in seagull *E. coli* at 36.5 psu and sewage *E. coli* at 30.0 psu (**Figure 2**).

**Figure 1.** Effect of temperature and salinity on the survival of ATCC 8739 and ATCC 35218 *E. coli* strains in the absence

of solar radiation (mean values ± SD).

$$T\_{90} = \frac{-\ln(0.1)}{k} \tag{2}$$

where *N*<sup>0</sup> and *N*<sup>t</sup> are the number of culturable *E. coli* at the beginning of the experiment and at the time of subsampling (*t*).

The inactivation energy S90 (the insolation required to reduce culturability by 90%) was calculated by Eq. (3),

$$S\_{y0}(\text{Vhm}^{-2}) = \sum\_{\nu=0}^{T\_w} E\_\nu \, T\_\nu \tag{3}$$

where *E*n (Wm−2) = the intensity of solar radiation.

### **3. Results and discussion**

### **3.1. Experiments in the absence of solar radiation**

In the natural range of temperature and salinity, the T90 values in this study were 31.9–51.7 h (mean 42.5 h) for *E. coli* ATCC 35218 and 29.4–37.9 h (mean 33.55 h) for *E. coli* ATCC 8739 (**Figure 1**). The results were mostly consistent with the results of previous studies. The T90 values of *E. coli* reported in earlier studies were from 26 h, which was found in estuarine water at 20°C [20], to 115 h, which was found for fecal coliforms in seawater at 8–10°C [5].

The results of two‐way ANOVA revealed that there were no statistically significant (*p* > 0.05) separate and/or interactive effects of the variations in temperature and salinity in their natural range, on changes in T90 values of ATCC *E. coli* strains. Unlike the T90 values, these results were not consistent with those of previous studies. In general, most studies showed enhanced stability of indicator bacteria at lower temperatures [4, 5], although Anderson et al. [21] found a lower stability and negative effect. The increased inactivation of indicator microorganisms at higher temperatures was mostly attributed to increased metabolic activities in terms of reduced nutrient concentration [22], increasing predation, grazing and the deleterious effect of solar radiation [23]. One of the possible explanations for the absence of clear effect of temperature in our experiments might be the absence of grazing and predation, since seawater was sterilized by autoclaving. Anderson et al. [21] attributed a positive effect of temperature in their study mostly to sublethal stress that was induced by laboratory manipulation and related to the pre‐exposure history of the used isolates. In our additional experiments with *E. coli* ATCC 8739 at 6°C, significantly higher T90 values were recorded (**Figure 1**), indicating more significant negative effect of temperature in its natural range than at lower values. A positive effect of temperature in whole measured range was observed in seagull *E. coli* at 36.5 psu and sewage *E. coli* at 30.0 psu (**Figure 2**).

reduction to zero. After isolation and counting on selective agar plates, one colony of *E. coli* cells that survived T90 (the time required to reduce culturability by 90%) or longest was randomly selected from each plate of samples collected in the morning. Cultures were incubated on MMGB and processed in the same way as pure cultures of ATCC strains. A bacterial suspension was added to 500 mL borosilicate bottles, containing autoclaved, 36.5 psu seawater, and stored in dark at 4°C. The number of culturable *E. coli* was determined every 24 h until

172 *Escherichia coli* Escherichia coli - Recent Advances on Physiology, Pathogenesis and Biotechnological Applications - Recent Advances on Physiology, Pathogenesis and Biotechnological Applications

The concentration of *E. coli* in suspensions was determined by the direct method of epifluo‐ rescence microscopy [18]. The number of culturable *E. coli* cells in suspension and seawater was determined by the modified ISO/TS 16649‐1 method [19] using membrane filtration tech-

Data were processed using statistical package Statistica 8.0 (StatSoft Inc., 2007). High and sig-

can be expressed by Eq. (1), were found in the dark as well as in the presence of solar radia-

*Nt* = *N*<sup>0</sup> *e kt* (1)

The inactivation energy S90 (the insolation required to reduce culturability by 90%) was cal-

In the natural range of temperature and salinity, the T90 values in this study were 31.9–51.7 h (mean 42.5 h) for *E. coli* ATCC 35218 and 29.4–37.9 h (mean 33.55 h) for *E. coli* ATCC 8739 (**Figure 1**). The results were mostly consistent with the results of previous studies. The T90 values of *E. coli* reported in earlier studies were from 26 h, which was found in estuarine water

at 20°C [20], to 115 h, which was found for fecal coliforms in seawater at 8–10°C [5].

*n*=0 *T*90

tion. Consequently, T90 was derived from Eq. (1) and was calculated by Eq. (2),

 > 0.9, *p* < 0.01) fittings of raw die‐off data of *E. coli* to exponential function, which

are the number of culturable *E. coli* at the beginning of the experiment and

*<sup>k</sup>* (2)

*En Tn* (3)

their reduction to zero.

**2.3.** *E. coli* **determination**

**2.4. Statistical analysis**

and *N*<sup>t</sup>

culated by Eq. (3),

at the time of subsampling (*t*).

**3. Results and discussion**

nificant (*R*<sup>2</sup>

where *N*<sup>0</sup>

nique and expressed as CFU *E. coli*/100 mL.

