lable at ScienceDirect

Food Microbiology 26 (2009) 847–852

Contents lists avai

Food Microbiology

journal homepage: www.elsevier .com/locate/ fm

Assessment of the microbiological safety of edible dried seeds from retailpremises in the United Kingdom with a focus on Salmonella spp.

Caroline Willis a,*, Christine L. Little b, Satnam Sagoo b, Elizabeth de Pinna b, John Threlfall b

a Health Protection Agency, Food Water and Environmental Microbiology Network (Southampton Laboratory), Level B South Block, Southampton General Hospital, SO16 6YD, UKb Gastrointestinal, Emerging and Zoonotic Infections, Centre for Infections, Health Protection Agency, 61 Colindale Avenue, London NW9 5EQ, UK

a r t i c l e i n f o

Article history:Received 16 March 2009Received in revised form20 May 2009Accepted 21 May 2009Available online 27 May 2009

Keywords:SalmonellaEscherichia coliEdible seedsSesame seedsSunflower seedsPumpkin seedsMelon seedsAlfalfa seedsLinseed

* Corresponding author. Tel.: þ44 2380 777142; fax:Tel.:þ44 208 200 4400 (for all other authors).

E-mail address: caroline.willis@hpa.org.uk (C. Wil

0740-0020/$ – see front matter Crown Copyright � 2doi:10.1016/j.fm.2009.05.007

a b s t r a c t

Sesame seed products have recently been associated with a number of Salmonella outbreaks in the UKand elsewhere. Aside from sesame seeds, there is little published information on the prevalence ofSalmonella spp. in edible seeds. A study of 3735 samples of retail edible dried seeds in the UK wastherefore carried out between October 2007 and March 2008 to assess their microbiological safety inrelation to Salmonella contamination and levels of Escherichia coli, an indicator of faecal contamination.Overall, Salmonella was detected in 23 samples (0.6%), of which over half (57%) were sesame seeds. Otherseeds contaminated with Salmonella were linseed (1 sample), sunflower (1 sample), alfalfa (1 sample),melon (4 samples) and mixed seeds (3 samples). E. coli was detected in 9% of samples, with 1.5% con-taining unsatisfactory levels (�102/g). These included melon, pumpkin, sesame, hemp, poppy, linseed,sunflower and mixed seeds. The UK retailers affected by the detection of Salmonella in their productsrecalled the contaminated batches, and Food Standards Agency food alerts were issued to advise againstthe consumption of affected seed products. This study highlights the importance of good hygienepractices and effective decontamination procedures during the production of these products.

Crown Copyright � 2009 Published by Elsevier Ltd. All rights reserved.

1. Introduction

Dried seeds are popular as a ready-to-eat food, particularly dueto reports of their health benefits (Cooney et al., 2001; Ollis et al.,1999; Yu et al., 2005). However, in recent years there have beena number of reports of Salmonella incidents related to edible seedsand their products. For example, in 2001, an international Salmo-nella Typhimurium outbreak in Sweden, Norway, Germany andAustralia was linked with the consumption of helva, a sesame seed-based confectionery (de Jong et al., 2001). In 2002 and 2003, threeoutbreaks of Salmonella Montevideo infection were identified inAustralia and New Zealand (Unicomb et al., 2005), with tahini(sesame seed paste) imported from Egypt and Lebanon beingidentified as the source of infection. In Germany, Salmonella Agonafrom an untreated batch of aniseed imported from Turkey wasfound to be the cause of 36 cases of infantile gastroenteritis (Kochet al., 2005; Rabsch et al., 2005). Sprouted seeds, and alfalfa sproutsin particular, have also been associated with a number of outbreaks

þ44 2380 777143 (C. Willis).

lis).

009 Published by Elsevier Ltd. All

of Salmonella (Proctor et al., 2001; Van Beneden et al., 1999; Win-throp et al., 2003), including two in the EU in 2007 (Emberlandet al., 2007; Werner et al., 2007).

Different seed types require varying climates and environmentsfor growth and are therefore grown in many different areas of theworld. Linseed, poppy and hemp crops are grown in large fields,and harvested, cleaned and sorted mechanically in a similar way tocereal crops. Similarly, sunflower seeds are harvested from largefields and hulled mechanically. However, sunflowers are harvestedat such a height from the ground that they tend to be cleaner thanother seed crops (Pierce, 1970). The moisture content of sunflowerseeds following harvest may be as high as 20% (Salunkhe et al.,1992). These are dried to less than 10% moisture, usually by theapplication of forced air to the batch of seeds before the hulls areremoved mechanically (Salunkhe et al., 1992). In contrast, pumpkinand melon seeds are removed from the fruit in a relatively moistcondition (approx. 38%) (Pierce, 1970). The drying process for thesecrops may involve spreading the seeds on the ground to dry in thesun, or alternatively, drying in an oven or with the use of smoke(Bankole et al., 2005) before they are hulled with the aid of waterand then re-dried. Sesame seeds are harvested and then subjectedto a cleaning process to remove debris before being hulled bymeans of water (aqua-hulling), mechanical friction or a chemical

rights reserved.

