INTRODUCTIONMedicinal plants become contaminated by a varietyof sources such as heavy metals, insect larvae and seed,different bacteria, and fungi [1]. Heat and humidityof the environment, long-time drying, irrigation withcontaminated water, and lack of farmer training mayresult in considerable microbial contaminations andreduce the quality of plants. Moreover, microbialcontamination of plants may take place during unhealthycollection, cleaning, storage, transportation, andpackaging. Contact of herbal products with externalfactors such as plastic, glass, and other materials maylead to cross contamination.Medicinal plants can be contaminated by awide range of microorganisms, such as fungi,yeasts, protozoa, and viruses, most of which aretransferred from soil [2, 3]. Total microbial countis an important factor in determining the healthstatus or probable detection of a contaminationsource [4]. Yeast and mold are the most commoncontaminants of medicinal herbs. Various speciesof molds and yeasts that proliferate on food stuffsecrete metabolic toxic materials such as mycotoxins,which are harmful for humans and animals [5].The WHO (World Health Organization) has a largeamount of data in this direction [6].135Ameri A. et al. Foods and Raw Materials, 2020, vol. 8, no. 1, pp. 134–139Coliforms (Escherichia, Enterobacter, andKlebsiella) from the Enterobacteriaceae familyinhabit human and animal intestines. Most of them arenot pathogenic, although some E. coli strains couldbe highly pathogenic and cause food poisoning [7].According to Iran’s national standard, coliformcontamination in most dried vegetables should notexceed the maximum level of 1000 CFU/g, while thepresence of E. coli is not allowedI. Bacillus cereus iswidely distributed in the environment and some itsstrains are harmful for human health and can causefood poisoning. This bacterium secretes enterotoxin,hemolysin, and lecithinase C which are responsiblefor disease [9]. Dried vegetables contamination byBacillus cereus should not exceed the maximum level of100 CFU/g.Valeriana (Nardostachys jatamansi L.) from theValerianceae family is known for its anticonvulsant,sedative, anti-asthmatic, and cardiotonic properties [9].Fennel (Foeniculum vulgare L.) from theUmbelliferae family has culinary and medicinalproperties (anti-inflammatory, anti-spasmodic,diuretic, laxative, analgesic, antioxidant, and woundhealing)[10]. Licorice (Glycyrrhiza glabra L.)growing in Mediterranean countries, Central Asia,and Europe has a wide range of pharmacologicaleffects such as antioxidant activity, liver protection,and regulation of the immune system [11].Shirazi thyme (Zataria multiflora Boiss L.) is used in thesouth of Asia as tea or spice and in traditional medicineas a gastrointestinal disinfectant, diuretic, or an antiinflammatoryremedy [12].Contaminants such as microorganisms, heavymetals, and pesticides affect the quality and the efficacyof herbal products. Since it is impossible to remove allcontaminants, precautionary measures should be takento prevent or limit contamination [2, 3]. Therefore, ourstudy aimed to show the effect of these contaminationson consumer’s health.STUDY OBJECTS AND METHODSCollection and preparation of samples. For thisstudy, samples were randomly collected from medicinalherb retailers and drugstores of Ahvaz (Iran) fromDecember 2017 for 6 months. A total of 80 sampleswere used: 40 traditional (10 samples for each traditionalherb) and 40 industrial herbs from different companies(19 shirazi thyme, 6 fennel, 8 valeriana, and 7 licoricesamples). The amounts of industrial samples were notequal due to their insufficient availability.Total microbial count. Total microbial count wasperformed as described by Standard No. 5272, IranII.Different dilutions of medicinal herbs were prepared andI Standard No. 5939, Microbiology of dehydrated vegetablesspecifications.Institute of Standards and Industrial Research of Iran;2008.II Standard No. 5272, Microbiology of food and animal feedingstuffs-horizontal method for the enumeration of microorganismscolonycount technique at 30°C. Institute of Standards and IndustrialResearch of Iran; 2007.cultured on Plate count agar (PCA, Merck, Germany).Triplicate plates for each dilution were cultured andincubated for 72 h at 30°C. Then, the average of countedcolonies was measured taking into account the dilutioncoefficient.Mold and yeast count. Fungal count was