*<sup>T</sup>*<sup>90</sup> <sup>=</sup> −ln(0.1) \_\_\_\_\_\_

*S*90(Whm−2 ) = ∑

where *E*n (Wm−2) = the intensity of solar radiation.

**3.1. Experiments in the absence of solar radiation**

**Figure 1.** Effect of temperature and salinity on the survival of ATCC 8739 and ATCC 35218 *E. coli* strains in the absence of solar radiation (mean values ± SD).

**Figure 2.** Effect of temperature and salinity on the survival of *E. coli* strains originating from seagull feces and sewage in the absence of solar radiation (mean values ± SD).

Although not statistically significant, *E. coli* ATCC 35218 survived longer than *E. coli* ATCC 8739 in almost all tested conditions. A significantly higher T90 was found in pure culture of *E. coli* isolated from seagull feces, 28.5–173 h (mean 72.8 h) and particularly in *E. coli* isolated from sewage outfall, 150–390 h (mean 278.6) (**Figure 3**). This suggested the importance of cell strain and origin to their survival in sea water and partially contributed to better understanding of the variations in results of previous studies. Since all tested *E. coli* strains were exposed to the same environmental conditions, the obtained results probably illustrated the extensive genetic and phenotypic diversity exhibited within *E. coli* strains, which could explain the different survival abilities in this study and aquatic environments generally [31]. On the basis of genomic information, *E. coli* species have been divided into eight phylogenetic groups, A, B1, B2, C, D, E, F, and clade I [32]. Strains belonging to phylogroups A and B1 are highly adapted to humans and vertebrate animals, the A phylogroup strains being predominant in humans and the B1 strains in animals [33, 34]. Significant differences in survival in water environment were found among strains belonging to different phylogroups. Recent studies showed that *E. coli* B1 strains can persist longer in water than strains of the other phylogroups [31] and supported the hypothesis that persistent genotypes have an adaptive advantage in the secondary habitat outside the host [35]. Consequently, once released into water, *E. coli* strains could be selected on the basis of their survival ability, and the resulting population differed from the original one in terms of phenotypic traits. *E. coli* strain isolated from sewage outfall and used in this study was in the seawater for an unknown period and probably passed selection that explains its superiority in survival

**Figure 3.** Effect of temperature and salinity in its wider range (15.0, 30.0 and 36.5 psu) on the survival of ATCC 8739 *E.* 

Effect of Environmental Conditions on *Escherichia coli* Survival in Seawater

http://dx.doi.org/10.5772/67912

175

*coli* strain in the absence of solar radiation (mean values ± SD).

over other tested strains.

Most studies showed a negative correlation between survival time of indicator bacteria, including *E. coli*, and salinity [3, 7, 24, 25]. Troussellier et al. [26] found negative effect of salinity only in the presence of solar light. An unclear correlation between T90 and salinity in our study was partly attributed to a narrow salinity range. In order to extend the range of salinity and to clarify its effect on T90, we performed additional experiments with *E. coli* ATCC 8739 at 15 psu. At higher temperatures (18 and 24°C), negative effect of lower salinity was less pronounced than the effect of salinity in its natural range, while at 6°C, lower salinity significantly lowered T90, making the effect of salinity additionally unclear (**Figure 3**). A negative effect of salinity is attributed to specific characteristics of sea water, such as osmotic pressure [27] and the toxicity of inorganic salts [6]. Survival largely depends on the osmoregulatory ability of each strain or group of bacteria [28]. In order to equalize osmotic pressure and avoid drastic loss of water from the cytoplasm, bacterial cells can accumulate or synthesize specific osmoprotectant molecules [29] which increase cell resistance to seawater. This depends, among other things, on preadaptation media. Since *E. coli* in experiments was incubated on MMGB media, accumulation of glutamate, which is recognized as osmoprotectant [30], probably caused an unclear effect of salinity on *E. coli* survival.

**Figure 3.** Effect of temperature and salinity in its wider range (15.0, 30.0 and 36.5 psu) on the survival of ATCC 8739 *E. coli* strain in the absence of solar radiation (mean values ± SD).

Although not statistically significant, *E. coli* ATCC 35218 survived longer than *E. coli* ATCC 8739 in almost all tested conditions. A significantly higher T90 was found in pure culture of *E. coli* isolated from seagull feces, 28.5–173 h (mean 72.8 h) and particularly in *E. coli* isolated from sewage outfall, 150–390 h (mean 278.6) (**Figure 3**). This suggested the importance of cell strain and origin to their survival in sea water and partially contributed to better understanding of the variations in results of previous studies. Since all tested *E. coli* strains were exposed to the same environmental conditions, the obtained results probably illustrated the extensive genetic and phenotypic diversity exhibited within *E. coli* strains, which could explain the different survival abilities in this study and aquatic environments generally [31]. On the basis of genomic information, *E. coli* species have been divided into eight phylogenetic groups, A, B1, B2, C, D, E, F, and clade I [32]. Strains belonging to phylogroups A and B1 are highly adapted to humans and vertebrate animals, the A phylogroup strains being predominant in humans and the B1 strains in animals [33, 34]. Significant differences in survival in water environment were found among strains belonging to different phylogroups. Recent studies showed that *E. coli* B1 strains can persist longer in water than strains of the other phylogroups [31] and supported the hypothesis that persistent genotypes have an adaptive advantage in the secondary habitat outside the host [35]. Consequently, once released into water, *E. coli* strains could be selected on the basis of their survival ability, and the resulting population differed from the original one in terms of phenotypic traits. *E. coli* strain isolated from sewage outfall and used in this study was in the seawater for an unknown period and probably passed selection that explains its superiority in survival over other tested strains.