C. Willis et al. / Food Microbiology 26 (2009) 847–852848

process (using a caustic soda solution) (Brockmann et al., 2004;Wittenberg, 2007). Finally, they are dried, often mechanically, ina stream of hot air.

During growth, the crops are likely to be exposed to a widerange of microbial contamination from many sources includingsoil, manure, irrigation water, wild birds and animals (Doyle andErickson, 2008). Further potential for microbial contaminationmay occur during post-harvest processing. Whilst the dryingprocess for some seed products involves a heat treatment, this isnot always the case, and where heat is used it is not necessarily ata sufficient temperature to ensure that all pathogenic bacteria arekilled.

There is little published data on the microbiological quality ofedible dried seeds. Some surveys of sesame seed products havebeen carried out in the context of international outbreaks ofSalmonella associated with the consumption of these products in2001 and 2003. For example, Brockmann et al. (2004) examined117 samples of sesame seed products on retail sale in Germany in2001. Of these, 11 (9.4%) were found to be contaminated withSalmonella species. Two investigations of sesame seed productsfrom retail premises in England and Wales were carried out in2001 and 2003. In 2001, Salmonella Typhimurium was detected in6 of 151 halva samples. These represented two brands producedby a single manufacturer in Turkey (C Little, HPA, unpublishedobservations). In 2003, 160 sesame products were examined, ofwhich Salmonella was detected in 25 tahini samples from Lebanonand Cyprus (S Surman-Lee, HPA personal communication). InAustralia, the National Enteric Pathogen Surveillance Schemereported the isolation of 17 Salmonella serovars from sesameseeds and their products over 30 evaluations between 1985 and2001 (O’Grady et al., 2001). In 2006, a pan-London study of edibleseeds showed that 2.0% (7 of 367 analysed samples) werecontaminated with Salmonella spp. (S Surman-Lee, HPA personalcommunication). In addition, a number of studies have investi-gated the microbiological quality of edible nuts, which are similarproducts to seeds in terms of physical composition and produc-tion processes. For example, a study of raw almonds arriving atalmond processors between 2001 and 2005 indicated a Salmo-nella prevalence of 0.87% (81 of 9274 samples) (Danyluk et al.,2007). Moreover, in a survey of 921 samples of peanut, cashew,almond, Brazil nut and hazelnut kernels, Salmonella was detectedin one raw almond sample (overall prevalence of 0.1%) (Eglezoset al., 2008). An international outbreak of Salmonella in 2001 waslinked with the consumption of imported peanuts (Kirk et al.,2004). The findings of these studies on seeds and nuts indicatethat edible seeds could be a potential source of foodborne illnessin the UK.

In response to these investigations, the Local Authorities Co-ordinators of Regulatory Services (LACORS) and the HealthProtection Agency (HPA) Co-ordinated Food Liaison Group pro-gramme undertook a microbiological study with the aim ofassessing the microbiological safety of edible dried seeds on retailsale in the UK, with a particular focus on the detection of Salmonellaspecies. A range of edible dried seeds were sampled and examinedover a six month period to provide further data on the microbiologyof these products and to highlight potential problems with theirproduction and use.

2. Methods

2.1. Sample collection

A total of 3735 samples of edible seeds were collected from 3390retail premises by sampling officers from 317 Local Authority Envi-ronmental Health Departments (EHD) across the UK between 1

October 2007 and 31 March 2008. Sampling officers were requestedto include at least one sample of sesame seeds on every occasion thatthey collected samples. This level of scrutiny was deemed necessarybecause of the previous reports of Salmonella contamination ofsesame seed products. Seeds coated with chocolate, yoghurt or othercoatings, flavoured with seasonings (spices, salt, etc.), or thosecontaining dried fruits were specifically excluded from the study.Registered retail premises lists held by EHDs were used to derive anapproach to sampling. Retail premises were selected at random fromEHDs’ database of food businesses via a random number generatoror every 10th entry, and if suitable, samples were collected. Samples(of at least 50 g) were collected and transported in accordance withthe Food Standards Agency Food Law Code of Practice (Food Stan-dards Agency, 2006) and the Local Authorities Co-ordinators ofRegulatory Services (LACORS) guidance on microbiological foodsampling (Local Authorities Co-ordinators of Regulatory Services,2006). These were examined by 32 Official Control Laboratories.Information on samples was obtained by observation and enquiryand recorded on a standard questionnaire.

