Introduction

Livestock drinking water is one of the most important sources of bacterial contamination in farm animals (Khan et al., 2016). Determi- nation of water contamination to fecal con- taminants helps us to prevent the water-borne diseases. Microorganisms do not always af- fect the appearance or taste of water but af- fect animal health and milk quality, therefore, continuous microbiological investigation of drinking water of animals is necessary (Din et al., 2014; Ramona, 2015).

Raw water contains two classes of micro- organisms, including permanent microor- ganisms that are naturally inhabited by water and have low nutritional requirements and transient microorganisms that are transmitted from the environment by soil, humans, and animals. Pathogenic types fall in the second category (Shaghaghi, 2019). Total microbial analysis of water is ideal for determining its health quality, but since it is not easy and in- expensive, so agents that are known as indi- cators, such as bacteria in the Coliform group, are used to measure microbial  contamina- tion of water (Shaghaghi, 2019; Shimie and Yousefi,  1997). Escherichia  coli   O157: H7   is

one  of  the  most  dangerous  and  pathogenic

serotypes in humans, while in cattle, it does not cause any clinical disease other than di- arrhea, and animals mainly act as carriers of the bacterium to humans (Bindu et al., 2010). Some other bacteria such as Streptococcus (a non-Coliform indicator of fecal contamina- tion in water) are also considered as a micro- bial contaminant of drinking water (El Emam and El Jalii, 2010). Bacteria enter the drinking water from various sites such as oral and nasal discharge, feces and urine (Van Emon, 2015). The risk of transmission of diseases through stagnant waters is much higher than that of running water (Pooyanmehr et al., 2008).

Most dairy farmers consider that microbial contamination of drinking water is inevitable and cattle are resistant to it, so, less attention have been paid to it and its result will be a re- duction in livestock production (Pooyanmehr et al., 2008). There is some documentation about physical and chemical status of drink- ing water in farms, but little is known about the microbial quality of livestock water in Iran (El Emam and El Jalii, 2010). This study was performed to evaluate the microbial quality of drinking water in cows and compare it with existing standards to eliminate deficiencies in this field.

Materials and methods

In late fall and early winter, 2018, from 30 industrial and semi-industrial dairy farms lo- cated in 4 districts of Mashhad suburb, Iran (north, northwest, east, southeast), samples of drinking water were taken from four parts in- cluding water tanks, beginning (inlet) and end (outlet) of drinking water troughs of cows and water buckets of calves, placed in sterile fal- cons and transferred to the lab with cold chain preservation.

Using the most probable  number  (MPN) or the multiple tubes method (9 tubes meth- od), total and fecal Coliforms were counted. For this purpose, by using sterile pipette, 10 ml of water sample was inoculated in to the three tubes containing 20 ml of selective en- richment medium (lauryl sulfate) with double strength, and in the second and third three tubes, containing single strength concentra- tion of lauryl sulfate (according to the manu- facturer instructions), were inoculated with 1 ml and 0.1 ml of water sample, respectively.

All test tubes contained durham tubes in an inverted position. The test tubes were incu- bated at 37 ° C for 48 h. The tubes containing

gas in the durum tube with obvious turbidity, were considered positive. From each positive tube, one loopful of medium was inoculated into confirmatory culture medium (brilliant green lactose bile broth) containing a durham tube, and incubated for 24 ±2 h at 37 ±1°C. If no gas or turbidity was observed during this period, incubation was continued for another 24±2 h. Then, the tubes containing turbidity and gas were considered positive and fecal Coliform counts were performed based on MPN table.

To count E.coli, a loopful from culture me- dium in the MPN method, which produced gas and turbidity were inoculated into tubes containing Escherichia coli (EC) broth whit durum tubes, and incubated at 44°C for 48 h, if gas or turbidity was observed, a loopful was inoculated into peptone water medium and incubated at 44° C for another 48 h. For all EC tubes which produced gas and turbidity, the IMViC tests consisting of indole, methyl red, Voges-Proskauer (VP) and Simon citrate tests were performed.