Most studies showed a negative correlation between survival time of indicator bacteria, including *E. coli*, and salinity [3, 7, 24, 25]. Troussellier et al. [26] found negative effect of salinity only in the presence of solar light. An unclear correlation between T90 and salinity in our study was partly attributed to a narrow salinity range. In order to extend the range of salinity and to clarify its effect on T90, we performed additional experiments with *E. coli* ATCC 8739 at 15 psu. At higher temperatures (18 and 24°C), negative effect of lower salinity was less pronounced than the effect of salinity in its natural range, while at 6°C, lower salinity significantly lowered T90, making the effect of salinity additionally unclear (**Figure 3**). A negative effect of salinity is attributed to specific characteristics of sea water, such as osmotic pressure [27] and the toxicity of inorganic salts [6]. Survival largely depends on the osmoregulatory ability of each strain or group of bacteria [28]. In order to equalize osmotic pressure and avoid drastic loss of water from the cytoplasm, bacterial cells can accumulate or synthesize specific osmoprotectant molecules [29] which increase cell resistance to seawater. This depends, among other things, on preadaptation media. Since *E. coli* in experiments was incubated on MMGB media, accumulation of glutamate, which is recognized as osmoprotec-

**Figure 2.** Effect of temperature and salinity on the survival of *E. coli* strains originating from seagull feces and sewage in

174 *Escherichia coli* Escherichia coli - Recent Advances on Physiology, Pathogenesis and Biotechnological Applications - Recent Advances on Physiology, Pathogenesis and Biotechnological Applications

tant [30], probably caused an unclear effect of salinity on *E. coli* survival.

the absence of solar radiation (mean values ± SD).

Anderson et al. [7] found significant differences in the survival of indicator bacteria, *E. coli* and enterococci, depending on their origin. Among three investigated sources of bacteria; soil, sewage and dog feces, bacteria that originated from soil showed the highest resistance to environmental factors. These results are very important, particularly because apart human feces, there are many potential sources of enteric bacteria that enter the sea, such as untreated piggery effluents [36], and the feces of wild birds [37] and other warm‐blooded animals.

 *T*90(*h* ) = 31.305 *e* −0.17*EUVA* (4) According to the function, when *E*UVA = 0 Wm−2, T90 value of *E. coli* ATCC 8739 amounted to 31.305, which corresponds to its value in the previous experiments in the absence of sunlight

Effect of Environmental Conditions on *Escherichia coli* Survival in Seawater

http://dx.doi.org/10.5772/67912

177

**Figure 4.** Effect of the UVA part of the solar radiation spectrum on the survival of ATCC 8739 and ATCC 35218 *E. coli*

**Variable rp Beta a b R2**

S 0.380 0.143 0.011 UVA −0.830\* −0.955 −0.024 PAR 0.211 0.323 0.001

S 0.319 0.107 0.711 UVA −0.8465\* −0.816 −0.131 PAR 0.438 0.924 0.017

T −0.116 −0.208 1.183 −0.011 0.8797\*

T −0.463 −1.043 7.882 −0.488 0.911\*

**Table 1.** Simultaneous effect of temperature (T), salinity (S), and solar radiation (UVA and PAR) on the T90 of *E. coli* (rp coefficient of partial correlation; beta‐regression coefficients; a—intercept; b—coefficient of multiple linear regression;

(**Figure 1**), which confirms the validity of function.

strains.

*E. coli* ATCC 35218

*E. coli* **ATCC 8739**

\*p < 0.01.

—coefficient of multiple determination).

*R2*

In seagull *E. coli*, a positive effect of salinity was recorded in whole range of temperature, and at 36.5 psu higher temperature enhanced bacteria survival (**Figure 2**). Some *E. coli* cells may adapt to a range of concentrations of sodium chloride, possibly by means of osmoregulatory mechanism induced by salts or some structural modification in the outer membrane of bacteria [27]. Seagulls are marine birds and sea environment is their natural habitat. They feed mainly marine animals that contain high content of sodium chloride and drink both, fresh and seawater. Before being processed and excreted through salt gland in order to maintain osmotic balance, sodium chloride is present in intestinal tract content in a relative high concentration [38]. Consequently, seagull intestinal microflora, including *E. coli*, is highly adapted to saline environment. Such cells probably survive longer in seawater than other cells. Seagulls present a big problem to shellfish farms because they usually gather on buoys and other farm constructions, and huge quantities of feces enter the sea and contaminate it. Since up to 107  CFU *E. coli* g−1 was found in seagull feces in this study, and the results suggested better survival of *E. coli* originating from seagull feces, there is higher probability of their accumulation in filter feeding bivalves than *E. coli* that survive shorter.