2.2. Sample examination

Escherichia coli was enumerated and the presence of Salmonellasought in accordance with HPA Standard Methods F22 and F13respectively (Health Protection Agency, 2005; Health ProtectionAgency, 2008). Where Salmonella was detected, and sufficientsample was available, Salmonella was enumerated by a MostProbable Number (MPN) 10-tube method. This involved preparinga 1 in 10 dilution of the sample by adding 900 ml of BufferedPeptone Water to 100 g of sample. Ten aliquots, each of 100 g, of thisdilution were then dispensed into separate sterile containers. Thesewere incubated and sub-cultured as described in HPA StandardMethod F13 (Health Protection Agency, 2008). The number ofaliquots from which Salmonella was detected was compared to anMPN 10-tube table (adapted from ISO/FDIS 7218:2007 (Interna-tional Organisation for Standardisation, 2007)). All isolates ofSalmonella were sent to the Laboratory of Gastrointestinal Patho-gens (LGP), HPA Centre for Infections, for further characterisation.This included sero-typing (Bale et al., 2007; Popoff and Le Minor,2001), phage typing (Chambers et al., 1987) and antimicrobialsensitivity testing (Frost, 1994).

Microbiological results were compared to the HPA (PHLS)Guidelines for the microbiological quality of some ready-to-eatfoods sampled at the point of sale (PHLS, 2000). According to theseguidelines, the absence of Salmonella in 25 g is satisfactory, whilstthe detection of Salmonella in 25 g is unacceptable, and is consid-ered potentially injurious to health and/or unfit for humanconsumption (i.e. contravenes Article 14 Food Safety Requirementsof Regulation (EC) No.178/2002 (the General Food Law Regulation)).An E. coli level of <20/g is considered satisfactory, between 20 and<100/g is acceptable and �100/g is unsatisfactory.

2.3. Statistical analysis

Descriptive and statistical analysis of the data was undertakenusing Microsoft Excel. Relative proportions were compared usingthe Chi squared test (c2) and Fisher’s Exact Test. A probability valueof less than 5% was defined as significant.

3. Results

3.1. Salmonella contamination of edible seeds

The types of seeds sampled are shown in Table 1. Overall,Salmonella species were detected in 23 of 3735 samples (0.6%). This

Table 1Edible dried seed types sampled in relation to presence of Salmonella and unsatis-factory levels of E. coli (i.e. �102/g).

Seed type No. ofsamplesN ¼ 3735 (%)

No. withSalmonelladetected (%)

No. withE. coli�102/g (%)

Single seed types 3385 (90.6) 20 (0.6) 47 (1.4)Alfalfa 58 (1.6) 1 (1.7) –Hemp 121 (3.2) – 2 (1.7)Linseed (flax) 284 (7.6) 1 (0.4) 1 (0.4)Melon (egusi) 47 (1.3) 4 (8.5) 6 (12.8)Poppy 202 (5.4) – 3 (1.5)Pumpkin 886 (23.7) – 23 (2.6)Sesame 771 (20.6) 13 (1.7) 8 (1.0)Sunflower 976 (26.1) 1 (0.1) 4 (0.4)Other (water melon, celery) 40 (1.1) – –

Mixed seed types 350 (9.4) 3 (0.9) 8 (2.3)2 seed types 57 (1.5) – 1 (1.8)3 seed types 80 (2.1) 1 (1.3) 4 (5.0)4 seed types 139 (3.7) – 3 (2.2)5 seed types 71 (1.9) 2 (2.8) –>5 seed types 3 (0.1) – –

Total 3735 23 (0.6) 55 (1.5)

C. Willis et al. / Food Microbiology 26 (2009) 847–852 849

included 20 samples of a single seed type (alfalfa, linseed, melon (or‘‘egusi’’), sesame and sunflower) and three samples of mixed seeds;two containing five seed types (pumpkin, sunflower, sesame,linseed and hemp) and one containing three types (pumpkin,sunflower and sesame). Over half (57%; 13/23) of samples con-taining Salmonella were sesame seeds. Moreover, mixed seeds thatcontained sesame seeds were contaminated with Salmonella (1.3%;3/228), whereas those without sesame seeds were not (0%; 0/122).However, the frequency of contamination of sesame seeds withSalmonella (1.7%; 13/771) was equal to the contamination rate ofalfalfa seeds (1.7%; 1/58) and significantly lower than that of melonseeds (8.5%; 4/47; p ¼ 0.013; Fisher’s Exact test). Salmonella wasenumerated in six samples. In four of these, the Most ProbableNumber was <0.1/g, whilst counts in the other two samples were0.1 and 0.2/g respectively.

Table 2Salmonella subtypes isolated from edible dried seed samples.