performedaccording to Standard No. 10899, IranIII. Differentdilutions of medicinal herbs were inoculated onSabouraud dextrose agar (SDA, Merck, Germany) intriplicates and incubated at 25°C for 5 days. Then, theaverage number of molds and yeasts per gram of herbwas estimated.Bacillus cereus detection. To detect and countB. cereus (Standard No. 2324, Iran), dilutions ofmedicinal herbs were prepared and cultured in triplicateon Mannitol-egg yolk-polymyxin (MYP) agar (Merck,Germany) at 30°C for 48 h. The agar contained an eggyolk emulsion and polymyxin B sulfate (ShijiazhuangPharma, China)IV. Large and pink colonies (lackof manitol fermentation) with a sedimentary halo(lecithinase producer) were counted as probableB. cereus. To confirm the suspected colonies, ahemolysis test was performed on Blood agar (Merck,Germany).Coliform detection and enumeration. Coliformdetection and enumeration were performed accordingto Standard No. 9263, IranV. Different dilutions ofmedicinal herbs were inoculated (pour plate and twolayerculture) in triplicate on Crystal violet neutralred bile lactose (VRBL, Merck, Germany) agar andincubated at 37°C for 24 h. Typical red purple colonieswere confirmed on Brilliant green bile lactose (BGBL)broth (Merck, Germany) contained in Durham tubes attwo temperatures (37 and 44°C) for 24 h.Escherichia coli detection. Following coliformdetection, positive BGBL tubes (gas production) wereinoculated into peptone water and incubated at 44°C for48 h. Gas production in BGBL and production of indolein peptone water were recorded for presence of E. coliVI.Statistical analysis. Analysis of data was performedusing SPSS statistical software. The significance of theresults was evaluated by McNemar nonparametric testwith significance level of P < 0.05.RESULTS AND DISCUSSIONTotal microbial contamination. According to theresults of total microbial count (Fig. 1), 45% of the totalIII Standard No. 10899, Microbiology of food and animal feedingstuffs-horizontal method for the enumeration of yeasts and molds.Institute of Standards and Industrial Research of Iran; 2008.IV Standard No. 2324, Microbiology of food and animal feedingstuffs-horizontal method for the enumeration of presumptive Bacilluscereus-colony count technique at 30°C. Test method. Institute ofStandards and Industrial Research of Iran; 2006.V Standard No. 9263, Microbiology of food and animal feeding stuffshorizontalmethod for the enumeration of coliforms-colony counttechnique. Institute of Standards and Industrial Research of Iran; 2007.VI Standard No. 2946, Microbiology of food and animal feedingstuffs-detection and enumeration of presumptive Escherichiacoli. Most probable number technique. Institute of Standards andIndustrial Research of Iran; 2005.136Ameri A. et al. Foods and Raw Materials, 2020, vol. 8, no. 1, pp. 134–139herbs (36 samples out of 80) showed contaminationover the limit (105 CFU/g). Of them, 19 (48%) and17 (43%) were traditional and industrial, respectively.The microbial load in the samples with contaminationover the limit varied from 5.03 ± 0.03 to 6.25 ± 0.03 logCFU/g. This ranged from 5.14 ± 0.01 to 6.25 ± 0.03 inthe traditional samples and from 5.03 ± 0.3 to6 ± 0.05 log CFU/g in the industrial samples. However,there was no significant difference between thetotal microbial contamination in the traditional andindustrial samples (P > 0.05). Among the studied herbs,valeriana (Nardostachys jatamansi L.) and licorice(Glycyrrhiza glabra L.) showed the highest and thelowest contamination ‒ 56 and 18%, respectively. Also,the total microbial contamination of shirazi thyme(Zataria multiflora Boiss L.) and fennel (Foeniculumvulgare L.) was over the limit in 55% and 44% of thesamples, respectively.Fungal contamination. The results of fungalcontamination are presented in Fig. 2. As we cansee, 61 (77.5%) out of 80 samples had mold andyeast contamination over the limit (103 CFU/g).They comprised 34 (85%) traditional and 27 (67.5%)industrial samples. The fungal load in the samples withcontamination over the limit varied from 3.02 ± 0.00 to4.78 ± 0.06 log CFU/g. This reached from 3.02 ± 0.00 to4.78 ± 0.06 in the traditional samples and from 3.04 ±0.01 to 4.60 ± 0.01 log CFU/g in the industrial samples.It should be mentioned that the traditional sampleswere significantly contaminated with mold and yeast(P < 0.05). Meanwhile, valeriana (100%) and licorice(47%) showed the highest and the