### **3.2. Experiments in the presence of solar radiation**

The intensity of solar radiation in this study was in the range 258–693 Wm−2, with a mean contribution of the UVA spectrum of 9.3–11.1%. In the experiments carried out in the laboratory, a strong (*R*<sup>2</sup>  > 0.82) and statistically significant (*p* < 0.01) negative correlation between the T90 of *E. coli* and the intensity of the UVA spectrum of solar radiation was found (**Figure 4**). The T90 of *E. coli* ATCC 35218 was in the range 0.30–0.82 h, whereas the T90 of *E. coli* ATCC 8739 ranged from 0.31 to 5.93 h. A significant, 15‐ to 70‐fold reduction in T90, compared with that recorded in the absence of solar radiation indicated the dominant effect of solar radiation on the survival of *E. coli* in seawater. Multiple linear regressions were used to investigate the combined effect of temperature, salinity, and the intensity of solar radiation on changes in the T90 of *E. coli* (**Table 1**). High and statistically significant (*p* < 0.01) coefficients of multiple determination (*R*<sup>2</sup> ) were found, which revealed that most of the variance in T90 can be explained by independent variables. However, only variations in solar radiation had a significant effect on T90.

Because of inadequate weather conditions (a relative cloudy sky), in situ experiments were carried out at lower intensities of solar radiation. We also found a strong vertical decline in the intensity of solar radiation (UVA and PAR) and a significant decrease in the contribution of the UVA spectrum in the water column (**Figures 5** and **6**). Therefore, only a few data from surface layer could be used in our calculations. Unlike the experiments carried out in the laboratory, T90 of *E. coli* fitted exponential function (**Figure 7**) which can be expressed by Eq. (4)**.**

$$T\_{s0}(h \text{ )}=\Im 1.\Im 05 \text{ } e^{-0.17E\_{\text{max}}}\tag{4}$$

According to the function, when *E*UVA = 0 Wm−2, T90 value of *E. coli* ATCC 8739 amounted to 31.305, which corresponds to its value in the previous experiments in the absence of sunlight (**Figure 1**), which confirms the validity of function.

Anderson et al. [7] found significant differences in the survival of indicator bacteria, *E. coli* and enterococci, depending on their origin. Among three investigated sources of bacteria; soil, sewage and dog feces, bacteria that originated from soil showed the highest resistance to environmental factors. These results are very important, particularly because apart human feces, there are many potential sources of enteric bacteria that enter the sea, such as untreated piggery effluents [36], and the feces of wild birds [37] and other warm‐blooded animals.

176 *Escherichia coli* Escherichia coli - Recent Advances on Physiology, Pathogenesis and Biotechnological Applications - Recent Advances on Physiology, Pathogenesis and Biotechnological Applications

In seagull *E. coli*, a positive effect of salinity was recorded in whole range of temperature, and at 36.5 psu higher temperature enhanced bacteria survival (**Figure 2**). Some *E. coli* cells may adapt to a range of concentrations of sodium chloride, possibly by means of osmoregulatory mechanism induced by salts or some structural modification in the outer membrane of bacteria [27]. Seagulls are marine birds and sea environment is their natural habitat. They feed mainly marine animals that contain high content of sodium chloride and drink both, fresh and seawater. Before being processed and excreted through salt gland in order to maintain osmotic balance, sodium chloride is present in intestinal tract content in a relative high concentration [38]. Consequently, seagull intestinal microflora, including *E. coli*, is highly adapted to saline environment. Such cells probably survive longer in seawater than other cells. Seagulls present a big problem to shellfish farms because they usually gather on buoys and other farm constructions, and huge quantities of feces enter the sea and contaminate it. Since up to 107

*E. coli* g−1 was found in seagull feces in this study, and the results suggested better survival of *E. coli* originating from seagull feces, there is higher probability of their accumulation in filter

The intensity of solar radiation in this study was in the range 258–693 Wm−2, with a mean contribution of the UVA spectrum of 9.3–11.1%. In the experiments carried out in the

between the T90 of *E. coli* and the intensity of the UVA spectrum of solar radiation was found (**Figure 4**). The T90 of *E. coli* ATCC 35218 was in the range 0.30–0.82 h, whereas the T90 of *E. coli* ATCC 8739 ranged from 0.31 to 5.93 h. A significant, 15‐ to 70‐fold reduction in T90, compared with that recorded in the absence of solar radiation indicated the dominant effect of solar radiation on the survival of *E. coli* in seawater. Multiple linear regressions were used to investigate the combined effect of temperature, salinity, and the intensity of solar radiation on changes in the T90 of *E. coli* (**Table 1**). High and statistically significant

of the variance in T90 can be explained by independent variables. However, only variations

Because of inadequate weather conditions (a relative cloudy sky), in situ experiments were carried out at lower intensities of solar radiation. We also found a strong vertical decline in the intensity of solar radiation (UVA and PAR) and a significant decrease in the contribution of the UVA spectrum in the water column (**Figures 5** and **6**). Therefore, only a few data from surface layer could be used in our calculations. Unlike the experiments carried out in the laboratory, T90 of *E. coli* fitted exponential function (**Figure 7**) which can be expressed by Eq. (4)**.**

 > 0.82) and statistically significant (*p* < 0.01) negative correlation

) were found, which revealed that most

feeding bivalves than *E. coli* that survive shorter.

laboratory, a strong (*R*<sup>2</sup>

**3.2. Experiments in the presence of solar radiation**

(*p* < 0.01) coefficients of multiple determination (*R*<sup>2</sup>

in solar radiation had a significant effect on T90.