Salmonella sero/phage type (PT) Number ofsamples

AntibioticResistance Profile

S. Agona 3PT 3 1 SensitivePT UTa 1 SensitivePT RDNCb 1 Sensitive

S. Bergen 1 SensitiveS. Binza 1 SensitiveS. Chittagong 1 SensitiveS. Drypool 5 SensitiveS. Montevideo 1 SensitiveS. Newport 1 SensitiveS. Schwarzengrund 1 SensitiveS. Sculcoates 1 Ac

S. Senftenberg 1 SensitiveS. Tennessee 1 SensitiveS. Unnamed (I 47: z4,z23:-) 3 SensitiveS. Unnamed (I 3,10: y:-) 1 SensitiveS. Virchow 2

PT 31 1 SensitivePT 51 1 Sensitive

a UT, Untypeable.b RDNC: Reacts with typing phages, but does not conform to a known type.c A: Ampicillin resistant.

3.2. Salmonella subtypes present in edible seeds

Seventeen Salmonella subtypes were isolated from 23 samplesin this study (Table 2). Salmonella Drypool was isolated from fiveseparate samples, of which four were sesame seeds from India andone was a mixture of pumpkin, sunflower and sesame seeds(country of origin unknown). Salmonella Unnamed (47: z4,z23:-)was detected in three samples, of which two were sesame seedsfrom India and one was egusi seed (country of origin unknown).Although S. Agona and S. Virchow were also isolated from multiplesamples, no individual phage type was isolated from more than onesample. Twenty-two of the 23 Salmonella isolates were sensitive toall antibiotics tested (Table 2). However, one (S. Sculcoates)demonstrated resistance to ampicillin.

3.3. E. coli levels in edible seeds

E. coli was present at a level of �3/g in 339 samples (9%). Countsof �102/g (range 1.0 � 102–>1.0 � 103/g) were detected in 55samples (1.5%), as shown in Table 1. Of these, 23 samples (42%) werepumpkin seed products. Only two samples in which Salmonellaspecies were detected also had an E. coli count of �102/g (240/g,1100/g), whilst in 17 samples contaminated with Salmonella (74%),E. coli was not detected (<3/g).

3.4. Relationship between country of origin and microbiologicalquality of seeds

The samples collected represent at least 42 countries of origin(Table 3), with 281 samples containing seeds of more than onecountry. One fifth (20%) of all samples were produced in China, and7% in India. Salmonella species were detected in seeds produced inArgentina, Burkina Faso, Guatemala, India, Nigeria, USA and WestAfrica (Table 3). Salmonella spp. were also detected in three sampleswhich were labelled as produce of more than one country and fivefor which the country of origin was unknown, but were notdetected in any samples originating from individual countrieswithin the European Union.

E. coli levels of �102/g were detected in seeds produced inArgentina, Austria, China, France, India, Netherlands, UK, USA andWest Africa. Of the seed samples labelled as produce of more thanone country, 3.2% contained E. coli at �102/g, as did 1.3% of samplesof unknown country of origin.

3.5. Relationship between packaging and microbiological qualityof seeds

The majority (97.2%; 3632/3735) of samples collected werepre-packed, with 2.8% (103) of samples being taken from opendisplay containers. All samples contaminated with Salmonella andE. coli at �102/g were pre-packed. Salmonella spp. were detectedin 12 different brands of seed products, with the maximumnumber of contaminated samples of a single brand being five.However, eight samples of sesame seeds contaminated withSalmonella comprised two different brands, both produced inIndia. No duplication of batch numbers was observed betweencontaminated samples.

3.6. Association between organic status and microbiologicalquality of seeds

Almost one quarter of seed samples (24%; 893/3735) werelabelled as organic. There was no significant difference in theproportion of seeds labelled as organic (0.2%; 2/893) or not (0.7%;21/2842) that were found to be contaminated with Salmonella.

Table 3Country of origin of edible dried seeds in relation to the presence of Salmonella andelevated levels of E. coli (i.e. �102/g).

Country No. of samplesN ¼ 3735 (%)

No. withSalmonelladetected (%)

No. withE. coli�102/g (%)

EC Countries 490 (13.3) 0 5 (1.0)Austria 17 (0.5) – 1 (5.9)Czech Republic 3 (0.1) – –France 136 (3.6) – 1 (0.7)Germany 38 (1.0) – –Greece 6 (0.2) – –Hungary 3 (0.1) – –Italy 16 (0.4) – –Netherlands 79 (2.1) – 1 (1.3)Spain 5 (0.1) – –UK 180 (4.8) – 2 (1.1)Other EC countries 7 (0.2) – –

