Microbial Risk Assessment of Vegetables Irrigated with Akaki River Water in Addis Ababa.

Sisay Derso Mengesha (  sisdres23@yahoo.com ) Ethiopian Public Health Institute https://orcid.org/0000-0002-5753-5677 Yosef Beyene Asfaw EPHI: Ethiopian Public Health Institute Abel Weldetinsae Kidane EPHI: Ethiopian Public Health Institute Kirubel Tesfaye Teklu EPHI: Ethiopian Public Health Institute Melaku Gizaw Serte EPHI: Ethiopian Public Health Institute Moa Abate Kenea EPHI: Ethiopian Public Health Institute Daniel Abera Dinssa EPHI: Ethiopian Public Health Institute Mesay Getachew Woldegabriel EPHI: Ethiopian Public Health Institute Tsigereda Assefa Alemayehu EPHI: Ethiopian Public Health Institute Aderajew Mekonnen Girmay EPHI: Ethiopian Public Health Institute

e uent of wastewater from numerous industries and other gray water from the habitat of the Addis Ababa city is currently used for irrigation purpose as other developing urban areas practiced (Gashaye, 2020; Janeiro, Arsénio, Brito, & van Lier, 2020; Ungureanu, Vlăduț, & Voicu, 2020). As all streams of Addis Ababa, the Akaki River is enormously polluted by anthropogenic impacts from upstream to down (

Study areas
The study was conducted in vegetable farmlands irrigated with Akaki River water in Addis Ababa, Ethiopia. During our reconnaissance survey, we have identi ed 14 farmlands in and near Addis Ababa for this study. These farmlands include ( gure 1) Burayu Kera, Burayu Gefersa, Lome Meda, Koka, Mekanissa Teklehaymanot, Mekanissa, Kera, Bihere Tsige, and Shitu along the Little Akaki River and Peacock1, Peacock2, Akaki Tirunesh Beijing, and Akaki Beseka in the catchment of Great Akaki River. The Aba Samuel farmland is located downstream of the Akaki Rivers and used water from the manmade dam. The description of each sampling point is summarized in Table 1.

Research Design
A laboratory-based cross-sectional study was conducted on vegetables that are irrigated with Akaki River water during dry and wet seasons.

Sample collection and analysis
Fresh vegetables such as Ethiopian Kale, Lettuce, cabbage, and spinach were collected from farms irrigated by the Akaki River in the dry and wet seasons. During the dry month of February 2017, and in the wet season of August 2016, 38 and 15 vegetable samples were collected, respectively. The Vegetable samples were taken from every corner of the plots to make sampling representative. 500 gm of vegetables were collected from each sampling site in sterilized plastic bags. Besides, 500 ml water samples were collected from the Akaki River irrigation area using sterilized bottles. Collected vegetables and water samples were stored in a refrigerator at 4 o C to prevent lysis of microorganisms and subjected to analysis within 24 hours.
Microbial analyses of vegetables and water samples were done at Ethiopian Public Health Institute (EPHI) microbiology laboratory. The membrane ltration method was used for microbiological analysis of water samples following Ethiopian standard agency method number ES ISO 4833 (ESA, 2013) as mentioned by (Alemayehu et al., 2020). Enumeration of Total coliform, fecal coliform, and E. coli was done from vegetable samples using standard methods (Refai, 1979). 25 g of vegetable samples were taken from each vegetable sample, rinsed with 225 ml of peptone water, and shake with a stomacher to wash the bacteria from the vegetables and get into the solution. E-coli and non-E-Coli coliform count were done from the water samples using the standard method (U. J. . 1 ml, 10 ml, and 50 ml of the water samples were ltered using 0.45 whole membrane ltration for each water sample and the membranes were incubated at 44 O c in different Petri-dishes that have appropriate culture media for coliforms. The growth bacterial colonies were counted from those Petri-dishes that are convenient to count. Con rmation for E-coli was done using an indol test from the counted bacterial colonies. For vegetables, the number of bacterial colonies count was reported per gram of vegetables after calculating with the dilution factors and for water samples, it was reported per 10 ml of a water sample. Quality controls for analysis were done using standard organisms as a positive control and blank to check contamination during analysis. Tests that did not pass the quality control were repeated.