 CFU

**Figure 4.** Effect of the UVA part of the solar radiation spectrum on the survival of ATCC 8739 and ATCC 35218 *E. coli* strains.


**Table 1.** Simultaneous effect of temperature (T), salinity (S), and solar radiation (UVA and PAR) on the T90 of *E. coli* (rp coefficient of partial correlation; beta‐regression coefficients; a—intercept; b—coefficient of multiple linear regression; *R2* —coefficient of multiple determination).

**Figure 5.** Attenuation of irradiance (% of surface value) as a function of sea depth.

Solar light is considered to be the most important factor to bacterial reduction in the sea, although its effects are restricted to shallow depths [5], also observed in this study. The negative effects of solar radiation to bacterial cells operate via two different mechanisms: firstly, direct photobiological mechanisms break DNA bonds in bacterial cells [39, 40], while secondly, indirect photochemical mechanisms damage bacterial cells by photosensitized reactions initiated by some of endogenous and/or exogenous sensitizers [41]. Photochemical mechanisms become more important at higher wavelengths, and it is more injurious in the presence of oxygen [5]. Both effects cause a rapid decrease in colony forming ability [25, 26]. Dominant toxic effect of sunlight was also observed in previous studies. Šolić and Krstulović [4] found that within the investigated range of intensity of solar radiation (510–830 Wm−2), the T90 of fecal coliforms exponentially decreased by about 40% when the intensity of solar radiation increased by 100 Wm−2. Large differences (30‐ to 40‐fold) in survival with or without exposure to sunlight were observed by Fujioka et al. [25], who noted a T90 of fecal coliforms exposed to solar radiation in the range of 0.5–1.5 h. The reduction in survival time in the presence of solar radiation was also recorded by Chandran and Hatha [20], where the T90 of *E. coli*

**Figure 7.** Effect of the lower intensity of UVA part of the solar radiation spectrum on the survival of ATCC 8739 *E. coli*

Effect of Environmental Conditions on *Escherichia coli* Survival in Seawater

http://dx.doi.org/10.5772/67912

179

According to Davies‐Colley et al. [42], T90 is a better indicator for expressing inactivation in the absence of solar radiation and also inactivation caused by solar radiation of an equable and higher irradiance, whereas the S90 better expresses the inactivation caused by solar radiation of variable intensity, as exists in nature. The energy of solar radiation absorbed by bacterial cells and which is responsible for cell injury is a product of the intensity of solar radiation and

was reduced from 26 h in the dark to 4 h in the presence of solar radiation.

exposure time.

strain.

**Figure 6.** The reduction in the contribution of the UVA component of solar radiation as a function of sea depth.

The negative effect of sunlight exposure was also confirmed in the study with wild cultures of *E. coli* kept at 4°C. A significantly lower survival (T90 = 98 ± 22 h, n = 18) was found in *E. coli* sampled at noon than in those sampled in the morning (T90 = 153 ± 35 h, n = 18). This might be attributed to prolonged exposure to solar radiation and its higher intensity.

**Figure 7.** Effect of the lower intensity of UVA part of the solar radiation spectrum on the survival of ATCC 8739 *E. coli* strain.

Solar light is considered to be the most important factor to bacterial reduction in the sea, although its effects are restricted to shallow depths [5], also observed in this study. The negative effects of solar radiation to bacterial cells operate via two different mechanisms: firstly, direct photobiological mechanisms break DNA bonds in bacterial cells [39, 40], while secondly, indirect photochemical mechanisms damage bacterial cells by photosensitized reactions initiated by some of endogenous and/or exogenous sensitizers [41]. Photochemical mechanisms become more important at higher wavelengths, and it is more injurious in the presence of oxygen [5]. Both effects cause a rapid decrease in colony forming ability [25, 26]. Dominant toxic effect of sunlight was also observed in previous studies. Šolić and Krstulović [4] found that within the investigated range of intensity of solar radiation (510–830 Wm−2), the T90 of fecal coliforms exponentially decreased by about 40% when the intensity of solar radiation increased by 100 Wm−2. Large differences (30‐ to 40‐fold) in survival with or without exposure to sunlight were observed by Fujioka et al. [25], who noted a T90 of fecal coliforms exposed to solar radiation in the range of 0.5–1.5 h. The reduction in survival time in the presence of solar radiation was also recorded by Chandran and Hatha [20], where the T90 of *E. coli* was reduced from 26 h in the dark to 4 h in the presence of solar radiation.

According to Davies‐Colley et al. [42], T90 is a better indicator for expressing inactivation in the absence of solar radiation and also inactivation caused by solar radiation of an equable and higher irradiance, whereas the S90 better expresses the inactivation caused by solar radiation of variable intensity, as exists in nature. The energy of solar radiation absorbed by bacterial cells and which is responsible for cell injury is a product of the intensity of solar radiation and exposure time.