Non-EC Countries 1472 (39.4) 15 (1.0) 22 (1.5)Africa 25 (0.7) – –Argentina 13 (0.3) 1 (7.7) 1 (7.7)Australia 16 (0.4) – –Burkina Faso 18 (0.5) 1 (5.6) –Canada 8 (0.2) – –China 750 (20.0) – 8 (1.1)Egypt 12 (0.3) – –Guatemala 22 (0.6) 1 (4.5) –India 249 (6.7) 8 (3.2) 7 (2.8)Nicaragua 25 (0.7) – –Nigeria 34 (0.9) 1 (2.9) –Paraguay 28 (0.7) – –South America 15 (0.4) – –Thailand 8 (0.2) – –Turkey 26 (0.7) – –Uruguay 8 (0.2) – –USA 182 (4.9) 1 (0.5) 2 (1.1)Venezuela 10 (0.3) – –West Africa 7 (0.2) 2 (28.6) 4 (57.1)Other non-EC countries 16 (0.4) – –More than one country 281 (7.5) 3 (1.1) 9 (3.2)Not known 1492 (39.9) 5 (0.3) 19 (1.3)

Total 3735 23 (0.6) 55 (1.5)

C. Willis et al. / Food Microbiology 26 (2009) 847–852850

However, a significantly higher proportion of organically producedseed samples had unsatisfactory levels of E. coli (�102/g) (2.4%; 21/893) compared to those that were not (1.2%; 34/2842) (p ¼ 0.017).

4. Discussion

Products of plant origin are increasingly being associated withoutbreaks of infection of both Salmonella and other pathogens(Little and Gillespie, 2008; Sivapalasingam et al., 2004). There aremany opportunities for microbial contamination of these productspre-, during and post-harvest, and in many cases, the products areconsumed raw or following minimal processing. In the case ofedible dried seeds, these are frequently eaten as snack foods orincorporated into meals without further cooking. Results fromthis study have demonstrated that, whilst the vast majority ofedible dried seeds sampled at retail were of satisfactory oracceptable microbiological quality (98.0%; 3659/3735), a smallproportion of samples (0.6%) were contaminated with Salmonellaspp., which is unacceptable. Ready-to-eat foods contaminatedwith Salmonella spp. are unsafe. They are considered to be inju-rious to health and unfit for human consumption and theytherefore contravene the food safety requirements (Article 14) ofRegulation (EC) No.178/2002 (European Commission, 2002).Where Salmonella was detected from samples examined as partof this study, the UK retailers affected publicly recalled thecontaminated batches and full investigations were undertaken.

A number of Food Standard Agency food alerts were also issued,advising consumers not to eat the affected batches (FoodStandards Agency, 2009).

The recent international outbreaks of salmonellosis havedemonstrated that major health problems can arise fromconsumption of contaminated ready-to-eat sesame seed products ifhygiene practices break down (de Jong et al., 2001; Unicomb et al.,2005). Although there is no direct evidence that the contaminatededible seeds detected in this survey were responsible for any casesof human illness, six of the identified subtypes were reported incases of human infection in England and Wales during the studyperiod (S. Montevideo (27 cases), S. Newport (62), S. Schwarzen-grund (24), S. Senftenberg (14), S. Tennessee (8) and S. VirchowPT31 (2)).

The prevalence of Salmonella spp. in sesame seeds in the UK in2007/8 (1.7%) was significantly lower than that previously foundin Germany in 2001 (Brockmann et al., 2004) (p < 0.0001) and inEngland in 2006 (2.5%) (S Surman-Lee, HPA, personal communi-cation) (p ¼ 0.0144). However, the overall prevalence of Salmonellaspp. in edible seeds in this survey (0.6%) is not significantly differentto that in surveys of nuts in Australia (0.1%) (Kirk et al., 2004) andthe United States (0.87%) (Danyluk et al., 2007).

Where Salmonella was enumerated from seeds in this study,counts were low (all <1/g). However, small numbers (ranging from<0.03 to 0.46 organisms/gram) of salmonellae in a sesame-seedproduct (tahini) have been reported to cause a large outbreak ofsalmonellosis in Australia and New Zealand (Unicomb et al., 2005).This indicates that even low doses of Salmonella in these types ofproduct may lead to infection.

The most frequently contaminated seed type was melon seeds,also known as egusi, of which 8.5% contained Salmonella and12.8% contained unsatisfactorily high levels of E. coli (i.e. �102/g).In contrast, 1.7% of both sesame and alfalfa seeds were contami-nated with Salmonella, whilst 1% of sesame seeds and none of thealfalfa seeds examined contained unsatisfactory E. coli levels.Although labelling on some of the packs of melon seeds includedinstructions to cook thoroughly, this seed type was included inthe study as the cooking instructions did not include specific timeand temperature directions, and it was considered that the seedsmight be eaten with no or inadequate heat treatment. It isinteresting to note that in 2007 the EC was notified of 17 batchesof melon seeds (all from Nigeria) contaminated with aflatoxins(European Commission, 2008), which are a further indicator ofpoor control of microbial growth during production and storageof these seeds.