lowest contamination,respectively. Also, 79% and 75% of shirazi thymeand fennel, respectively, showed over the limit fungalcontamination.Bacillus cereus contamination. B. cereuscontamination in 44 samples (55%) of the total herbsexceeded the limit (103 CFU/g) (Fig. 3), including23 (57.5%) traditional and 21 (52.5%) industrial herbs.Over the limit B. cereus contamination varied from2.03 ± 0.03 to 3.84 ± 0.06 log CFU/g. In the traditionalsamples, it ranged from 2.03 ± 0.03 to 3 ± 0.06 logCFU/g and in the industrial samples, from 2.03 ± 0.03to 3.84 ± 0.06 log CFU/g. Our results showed that therewas no significant difference between the traditional andindustrial samples in B. cereus contamination (P > 0.05).The contamination in fennel (94%) and valeriana (66%)was significantly higher (P < 0.05). Shirazi thyme (41%)and licorice (29%) showed a lower level of B. cereuscontamination.Coliform contamination. Coliform contaminationwas found over the limit (103 CFU/g) in 44 samples(55%) of the total herbs. Among these samples,29 (72.5%) and 15 (37.5%) were from traditional andindustrial herbs, respectively (Fig. 4). Over the limitcoliform contamination ranged from 3.01 to 4.16 ±0.03 log CFU/g, namely from 3.01 to 4.16 ± 0.03 inthe traditional samples and from 3.03 ± 0.02 to 4.15 ±0.03 log CFU/g in the industrial samples. According tothe results, the traditional samples showed a significantlyhigher coliform contamination than the industrialsamples (P < 0.05). Fennel (81%) and valeriana (66%)Figure 1 Total microbial contamination in medicinal herbs.Total is number of samples. Positive ‒ contamination over thelimit (105 CFU/g). Negative ‒ contamination below the limit020406080Industerial&TraditionalTraditional IndusterialSamples numberSample typeTotal Posetive Negative020406080Industerial&TraditionalTraditional IndusterialSamples numberSample typeTotal Posetive Negative020406080Industerial TraditionalSamples numberTotal 020406080Industerial &TraditionalTraditional IndusterialSamples numberSample typeTotal Posetive NegativePositiveIndustrial IndustrialFigure 2 Mold and yeast contamination in medicinal herbs.Total is number of samples. Positive ‒ contamination over thelimit (103 CFU/g). Negative ‒ contamination below the limit.* P < 0.05IndusterialNegative020406080Industerial&TraditionalTraditional IndusterialSamples numberSample typeTotal Posetive Negative020406080Industerial &TraditionalTraditional IndusterialSamples numberSample typeTotal Posetive Negative020406080Industerial TraditionalSamples numberTotal IndusterialPositiveIndustrial IndustrialFigure 3 Bacillus cereus contamination in medicinal herbs.Positive ‒ contamination over the limit (102 CFU/g). Negative‒ contamination below the limit. * P < 0.05IndusterialNegative020406080Industerial&TraditionalTraditional IndusterialSamples numberSample typeTotal Posetive Negative020406080Industerial &TraditionalTraditional IndusterialSamples numberSample typeTotal Posetive Negative020406080Industerial TraditionalSamples numberTotal IndusterialPositiveIndustrial Industrial137Ameri A. et al. Foods and Raw Materials, 2020, vol. 8, no. 1, pp. 134–139revealed a very high level of contamination (P < 0.01),while the lowest level was recorded in licorice (41%) andshirazi thyme (41%).Escherichia coli contamination. There shouldbe no E. coli contamination in dried vegetables [8].According to our results, 25 samples (31.2%) of thetotal herbs showed E. coli contamination. Of them,19 (47.5%) and 6 (15%) were from traditional andindustrial herbs, respectively (Fig. 5). E. coli contaminationin the traditional samples was significantlyhigher than in the industrial samples (P < 0.05). Thecontamination in fennel (43.7%) and valeriana (33.3%)was significantly higher (P < 0.05) than that in licorice(23.5%) and shirazi thyme (27.5%).One of the important aspects is contamination ofmedicinal herbs by different types of harmful factorssuch as microbes, heavy metals, as well as radioactiveand chemical materials [2]. Our results, in many cases,indicated high contamination of the herbs under studywith different microbial agents. These contaminationscould occur during different stages of cultivation,extraction, drying, packing or distribution [2, 3, 13].In our study, the microbial contamination level in thetraditional samples was higher than that in the industrialsamples. This result could be due to different productionand packaging