The negative effect of sunlight exposure was also confirmed in the study with wild cultures of *E. coli* kept at 4°C. A significantly lower survival (T90 = 98 ± 22 h, n = 18) was found in *E. coli* sampled at noon than in those sampled in the morning (T90 = 153 ± 35 h, n = 18). This might be

**Figure 6.** The reduction in the contribution of the UVA component of solar radiation as a function of sea depth.

attributed to prolonged exposure to solar radiation and its higher intensity.

**Figure 5.** Attenuation of irradiance (% of surface value) as a function of sea depth.

178 *Escherichia coli* Escherichia coli - Recent Advances on Physiology, Pathogenesis and Biotechnological Applications - Recent Advances on Physiology, Pathogenesis and Biotechnological Applications

The mean values of S90 recorded in this study were 250 Whm−2 for *E. coli* ATCC 35218, and 610 Whm−2 for *E. coli* ATCC 8739. We found a medium‐strong (*R*<sup>2</sup>  = 0.457) and strong (*R*<sup>2</sup>  = 0.8282) statistically significant (*p* < 0.01) negative linear correlation between the intensity of the UVA spectrum of solar radiation and S90 values for both *E. coli* strains (**Figure 8**).

A high and significant correlation between the intensity of the UVA spectrum of solar radiation with T90 and S90 in this study suggests that within the studied spectrum of solar radiation, this part of the measured spectrum was most responsible for the inactivation of microorganisms. The linear regression‐slope coefficients suggested the wavelengths 320–360 nm to be the most bactericidal within the UVA spectrum (**Table 2**).

Acra et al. [43] found that up to 70% of the solar inactivation of bacteria can be attributed to the effect of the UVA part of the solar radiation spectrum, whereas Sinton et al. [5] found that 50% of the inactivation of indicator microorganisms could be attributed to solar radiation wavelengths up to 360 nm, with a wavelength of about 330 nm being most bactericidal. Calkins and Barcelo [44] find the UVB (280–330 nm) portion of the solar spectrum the most bactericidal, causing direct photobiological DNA damage. Although bactericidal effect of solar radiation is mainly associated with shorten wavelengths, the synergistic effect of UVB, UVA, and PAR explained most variations found in culturability of *E. coli* [45].

The values of S90 observed in this study were significantly lower than those in similar studies. S90 of fecal coliforms found by Gameson [22] was 1.290 Whm−2, whereas Sinton et al. [5] found a significantly higher value, 1.660 Whm−2 (6.0 MJm−2). Significant differences could be attributed to the different intensity of solar radiation and to variations in intensity, but also to

**Table 2.** Contribution of different wavelengths of measured solar radiation to T90 of *E. coli* expressed through slopes

—coefficient of multiple determination; *p*—level of

**320–340 340–360 360–380 380–400 400–700 (PAR)**

Effect of Environmental Conditions on *Escherichia coli* Survival in Seawater

http://dx.doi.org/10.5772/67912

181

b −0.147 −0.124 −0.072 −0.052 −0.001 *R*<sup>2</sup> 0.442 0.663 0.848 0.848 0.192 *p* <0.01 <0.01 <0.01 <0.01 >0.05

b −1.095 −0.812 −0.488 −0.364 −0.014 *R*<sup>2</sup> 0.741 0.834 0.851 0.864 0.635 *p* <0.01 <0.01 <0.01 <0.01 <0.01

Lower T90 and S90 values in this study might partly be explained by the previous growth history of cell cultures until exposure to the environment, different origin and strain of bacterial cells, as well as by the different intensity of solar radiation and other environmental factors. There are many ways in which bacterial cells can enter the sea. In some cases, they are discharged directly from boats or bathers, and in others, they remain for different periods in wastewater reservoirs and/or are carried out in the sea through natural rivers or artificial conduits [14]. The change from a nutrient‐rich environment to nutrient‐poor one, places *E. coli* cells under starvation stress, which is less pronounced in bacteria from the stationary‐ phase because they have already experienced a starvation adaptation period. As a response to starvation stress, bacterial cells can induce protective mechanisms against UVA stress [46]. Furthermore, starved cultures taken in the stationary‐phase of growth also showed higher osmotic [47] and temperature [48] resistance than those taken in a logarithmic‐phase. This is due to specific protein synthesis during stationary‐phase starvation [43]. In this study, *E. coli* cells were taken during the logarithmic‐phase and were directly transferred to phosphate buffer without washing of nutrients. Since these cells did not experience starvation during the growth phase or during culture transfer, this probably made them less resistant to the negative effects of seawater and oxidative processes of solar radiation than cells that had experienced starvation [47, 49, 50]. Consequently, wild bacterial cells that have not experienced growth in a rich medium and are taken from the stationary phase of growth should show a higher resistance than cultivated cells, as was confirmed in this study. We found a statistically significant, at least threefold lower T90 (50.6 ± 6 h, n = 18) in cultivated wild *E. coli* cells than in their mother cells that were isolated from seawater (153.3 ± 35 h, n = 18), although the latter were pre‐exposed to solar light and hostile environment for unknown period before seawater

different groups of microorganisms tested.

of linear regression (b—coefficient of multiple linear regression; *R2*

**Wavelength (nm)**

*E. coli* **ATCC 35218**

*E. coli* **ATCC 8739**

significance).

samples collected.