E. coli was included in this study as it is traditionally consideredto be a useful marker of faecal contamination of food products, andtherefore indicates the potential risk of contamination with faecalpathogens (Roberts and Greenwood, 2003). However, resultsobtained in this study indicate that there was no significant asso-ciation between an elevated E. coli count and the detection ofSalmonella in seeds. This finding is supported by a study carried outin the United States in 1977 (Andrews et al., 1979), which involvedthe microbiological analysis of 1960 samples of ‘‘health foods’’including seeds. Salmonella was detected in four packs of sunflowerseeds of a single brand and a single sample of alfalfa seeds. In thesunflower seed samples, faecal coliforms were not detected, and inthe alfalfa seeds, the faecal coliform count (Most Probable Number)was 0.4/g. Moreover, a recent LACORS/HPA study of fresh herbs alsofound that E. coli was not a reliable indicator for Salmonella pres-ence (Elviss et al., 2009).

The contaminated seed samples in this study were from severaldifferent producers, indicating that the results were not simply dueto hygiene lapses by a single producer, but rather that contamina-tion problems may be relatively widespread throughout the

C. Willis et al. / Food Microbiology 26 (2009) 847–852 851

industry and in many different countries. These results highlightthe importance of good hygiene practices at all stages of productionin order to avoid contamination of the final product with patho-genic bacteria. However, there is little published guidance on goodhygiene practices for use by the seed-producing industry. Criticalpoints for controlling the microbiological quality of edible seedsappear to be the cleaning and drying processes. The method andspeed of drying seeds are variable, with seeds such as pumpkin andmelon frequently being laid out on the ground to dry in the sun, andsesame and sunflower seeds being dried by means of a hot airstream whilst being agitated mechanically (Bankole et al., 2005;Salunkhe et al., 1992). Where appropriate, a heat treatment as partof the drying process is desirable and, in some instances, is insistedon by purchasers. In one study, optimum drying of pumpkin seedswas found to be achieved by an airflow speed of 0.8 m/s and an airtemperature of 60 �C (Sito et al.,1999). Temperatures of 80 or 100 �Cimpaired the quality of the seeds. However, if a temperature of60 �C were to be used, a holding time of several minutes would berequired to ensure a significant reduction in numbers of pathogenicorganisms. For example, at 65 �C, it has been demonstrated thatnumbers of Salmonella can be reduced by 1 log/min, and thereforea 10 min heat treatment would ensure a 10 log reduction in counts(Adams and Moss, 1995).

This study has highlighted the potential public health riskassociated with the consumption of edible dried seeds, due tocontamination with Salmonella spp., and has drawn attention to therequirement for clearer guidance on good hygiene practicesthroughout all stages of seed production, pre-, during and post-harvest. Such guidance would aid purchasers and suppliers of seedproducts in ensuring that appropriate practices are put in place byproducers worldwide.

Acknowledgements

The authors would like to thank the staff in EnvironmentalHealth Departments throughout the UK and staff in the HPAlaboratories and other Official Control Laboratories for theircontributions to this study. Thanks are extended to LGP, HPA Centrefor Infections for typing Salmonella isolates, to Gemma Cantelo atLACORS for co-ordinating the participation of EnvironmentalHealth Practitioners and advice from the LACORS Food Examinationand Food Hygiene Focus Groups, to the HPA Regional Food, Waterand Environmental Co-coordinators Forum for their contributionand advice on the sampling protocols of this study, and to LillianHucklesby for data collation and validation.

References

Adams, M., Moss, M., 1995. Food Microbiology. Royal Society of Chemistry,Cambridge.

Andrews, W., Wilson, C., Poelma, P., Romero, A., Mislivec, P., 1979. Bacteriologicalsurvey of sixty health foods. Appl. Environ. Microbiol. 37, 559–566.

Bale, J., de Pinna, E., Threlfall, E.J., Ward, L., 2007. Kauffmann–White Scheme – 2007:Salmonella Identification – Serotypes and Antigenic Formulae. Health Protec-tion Agency, London.

Bankole, S., Osho, A., Joda, A., Enikuomehin, O., 2005. Effect of drying method on thequality and storability of ’egusi’ melon seeds (Colocynthis citrullus L.). AfricanJ. Biotechnol. 4, 799–803.

Brockmann, S., Piechotowski, I., Kimmig, P., 2004. Salmonella in sesame seedproducts. J. Food Prot. 67, 178–180.

Chambers, R., McAdam, P., de Sa, J., Ward, L., Rowe, B., 1987. A phage-typing schemefor Salmonella virchow. FEMS Microbiol. Lett. 40, 155–157.