Data Analysis
The microbial load of vegetables and average of microbial population per gram of vegetable is calculated after normalized by log transformation except fecal coliforms. Also, descriptive statistics were conducted for Escherichia coli and non-E.Coli coliforms.

Microbial Contamination of Water and fresh vegetables
The present study attempted to determine the level of vegetable contamination with Total Aerobic plate Count (TAC), Total coliform bacteria and fecal coliforms (FC) as well as their microbial loads through Total Aerobic Count (TAC), Total Coliform Counts (TCC), and fecal coliforms (FC). This study also showed the percentage of Akaki River water contamination with E. coli and non-E. Coli coliforms.

Microbial contamination of River water
In the dry period, 14 water samples were collected from the inlet of irrigation water at Akaki River. The result is summarized by each sampling point and the average count of Escherichia Coli and non-Escherichia Coli coliform populations. Table 2 shows the distribution of E. coli and Non-E.coli coliform in the Akaki River that is used for irrigation of farmland in Adds Ababa. The minimum E.coli population of 10 CFU per 10 ml was measured in Aba Samuel and the maximum E.coli population with 320 CFU/10 ml was found in Akaki Beseka and Mekanissa. On the other hand, the non-E. Coli population was found above 300 CFU/1ml in all water sampling points. Table 3, the study revealed that the overall mean of E.coli and Non-E.coli were 2.09 and >3.48 log 10 Colony Forming Unit (CFU) 10 m L −1 respectively.

Microbial Contamination of Vegetable in Dry period
During the dry period, 35 vegetable samples from the 14 farmlands were included for analysis. During this time, farmers were entirely dependent on river water for irrigation since less rainwater is presented.
Ethiopian Kale is one of the dominant vegetables in the Ethiopian food menu regardless of the household incomes. Based on the study, the bacteriological quality of Ethiopian Kale is varied among the sampling points and had different microbial quality indicators ( Table 4). The total coliform count in Ethiopian Kale varies from 1 log 10 CFU g -1 in Mekanissa and Shito to approximately 5 log 10 CFU g -1 in Peacock 2. While the minimum fecal coliform pollution (10 CFU/100 ml) was found in the majority of sampling farms and the maximum, 2800 CFU/100ml was from Peacock2 farm. Contrarily, the total aerobic plate count population was relatively high in all sampling points with the range of 3.3 log 10 CFU g -1 in Bihere Tsige farm to 6.8 log 10 CFU g -1 in Akaki Tirunesh Beijing area (Table 4). Table 5 presents the bacteriological quality of lettuce irrigated with the Akaki River. The population of total coliform in Lettuce ranges from 3.27-log10 CFU g -1 in Mekanissa to 4.8 log 10 CFU g -1 in Burayu Kera.
The fecal coliform population is below or equal to 60 CFU/100 ml in all sampling points except in Koka and Bihere Tsige where the FC population was about 240 and 400 CFU/100 ml, respectively. While the total aerobic plate count population varies from 3.6 log 10 CFU g -1 in Mekanissa to 5.8 log 10 CFU g -1 in Peacock 1.
The microbial load of Swiss chard during the dry period shows variation based on location and type of microbial indicators (Table 6). The total coliform in Swiss chard ranged from 2 log 10 CFU g -1 in Burayu Kera to a maximum population of 3.8 log 10 CFU g -1 in Akaki Tirurnesh Beijing. Similarly, the minimum population of total aerobic plate count was found in Burayu Kera (3.9 log 10 CFU g -1 ) while the maximum population of total aerobic plate count was found in the Koka farm area (5.5 log 10 CFU g -1 ).