Pour plate method was used to count fecal streptococci. Amount of 1 ml of the water sample were transferred to each Petri dish and 15 ml of sterilized KF (Kenner Fecal) medi- um at 45-50°C were added to each plate and mixed thoroughly, after solidifying the agar medium, plates were incubated at 37°C for 48 h, pink and purple to red colonies were count- ed as presumably fecal Streptococci. Number of 10 colonies were selected for the confir- matory tests. Catalase and gram staining tests were used to confirm fecal Streptococci. The bacterial count was calculated in CFU (colo- ny- forming unit) per ml of sample.

In order  to identify  the E. coli  O157: H7 se-

rotype, linear culture method was used in Sorbitol MacConkey agar and nutrient agar mediums  as well as PCR technique  by using

specific primer for O157: H7 antigen genes.

Statistical analysis

Statistical data analysis was performed by using SPSS software, v.22. Both normality test (Kolmogorov-Smirnov and Shapiro-Wilk) re- vealed the nonparametric data, therefore, the median was used as a valid statistical index to interpret the results. Friedman and Wilcoxon signed- rank tests were used to compare the rate of water contamination in different sam- ple sites with each other (overall and pairwise, respectively), also, One  sample Wilcoxon signed rank test was used to compare the data with existing standard. P ≤ 0.05 was consid- ered statistically significant.

Results

This survey showed that the most frequent bacterial contamination (total and fecal Co- liform) were in the outlet of bovine drink- ing troughs, next up, in the water buckets of calves, inlet of bovine drinking troughs and water tanks. For fecal Streptococcus, the lowest and highest frequency were in the water tanks and outlet of bovine drinking troughs, respectively. Another finding was the observation of E.coli O157: H7  serotype

only at the outlet of drinking trough in one

farm and serotype H7 Oumdifimd in two other dairy units (Table 1).

The highest contamination with indicator bacteria was observed at the outlet of bo- vine drinking troughs (Table 2). Pairwise comparison between sampling sites revealed no significant difference for total Coliform and fecal Streptococcal contamination be- tween buckets of calves with inlet of bovine drinking troughs and for fecal Coliform and streptococci, between buckets of calves with outlet of bovine drinking troughs. In other cases, significant difference were observed (P ≤0.05) (Table 3).Comparison between the

Table 1. Frequency of indicator bacteria contamination at the sampling sites of livestock drinking water

Table 2. Statistical indices of contamination with indicator bacteria at the sampling sites of livestock drinking water (Cfu/ml)

Bacteria                       Site                     Tank                  Inlet of drink-

Outlet of drink-

Calves’drink-

index

ing drought

ing drought

ing buckets

Total Coliform

Fecal Coliform

Fecal Streptococcus

Median                     0                              93                                  1100                                 350

Min- Max              0- 1100                     0- 1100                           4- 1100                           2- 1100

Median                     0                              43                                 121.5                                 68

Min- Max               0- 240                      0- 1100                           4- 1100                           0- 1100

Median                     0                              12                                  27.5                               10.50

Min- Max                0- 10                         0- 36                               0- 48                               0- 06

E.Coli O157:H7                     Median                     0                               0                                     0                                      0

Table 3. Statistical comparison of drinking water sampling sites for indicator bacterial contamination

rate of contamination of water samples with indicator bacteria and existing standards (Van Emon, 2015; Beede, 2006) revealed that, for total and fecal Coliform, values due to inlet and outlet of bovine drinking troughs and water buckets of calves were significant- ly higher than expected values. Whereas the rate of contamination with above bacteria in water tanks were significantly lower than the highest standard values (< 50 and < 10, re-

spectively) and in other cases no significant difference were found.

For fecal streptococci, no significant dif- ference was observed for the outlet of bovine drinking troughs compared to the highest ex- pected value (<30), but in others were ob- served (P ≤0.05).