Cooney, R., Custer, L., Okinaka, L., Franke, A., 2001. Effects of dietary sesame seeds onplasma tocopherol levels. Nutr. Cancer 39, 66–71.

Danyluk, M., Jones, T., Abd, S., Schlitt-Dittrich, F., Jacobs, M., Harris, L., 2007. Prev-alence and amounts of Salmonella found on raw California almonds. J. Food Prot.70, 820–827.

Doyle, M., Erickson, M., 2008. Summer meeting 2007 – the problems with freshproduce: an overview. J. Appl. Microbiol. 105, 317–330.

Eglezos, S., Huang, B., Stuttard, E., 2008. A survey of the bacteriological quality ofpreroasted peanut, almond, cashew, hazelnut, and Brazil nit kernels receivedinto three Australian nut-processing facilities over a period of 3 years. J. FoodProt. 71, 402–404.

Elviss, N., Little, C.L., Hucklesby, L., Sagoo, S., Surman-Lee, S., de Pinna, E.,Threlfall, E.J., 2009. Microbiological study of fresh herbs from retail premisesuncovers an international outbreak of salmonellosis. Int. J. Food MicrobiolAvailable at: doi:10.1016/j.ijfoodmicro.2009.01.015.

Emberland, K., Etherlberg, S., Kuusi, M., Vold, L., Jensvoll, L., Lindstedt, B., Nygard, K.,Torpdahl, M., Sorensen, G., Jensen, T., Lukinmaa, S., Niskanen, T., Kapperud, G.,2007. Outbreak of Salmonella Weltevreden infections in Norway, Denmark andFinland associated with alfalfa sprouts. Eurosurveillance 12, 3321. July–October2007.

European Commission, 2002. Regulation (EC) No 178/2002 of the EuropeanParliament and of the Council of 28 January 2002 Laying Down the GeneralPrinciples and Requirements of Food Law, Establishing the European FoodSafety Authority and Laying Down Procedures in Matters of Food Safety. OfficialJournal of the European Communities, L 31 1.2.2002, pp. 1–24.

European Commission, 2008. The Rapid Alert System for Food and Feed (FASFF)Annual Report 2007. European Communities, Luxembourg.

Food Standards Agency, 2006. Food Law Code of Practice. FSA, London. Availableat: www.food.gov.uk/enforcement/foodlaw/foodlawcop/copengland (accessed20.07.08).

Food Standards Agency, 2009. Food Alerts, 2007–2008. Food Standards Agency.Available at: http://www.food.gov.uk/enforcement/alerts (accessed 23.02.09).

Frost, J., 1994. In: Chart, H. (Ed.), Methods in Practical Laboratory Bacteriology. CRCPress, New York, pp. 73–82.

Health Protection Agency, 2005. Enumeration of beta-glucuronidase positiveEscherichia coli – Most Probable Number Method. National Standard Method F22. Health Protection Agency. http://www.hpastandardmethods.org.uk/pdf_sops.asp.

Health Protection Agency, 2008. Detection of Salmonella species. National StandardMethod F13. Health Protection Agency. http://www.hpa-standardmethods.org.uk/pdf_sops.asp.

International Organisation for Standardisation, 2007. Microbiology of Food andAnimal Feeding Stuffs – General Requirements and Guidance for Microbiolog-ical Examinations. ISO, Brussels. ISO/FDIS 7218: 2007.

de Jong, B., Andersson, Y., Giesecke, J., Hellstrom, L., Stamer, U., Wollin, R., 2001.Salmonella Typhimurium outbreak in Sweden from contaminated jars of helva(or halva). Eurosurveillance Weekly 5, 1715.

Kirk, M., Little, C., Lem, M., Fyfe, M., Genobile, D., Tan, A., Threlfall, J., Paccagnella, A.,Lightfoot, D., Lyi, H., McIntyre, L., Ward, L., Brown, D., Surnam, S., Fisher, I., 2004.An outbreak due to peanuts in their shell caused by Salmonella enterica sero-types Stanley and Newport – sharing molecular information to solve interna-tional outbreaks. Epidemiol. Infect. 132, 571–577.

Koch, J., Schrauder, A., Alpers, K., Werber, D., Frank, C., Prager, R., Rabsch, W.,Broll, S., Feil, F., Roggentin, P., Bockemuhl, J., Tschape, H., Ammon, A., Stark, K.,2005. Salmonella Agona outbreak from contaminated aniseed, Germany. Emerg.Infect. Dis. 11, 1124–1127.

Little, C.L., Gillespie, I.A., 2008. Prepared salads and public health. J. Appl. Microbiol.105, 1729–1743.