Microbial contamination of vegetables in Wet Season
During the rainy season, 18 vegetable samples were taken from the nine small farmlands. This number is lower than what we sampled during the dry season. In this season, most of the sampled vegetables were Ethiopian Kale and Lettuce. Table 7 presents microbial contamination of Ethiopian Kale during the rainy season. The population of Total coliform, Fecal coliform, and Total aerobic plate count ranges from 19 log 10 CFU g -1 , 0 CFU, and 3.9 log 10 CFU g -1 in Burayu Gefersa to 5.5 log 10 CFU g -1 , 1400 CFU/100 ml, and 5.9 log 10 CFU g -1 in Mekanissa, respectively.
During the rainy season, microbial contamination of Lettuce was very high compared to other vegetables.

Average microbial load by season
To measure the level of microbial contamination of fresh vegetables, which cultivated from the Akaki River, the authors calculated the mean value of each microbiological indicator and concerning the type of vegetables.
The mean count of TC, FC, and TAC on collected vegetables irrigated with Akaki River were 3.22, 1.37, and 4.72 in the dry season, and 3.87, 2.57, and 5.09-log10 CFU per gram in the wet season, respectively ( Table   9). All fresh Table 10 present the mean microbial count of fresh vegetables irrigated with Akaki River in wet and dry seasons. The mean Total coliform count, Fecal coliform, and total aerobic count of Ethiopian Kale were 3.22, 1.39, and 4.72 log 10 CFU g -1 in the dry season; and 3.49, 2.86, and 4.97 log 10 CFU g -1 in wet season respectively. Also, on Lettuce, the mean of Total coliform, Fecal coliform, and Total aerobic plate count was 3.15, 1.32, and 4.75 log 10 CFU g -1 in the dry season and 4.01, 1.02, and 5.49-log 10 CFU g -1 in the wet season. While the mean of Total coliform, Fecal coliform, and Total aerobic plate count of Swish chard were 3.17, 1.31, and 4.69 log10 CFU g -1 in the dry period. However, in the wet season only Total coliform (4.72 log 10 CFU g -1 ), and Total aerobic plate count (5.49 log 10 CFU g -1 ) were identi ed.

Discussion
The ndings of the study revealed that most of the water samples used for irrigation were highly contaminated and above the permissible level. This could be due to the absence of legal instruments that prevent the discharge of untreated wastewater from nearby factories and the residents of the city.
Furthermore, the mean value of E.coli and Non-E.coli of the sampled water was 2.09 and > 3.48 log10 CFU 10 mL − 1, which is higher than the WHO recommended standard (WHO, 2006). This could be due to the poor attention of government and other concerned bodies that they forget the health impact of wastewater on public health and the biotic environment. According to WHO standard, the Fecal coliforms level must not exceed 1000 counts 100 mL − 1 for the safe use of wastewater for irrigation of vegetables (WHO, 2006). Of those Fecal coliforms, E.coli is recommended as a good indicator to determine the hygiene condition at primary production of leafy vegetables and to validate and verify the application of good agriculture practice (Allende & Monaghan, 2015). However, in this study, most of the water samples were found to be above WHO standard for irrigation. Besides, this nding is also in agreement with previously conducted studies in the Akaki River Though the authors try to present the level of microbial contaminates among the vegetables, there is no speci cation placed for the permissible level of microbes for raw food being served in Ethiopia. However, Hazard Analysis and Critical Control Points-Total Quality Management (HACCPTQM) Technical Guidelines lay down the microbial quality for raw foods, where the food containing less than 4.4-6.69, 6.69-7.69 and greater than 7.69 log CFU g − 1 (aerobic plate count) is rated as good, average, poor and spoiled food, respectively (Aycicek, Oguz, & Karci, 2006). Based on these criteria, the ndings of the study indicated that most of the vegetables irrigated with Akaki River particularly in the dry season fall under poor categories. This may be due to a lack of dilution in the dry period and farmers are not use river water during the wet season.  , 1998). This may be due to the location of the river in the metropolitan city that owns the presence of many industries, commercial activities, and residential raw wastewater e uents that may be cause many communicable and non-communicable diseases to the residents.