**Figure 8.** The S90 of ATCC 8739 and ATCC 35218 *E. coli* strains as a function of the intensity of the UVA part of the solar radiation spectrum.


The mean values of S90 recorded in this study were 250 Whm−2 for *E. coli* ATCC 35218,

A high and significant correlation between the intensity of the UVA spectrum of solar radiation with T90 and S90 in this study suggests that within the studied spectrum of solar radiation, this part of the measured spectrum was most responsible for the inactivation of microorganisms. The linear regression‐slope coefficients suggested the wavelengths 320–360 nm to be the most

Acra et al. [43] found that up to 70% of the solar inactivation of bacteria can be attributed to the effect of the UVA part of the solar radiation spectrum, whereas Sinton et al. [5] found that 50% of the inactivation of indicator microorganisms could be attributed to solar radiation wavelengths up to 360 nm, with a wavelength of about 330 nm being most bactericidal. Calkins and Barcelo [44] find the UVB (280–330 nm) portion of the solar spectrum the most bactericidal, causing direct photobiological DNA damage. Although bactericidal effect of solar radiation is mainly associated with shorten wavelengths, the synergistic effect of UVB,

The values of S90 observed in this study were significantly lower than those in similar studies. S90 of fecal coliforms found by Gameson [22] was 1.290 Whm−2, whereas Sinton et al. [5] found a significantly higher value, 1.660 Whm−2 (6.0 MJm−2). Significant differences could be

**Figure 8.** The S90 of ATCC 8739 and ATCC 35218 *E. coli* strains as a function of the intensity of the UVA part of the solar

of the UVA spectrum of solar radiation and S90 values for both *E. coli* strains (**Figure 8**).

 = 0.8282) statistically significant (*p* < 0.01) negative linear correlation between the intensity

 = 0.457) and strong

and 610 Whm−2 for *E. coli* ATCC 8739. We found a medium‐strong (*R*<sup>2</sup>

180 *Escherichia coli* Escherichia coli - Recent Advances on Physiology, Pathogenesis and Biotechnological Applications - Recent Advances on Physiology, Pathogenesis and Biotechnological Applications

UVA, and PAR explained most variations found in culturability of *E. coli* [45].

bactericidal within the UVA spectrum (**Table 2**).

(*R*<sup>2</sup>

radiation spectrum.

**Table 2.** Contribution of different wavelengths of measured solar radiation to T90 of *E. coli* expressed through slopes of linear regression (b—coefficient of multiple linear regression; *R2* —coefficient of multiple determination; *p*—level of significance).

attributed to the different intensity of solar radiation and to variations in intensity, but also to different groups of microorganisms tested.

Lower T90 and S90 values in this study might partly be explained by the previous growth history of cell cultures until exposure to the environment, different origin and strain of bacterial cells, as well as by the different intensity of solar radiation and other environmental factors. There are many ways in which bacterial cells can enter the sea. In some cases, they are discharged directly from boats or bathers, and in others, they remain for different periods in wastewater reservoirs and/or are carried out in the sea through natural rivers or artificial conduits [14]. The change from a nutrient‐rich environment to nutrient‐poor one, places *E. coli* cells under starvation stress, which is less pronounced in bacteria from the stationary‐ phase because they have already experienced a starvation adaptation period. As a response to starvation stress, bacterial cells can induce protective mechanisms against UVA stress [46]. Furthermore, starved cultures taken in the stationary‐phase of growth also showed higher osmotic [47] and temperature [48] resistance than those taken in a logarithmic‐phase. This is due to specific protein synthesis during stationary‐phase starvation [43]. In this study, *E. coli* cells were taken during the logarithmic‐phase and were directly transferred to phosphate buffer without washing of nutrients. Since these cells did not experience starvation during the growth phase or during culture transfer, this probably made them less resistant to the negative effects of seawater and oxidative processes of solar radiation than cells that had experienced starvation [47, 49, 50]. Consequently, wild bacterial cells that have not experienced growth in a rich medium and are taken from the stationary phase of growth should show a higher resistance than cultivated cells, as was confirmed in this study. We found a statistically significant, at least threefold lower T90 (50.6 ± 6 h, n = 18) in cultivated wild *E. coli* cells than in their mother cells that were isolated from seawater (153.3 ± 35 h, n = 18), although the latter were pre‐exposed to solar light and hostile environment for unknown period before seawater samples collected.

As mentioned previously, *E. coli* ATCC 35218 survived longer than *E. coli* ATCC 8739 in almost all experimental conditions in the dark, but if exposed to solar radiation, *E. coli* ATCC 35218 survived significantly shorter than *E. coli* ATCC 8739. This suggests that these microorganisms probably do not have equally developed mechanisms of protection against various abiotic environmental factors and that the protection rate also depends on which mechanism is more effective under the same conditions.