Local Authorities Co-ordinators of Regulatory Services, 2006. LACORS Guidance onFood Sampling for Microbiological Examination. Issue 2. Local Authorities Co-ordinators of Regulatory Services. Available at: www.lacors.com (accessed20.07.08).

O’Grady, K., Powling, J., Tan, A., Valcanis, M., Lightfoot, D., Gregory, J., Lalor, K.,Guy, R., Ingle, B., Andrews, R., Crerar, S., Stafford, R., 2001. Salmonella Typhi-murium DT104-Australia, Europe. No. 20010822.1980. Available at: www.promedmail.org (accessed 20.07.08).

Ollis, T., Meyer, B., Howe, P., 1999. Australian food sources and intakes of omega-6and omega-3 polyunsaturated fatty acids. Ann. Nutr. Metab. 43, 346–355.

PHLS, 2000. Guidelines for the microbiological quality of some ready-to-eat foodssampled at the point of sale. Commun. Dis. Public Health 3, 163–167.

Pierce, R.,1970. Sunflower processing techniques. J. Am. Oil Chem. Soc. 47, 248A–269A.Popoff, M., Le Minor, L., 2001. Antigenic Formulas of the Salmonella Serovars, eighth

ed. WHO collaborating centre for reference and research on Salmonella, InstitutPasteur, Paris.

Proctor, M., Hamacher, M., Tortorello, M., Archer, J., Davis, J., 2001. Multistateoutbreak of Salmonella serovar Muenchen infections associated with alfalfasprouts grown from seeds pretreated with calcium hypochlorite. J. Clin.Microbiol. 39, 3461–3465.

Rabsch, W., Prager, R., Koch, J., Stark, K., Roggentin, P., Bockemuhl, J., Beckmann, G.,Stark, R., Siegl, W., Ammon, A., Tschape, H., 2005. Molecular epidemiology ofSalmonella enterica serovar Agona: characterization of a diffuse outbreak causedby aniseed-fennel-caraway infusion. Epidemiol. Infect. 133, 837–844.

Roberts, D., Greenwood, M., 2003. Practical Food Microbiology. Blackwell PublishingLtd.

Salunkhe, D., Chavan, J., Adsule, R., Kadam, S., 1992. World Oilseeds: Chemistry,Technology and Utilization. Springer.

Sito, S., Barcic, J., Ivancan, S., 1999. Effect of different airflow speeds upon theduration of the drying process of high moisture pumpkin seeds. Agric. Conspec.Sci. 64, 133–136.

Sivapalasingam, S., Friedman, C., Cohen, L., Tauxe, R., 2004. Fresh produce:a growing cause of outbreaks of foodborne illness in the United States, 1973through 1997. J. Food Prot. 67, 2342–2353.

C. Willis et al. / Food Microbiology 26 (2009) 847–852852

Unicomb, L., Simmons, G., Merritt, T., Gregory, J., Nicol, C., Jelfs, P., Kirk, M., Tan, A.,Thomson, R., Adamopoulos, J., Little, C., Currie, A., Dalton, C., 2005. Sesame seedproducts contaminated with Salmonella: three outbreaks associated with tahini.Epidemiol. Infect. 133, 1065–1072.

Van Beneden, C., Keene, W., Strang, R., Werker, D., King, A., Mahon, B., Hedberg, K.,Bell, A., Kelly, M., Balan, V., MacKenzie, W., Fleming, D., 1999. Multinationaloutbreak of Salmonella enterica serotype Newport infections due to contami-nated alfalfa sprouts. JAMA 281, 158–162.

Werner, S., Boman, K., Einemo, I., Erntell, M., Helisola, R., de Jong, B., Lindqvist, A.,Lofdahl, S., Meeuwisse, A., Ohlen, G., Olsson, M., Persson, I., Runehagen, A.,Rydevik, G., Stamer, U., Sellstrom, E., Andersson, Y., 2007. Outbreak of

Salmonella Stanley in Sweden associated with alfalfa sprouts, July–August 2007.Eurosurveillance 12, 3291.

Winthrop, K.L., Palumbo, M.S., Farrar, J.A., Mohle-Boetani, J.C., Abbott, S.,Beatty, M.E., Inami, G., Werner, S.B., 2003. Alfalfa sprouts and Salmonella Kott-bus infection: a multistate outbreak following inadequate seed disinfectionwith heat and chlorine. J. Food Prot. 66, 13–17.

Wittenberg, M.M., 2007. In: Wittenberg, M.M. (Ed.), New Good Foods:Essential Ingredients for Cooking and Eating Well. Ten Speed Press,Canada, pp. 141–166.

Yu, L., Zhou, K., Parry, J., 2005. Antioxidant properties of cold-pressed black caraway,cranberry and hemp seed oils. Food Chem. 91, 723–729.