conditions.Researchers have reported that the differencesin technological level and preparation, supply andproduction of medicinal herbs could affect theircontamination level [13–15]. Worldwide, a high levelof contamination has been reported in a variety ofmedicinal herbs. For example, Banerjee et al., in astudy of 154 dried plants collected from shops in India,showed that the total microbial count was over thelimit in 51% of the samples, and 97% of them had moldcontamination [16].Moreover, Abba et al., in a study of powderedmedicinal plants in Nigeria, reported that 87% ofthem had high microbial contamination [17]. Theircontamination level was significantly higher than that inour study (45%), which could be due to environmentalfactors, soil or inappropriate packing conditions. Somestudies in different locations showed that many of theinvestigated medicinal herbs were contaminated withvarious fungi [5, 16, 18, 19]. Alwakeel, in a study on 32samples of various medicinal plants in Saudi Arabia,showed Bacillus cereus as the most common microbialcontaminant [20]. Martins et al. found the same resultin more than 90% of the studied medicinal plants inPortugal [21].In our study, fennel and valeriana showed the highestand licorice showed the lowest levels of contaminationwith B. cereus and coliform. In a study of theantimicrobial activity of Turkish spices, fennel showeda lower antibacterial effect on B. cereus [22]. Moreover,Lang et al. in Austria reported that fennel had a lowerinhibitory effect on coliform than licorice [23]. It seemsthat the antimicrobial properties of medicinal herbscould also explain the differences in their contaminationlevels. In the previous studies, high microbialcontaminations were reported in valeriana [19, 24]. Wefound that valeriana had the highest contamination level.This result can be due to the fact that most of its activemedicinal ingredients are in the root of the plant, whichis in direct contact with soil, so more microorganismscan be transmitted to it [19]. In our study, licoriceshowed a considerably lower contamination level thanother plants. It could be due to differences in plantproduction, its active constituents, and distributionprocesses [3, 25]. In addition, most studies have shownthat licorice had higher antimicrobial and antifungalactivities than other plants, especially against B. cereusand E. coli [22, 35–38].CONCLUSIONIn our study, we tested 80 samples of traditionaland industrial herbs, such as valeriana, fennel, licorice,and shirazi thyme, for microbial contamination. TheFigure 4 Coliform contamination in medicinal herbs.Positive ‒ contamination over the limit (103 CFU/g).Negative ‒ contamination below the limit. * P < 0.05IndusterialNegative020406080Industerial &TraditionalTraditional IndusterialSamples numberSample typeTotal Posetive NegativePositiveIndustrial IndustrialFigure 5 Escherichia coli contamination in herbs. Total isnumber of samples. Positive ‒ presence of E. coli.Negative ‒ absence of E. coli. * P < 0.05020Industerial&TraditionalTraditional IndusterialSamples Sample typeTotal Posetive Negative020Industerial&TraditionalSamples Total 020406080Industerial &TraditionalTraditional IndusterialSamples numberSample typeTotal PPoosseittiivvee NegativeIndustrial Industrial138Ameri A. et al. Foods and Raw Materials, 2020, vol. 8, no. 1, pp. 134–139results of the experiment showed that microbial indicesin considerable percentage of the samples exceededthe standard limit. Valeriana had the highest totalmicrobial count and yeast/mold contamination, fennel ‒B. cereus and coliforms, while licorice was notmassively contaminated.We also revealed that fungal, coliform, and E. colicontamination in traditional herbs was considerablyhigher than that in industrial samples. However, therewas no significant difference between them in totalmicrobial count and B. cereus contamination.Thus, our results demonstrated the importance ofmonitoring medicinal plants contamination to controlthe quality of herbal products.CONTRIBUTIONMaryam Ekhtelat developed the original idea and theprotocol. Abdolghani Ameri, Maryam Ekhtelat and SaraShamsaei developed the protocol, analyzed the data, andwrote the manuscript.CONFLICT OF INTERESTThe authors declare no conflict of interest.ACKNOWLEDGMENTThis study was financially supported by AhvazJundishapur University of Medical Sciences andextracted from the thesis of Ms Sara Shamsaei (B-9605).