**References**

[1] Berg G, editor. The indicator system. Indicators of Viruses in Water and Food. Ann Arbor: Ann Arbor Science Publishers; 1978. pp. 1‐13. doi:10.1002/jobm.19800200609 [2] The European Parliament and the Council of the European Union. Regulation (EC) No 854/2004 of the European Parliament and of the Council of 29 April 2004 laying down specific rules for the organisation of official controls on products of animal origin intended for human consumption. Official Journal of the European Union. 2004;L226:83‐127. [3] The European Parliament and the Council of the European Union. Directive 2006/7/EC of 15 February 2006 concerning the management of bathing water quality and repealing directive

Effect of Environmental Conditions on *Escherichia coli* Survival in Seawater

http://dx.doi.org/10.5772/67912

183

[4] Šolić M, Krstulović N. Separate and combined effects of solar radiation, temperature, salinity and pH on the survival of faecal coliforms in seawater. Marine Pollution Bulletin.

[5] Sinton LW, Davies‐Coley RJ, Bell RG. Inactivation of enterococci and fecal coliforms from sewage and meatworks effluents in seawater chambers. Applied and Environmental

[6] Carlucci AF, Pramer D. An evaluation of factors affecting the survival of Escherichia coli in sea water. II. Salinity, pH and nutrients. Applied Microbiology. 1960;8(4):247‐250. [7] Anderson KL, Whitlock JE, Harwood DJ. Persistence and differential survival of fecal indicator bacteria in subtropical waters and sediments. Applied and Environmental

[8] Sivri N, Balci M, Durmus T, Seker DZ, Balkis N. Analysis of enteric bacteria distribution in the Gulf of Gemlik (Marmara Sea) by means of GIS. Fresenius Environmental Bulletin.

[9] Anderson IC, Rhodes M, Kator HI. Sublethal stress in *Escherichia coli*: a function of salin-

[10] Kapuscinski RB, Mitchell R. Solar radiation induces sublethal injury in *Escherichia coli* in

[11] Barcina I, González JM, Iriberri J, Egea L. Survival strategy of *Escherichia coli* and *Enterococcus faecalis* in illuminated fresh and marine systems. Journal of Applied

[12] Grimes DJ, Atwell RW, Brayton PR, Palmer LM. Rollins, DM, Roszak DB, Singleton FL, Tamplin ML, Colwell RR. The fate of enteric pathogenic bacteria in estuarine and marine

[13] Singh A, R Yeager, McFeters GA. Assessment of in vivo revival, growth, and pathogenicity of *Escherichia coli* strains after copper‐ and chlorine‐induced injury. Applied and

Microbiology. 2005;71(6):3041‐3048. doi:10.1128/AEM.71.6.3041‐3048.2005

ity. Applied and Environmental Microbiology. 1979;38(6):1147‐1152.

seawater. Applied and Environmental Microbiology. 1981;41(3):670‐674.

Bacteriology. 1990;68(2):189‐98. doi:10.1111/j.1365‐2672.1990.tb02565.x

environments. Microbiological Sciences. 1986;3(11):324‐329.

Environmental Microbiology. 1986;52(4):832‐837.

76/160/EEC. Official Journal of the European Union. 2006;L64:37‐51.

1992;24(8):411‐416. doi:10.1016/0025‐326X(92)90503‐X

Microbiology. 1994;60(6):2040‐2048.

2012;21(11):3224‐3232.

## **4. Conclusions**

This study showed that in the absence of solar radiation, there were no statistically significant effects of temperature and salinity in their natural range (12, 18 and 24°C; 30.0 and 36.5 psu) on the survival of two ATCC strains. The survival of *E. coli* ATCC 8739 was enhanced only at 6°C, suggesting a negative effect of temperature at temperatures lower than natural range. At 15.0 psu, significantly lower T90 values of *E. coli* ATCC 8739 were observed at lower temperatures (6 and 12°C) than at higher ones (18 and 24°C). In the natural range of temperature and salinity, mean T90 values varied significantly, but only as a function of bacterial strain and origin. The importance of bacterial strain and origin for the survival of *E. coli* was demonstrated by significantly longer survival of *E. coli* cells isolated from seagull feces and sewage outfall. In the same conditions, their survival surpassed the survival of ATCC strains up to 10‐fold. A dominate effect of solar light on *E. coli* survival was confirmed by recording a more rapid *E. coli* died‐off in cells exposed to a natural range of solar radiation. The T90 was 15‐ to 70‐fold shorter than in the absence of solar radiation. Within the investigated range of solar radiation (320–700 nm), only the effect of UVA spectrum was found to be statistically significant. The wavelengths 320–360 were the most bactericidal. The bacterial strain that had a shorter T90 in the absence of solar radiation showed a significantly higher T90 and S90 when exposed to solar light, suggesting different mechanisms of adaptation to studied abiotic environmental factors and a different efficiency of mechanisms under the same conditions. Cultivated bacteria showed threefold shorter T90 than their mother cells isolated from seawater. This suggested the importance of previous growth history of bacterial cells prior to exposure to the environment, for their adaptation capacity. The observed superiority of wild bacterial cultures over cultivated ones was probably a result of starvation experience during the stationary‐phase, and/or due to a lack of nutrients.

## **Author details**

Slaven Jozić\* and Mladen Šolić

\*Address all correspondence to: sjozic@izor.hr

Institute of Oceanography and Fisheries, Split, Croatia