Also, the rate of contamination with the bacterium at water tanks was significantly lower and at intlet of bovine drinking troughs

and water buckets of calves were higher than expected values (Tables 2 and 4).

There was no statistically significant dif- ference between different geographical lo-

cations (North: 14 units, Northwest: 6 units, East: 5 units, Southeast: 5 Unit) in terms of contamination of water tanks with indicator bacteria.

Table 4. Frequency of contamination rate and statistical comparison of drinking water sampling sites in terms of contamination with indicator bacteria by existing standards (colony per 100 ml).

For all three indicator bacteria, the expected value is less than one colony per 100 ml. In the case of total Coliform, fecal Coliform and fecal strepto- cocci for calves, the values >1, >1 and> 3, and for cows the values >15-50, >10, and >30 are problematic and unsafe, respectively.

Discussion

In this survey, the high frequency of con- tamination with indicator bacteria at the outlet of bovine drinking troughs compared to the other sampling sites (Table 1) can be related to the stagnation of water in outlet part (the inlet areas receive fresh water reg- ularly), continuous shedding of materials at- tached to the hair and body of the animals  (it often occurs as a result of sitting on a bed of manure or sticking of stool to the tail and lower area of the animal) into the drinking water or even inserting contaminated muzzle into the water troughs (due to the smelling

behavior of the lower part of other livestock in estrus cows). Therefore, continuous drain- age and cleaning of drinking water troughs is recommended (Pooyanmehr et al., 2008). It seems that the outlet part of water troughs to be a more realistic representative of the water consumed by livestock, because the number of livestock that drink water from inlet part is lower than the other parts, so,it is essential to identify the causes of its pollution.

However, the comparison between the frequency of bacterial contamination in the drinking water of cows and the bucket of calves due to their different hygienic and

breeding conditions does not seem reason- able, because calves’ water buckets are gen- erally specific and washed daily while the cows’ water troughs are public and rarely cleaned, but the results of this comparison are important managerially, because, in spite of the special attention paid by dairy farmer to the health of calves (due to their high sen- sitivity to intestinal diseases), the contamina- tion rate of their water bucket with all indi- cator bacteria were approximately similar to that of cows’ water troughs (Table 2).

High level of water contamination in calves' buckets especially due to fecal Coliform and Streptococci (Tables 2 and 3), indicate secondary contamination by feces or other environmental factor and poor manage- ment of neonatal drinking water. Obviously, the stagnation of water in calves’ buckets during the day compared to the inlet water trough may contribute to its greater contam- ination than the inlet parts. In one study, the high contamination of drinking water in one goat farm were attributed to manual trans- portation of water to the drink trough, hand contamination, and lack of respect for work- ers' personal hygiene during carrying water (Ramona, 2015), so, training dairy workers to accurately wash the buckets and follow the hygienic points during their filling and transporting to the place of consumption will be effective in reducing the rate of contami- nation (Pooyanmehr et al., 2008).

High levels of contamination with fecal streptococci, especially at the outlet water troughs (27.5 cfu/ml), confirms  the release of excreted substances into the water and in- dicates poor health in bovine drinking wa- ter (Tables 2 and 4). Keeping water tanks indoors and away from access to fecal and environmental pollutants such as dust, will justify lower abundance and significant re-

duction of fecal streptococci contamination in them than other sampling sites (Tables 1, 2 and 3). In a survey was performed in sudan, only 4.76% of the isolates were Streptococ- cus bacteria (El Emam and El Jalii, 2010).

If the contamination with fecal Coliforms is more than several times that of fecal Streptococci, it can be suspected to contam- ination by human sources and in contrary to the above condition, contamination by ani- mal ones is considered (Beede, 2006; Loop- er, 2012). In present study, fecal Coliform contamination in calves' water buckets was more than 6 times that of fecal Streptococci (Table 2), So it seems that its contamination has often been caused by human sewage, therefore, more attention to  human  health  is necessary. The presence of E.coli O157: H7 serotype contamination in outlet of bo- vine drinking troughs and its absence in oth- er sampling sites could be due to secondary contamination from human or animal sourc- es in stagnant water (Table 1).