Conclusion
The Akaki River was heavily contaminated with non-E. Coli & E. coli coliforms and did not meet the WHO guideline criteria for safe irrigation. In this study, target microorganisms commonly used as indicators for the hygiene status of foods frequently exceeded the HACCPTQM and ICMSF limit values for safe consumption. Freshly produced vegetables (especially Ethiopian Kale, lettuce, cabbage, and Swiss chard) had high contamination and they might contain pathogenic microorganisms and represent a risk for consumers regarding the foodborne disease. Therefore, this calls for farmer's and consumer's awareness of the dangers of contacting with Akaki River water and consuming pathogen-contaminated vegetables, and the need to insist on properly processed/stored sliced produce needs to be reawakened. Besides, to minimize potential risks associated with river water irrigation, the proper use of river water, as well as cheap and e cient methods to reduce microbial loads in microbiologically contaminated water used for irrigation, needs to be implemented.

Declarations
Ethics approval and consent to participate Not applicable.

Consent for publication
All authors have read the manuscript carefully and agreed to submit for publication.

Availability of data and material
The current study datasets are part of the Akaki River research project, which is not publicly available. The data can be accessed by ful lling the data sharing policy of the Ethiopian Public Health Institute.

Competing interests:
The authors declare that they have no competing interests.  Burayu Gefersa This site is located after the Geferessa dam and there is no known point source pollution. The area is dominated by residential and agricultural activity. In addition, there is a practice of animal grazing and community wash closes in the stream.

Lomimeda
Upstream land use includes vegetable farming, residential areas including condominiums with waste treatment ponds, and glass factory. There is a sign of a release of household and factory wastes.
Coca-cola area On this site, streams from the Winget side with known point sources such as marble factory, tannery industry jointly pass with the Gefersa stream. On the right sides of the river are houses and on the left are garages. Farming practices a few meters from the riverbanks.

Mekanissa-Teklehaymanot
Upstream land uses are residential houses, Alert hospital, Fistula hospital, Torhayloche hospital, and people bathing in the river. To the right, there are residential areas, animal farming unit and landslide, houses, and vegetable farming dominated the left side. Besides, during our visit to this site, we noticed that open defecation is hugely practiced on both sides of the river.
Mekanisa area Upstream land use includes an Alcohol factory, which is discharged to the Akaki River, to the left there is vegetable farming and residential houses to the right there is the practice of open defecation. Also, a stream from the Kera side joined this site with abattoir wastes.

Kera
Upstream land use includes the disposal of all the potential pollutants discharged from the upper stream such as the alcohol factory. The right side of the river is irrigated with vegetables and the left side of the river is the Addis Ababa abattoir. At downstream on the other side of the road just near to Japanese restaurant, there is also vegetable farming irrigating with the diverted river. Besides, the upper catchment is dominated by both residential and commercial activities.

Bihere Beheretsige
The upper catchment of this site is predominantly used as a residential area, and several automotive Garages have dominated the area.

Shitu behind Kality WWTP
"Shitu" is indicated the bad odour from the wastewater treatment plant located in the area. The upper catchment of the area is used for a residential area on top of the waste treatment plant. During our site visit, we noted that farmers directly irrigated their farmlands with over ow wastewater from lagoons and discharge wastewater after treatment.

Akaki Beseka
The upper stream is domesticated with animal raring units and the right side is adjoined by a condominium wastewater treatment station and to the left, there is farmland irrigated to produce vegetables.

Akaki Turneshi Beijing
The upper stream is domesticated with a hospital, beer factory road, and the right and left side is inhabited with vegetable farming by uses water pumpers. Besides, the second open market of Addis Ababa next to Merkato is found upstream of this site.
Peacock 1 upstream are situated with hotels, residential areas and the right side is the diversion of the river for irrigation and the left side is associated with construction material production and residential houses 100-200 meters away. Moreover, healthcare provider facilities are located above the farmland.

Peacock 2
Upstream are situated with hotels, residential areas on both sides of the sampling sites.
Aba Samuel This site is located the downstream of Akaki River. Intensive agricultural practice is carried out and animal grazing is common.           Map of study sites Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.