In a study was performed in Balochistan, Pakistan, the rate of Coliform contamination in buffaloes and dairy cattle was reported 17%. and attributed it to the open sewage system, decaying and rusting pipes, and use of inappropriate water troughs (Khan et al., 2016). In the present study, more attention should be paid to the high frequency of con- tamination (33.3% in tanks, up to 100% in outlet water troughs) compared to the recent study. In another study was conducted on drinking water in Quetta, Pakistan, in addi- tion to severe water pollution (especially due to E. coli), reduction the susceptibility of all isolated pathogens to a wide range of antimi- crobial drugs has been identified as a prob- lematic factor in the treatment of water born diseases (Din et al., 2014).

The purpose of this survey was not to in-

vestigate the risk factors associated with bacterial contamination of drinking water, but the field observations and filling out a questionnaire revealed some issues. For ex- ample, in one farm (where the total Coliform count at tank was >1100 Cfu/ml and the fecal Coliform at the inlet and outlet of the water trough were measured: 150 and 1100 Cfu/ml, respectively), the water was first transferred to an outdoor pool, then pumped into the water trough, the designer observa- tions also indicate that the cows' drinking water is very dirty. It is clear that the possi- ble use of the pool for hand washing or even swimming and exposure to dust can increase the chance of Coliform contamination. Also, in another highly contaminated farm (fecal Coliform: 240 cfu/ml), the water of well was first transported to the  cement  pond,  then to the tank and drinking trough by pipeline. Therefore, exposing the pond to animal and bird excreta and stagnating water in it will increase the chance of contamination. Oth- er points that this questionnaire made clear were: No records of microbial water testing or doing a test many years ago in some dairy units and no use of disinfectants for washing of water troughs in any of the dairies. De- spite the presence of five farms in the vicini- ty of the slaughterhouse, their drinking water contamination with the indicator bacteria in the tanks were very low (< 2 cfu/ml), proba- bly due to the use of village disinfected water (which is also used by people).

In all studied  farms,  underground   wa- ter (wells in the farm or outside of it) were used for bovine drinking. The  wide  range of contamination variations in tanks (mini- mum to maximum levels, especially due to total Coliform) (Table 2) may be related to the differences in health status of the tanks  or environmental and physical factors in the

well. Some conditions such as the presence of decaying plant or animal material are conducive to the survival  or  even  growth of waterborne microorganisms, while  oth-  er factors such as high salt, protozoans and bacteriophages can kill millions of bacteria in water. Some physical factors such as tem- perature, pressure, acidity, osmotic pressure (solute content) of water, penetration of air and sunlight into the water are also effective on the number and species of water-borne microorganisms. In general, deep well wa- ter often contains a small number of bacteria that are usually non-pathogenic (Shimie and Yousefi, 1997).

During the last few years, the quality of underground water in Mashhad plain has been severely reduced because of the sharp decline of water in aquifers following the drought and the inflow of various agricultur- al, industrial and urban pollutants in to the water, so, pay attention to the privacy of the wells is essential. It is said that the distance between the sewage and water wells should be sufficient to takes fifty days to reach the pollution to the water well (Alizadeh et al., 2009). One of the issues mentioned in the report of Alizadeh et al. (2009) is the over- whelming expansion of Mashhad city to- wards the adjacent villages and farms. This has caused the contamination of water wells of the farms to the household sewage or dis- charge of waste.

In one study, the absence of Coliform con- tamination in water wells was attributed to several factors, including: the long distance between water wells and human and animal wastewater, disinfection of water with chlo- rine and the proximity of water wells with the cobalt mines (due to the antimicrobial properties of cobalt compounds). The high concentration of nitrite, nitrate and chlorine