Characterization of microbial communities and fungal metabolites on field grown strawberries from organic and conventional production

International Journal of Food Microbiology 160 (2013) 313–322

Contents lists available at SciVerse ScienceDirect

International Journal of Food Microbiology journal homepage: www.elsevier.com/locate/ijfoodmicro

Characterization of microbial communities and fungal metabolites on field grown strawberries from organic and conventional production Birgit Jensen a,⁎, Inge M.B. Knudsen a, Birgitte Andersen c, Kristian Fog Nielsen c, Ulf Thrane c, Dan Funck Jensen a, d, John Larsen b, e a

Department of Plant and Environmental Sciences, University of Copenhagen, Thorvaldsensvej 40, DK-1871 Frederiksberg C, Denmark Department of Agroecology, Aarhus University, Forsøgsvej 1, DK-4200 Slagelse DTU Systems Biology, Technical University of Denmark, Søltofts Plads 221, DK-2800 Kgs. Lyngby, Denmark d Department of Forest Mycology and Plant Pathology, Swedish University of Agricultural Sciences, SLU, Box 7026, Almas Allé 5, 750 07 Uppsala, Sweden e Centro de Investigaciones en Ecosistemas, Universidad Nacional Autónoma de México, Apartado Postal 27-3, CP 58089 Morelia, Michoacán, Mexico b c

a r t i c l e

i n f o

Article history: Received 30 July 2012 Received in revised form 22 October 2012 Accepted 5 November 2012 Available online 13 November 2012 Keywords: Microbial communities Strawberry phyllosphere Biocontrol Organic production Secondary metabolites Mycotoxins

a b s t r a c t The background levels of culturable indigenous microbial communities (microbiotas) on strawberries were examined in a field survey with four conventional and four organic growers with different production practise and geographic distribution. The microbiota on apparently healthy strawberries was complex including potential plant pathogens, opportunistic human pathogens, plant disease biocontrol agents and mycotoxin producers. The latter group was dominated by Penicillium spp. and Aspergillus niger was also isolated. As expected, bacteria were the most abundant and diverse group of the strawberry microbiota followed by yeasts and filamentous fungi. No obvious correlation between grower practice and the strawberry microbiota was observed. Differences between microbiotas on strawberries from conventional systems with up to 10 fungicide spray treatments and organic production systems were insignificant. Mycotoxins were not detected in mature strawberries from any of the eight different growers neither in additional samples of low quality berries. However, isolates of Penicillium expansum and A. niger produced high amounts of mycotoxins when incubated on strawberries at 25 °C. Penicillium polonicum produced cyclopenol, cyclopenin, and viridicatin on the artificially infected berries, while Alternaria arborescens produced tenuazonic acid, Alternaria tenuissima produced altertoxin I and altenuene, and Trichoderma spp. produced several peptaibols. In conclusion, native strawberry microbiotas are highly diverse both in terms of taxonomic groups and functional traits that are important in relation to plant and human health. © 2012 Elsevier B.V. All rights reserved.

1. Introduction Strawberry is an important horticultural crop worldwide produced both conventionally and organically in open field and in plastic covered tunnels. Intensive pesticide treatments especially with fungicides are conducted in the conventional production, mainly due to disease problems caused by the grey mould pathogen Botrytis cinerea, while in tunnel production powdery mildew caused by Podosphaera aphanis is an emerging problem. The high number of pesticide application results in pesticide residues in more than 50% of strawberry samples, as analysed by the Danish Veterinary and Food Administration (Jensen et al., 2010). Alternative control measures are therefore desirable. Biocontrol in terms of microbial pest control agents (MPCAs) not only can be legally applied as a biological resource in

⁎ Corresponding author. Tel.: +45 35 33 33 34; fax: +45 35 33 33 00. E-mail address: [email protected] (B. Jensen). 0168-1605/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.ijfoodmicro.2012.11.005

organic production, but also can be exploited in conventional production e.g. in an integrated approach with fungicides. As for the pesticides it is crucial to make proper risk assessment analyses of MPCAs before their practical applications (Laengle and Strasser, 2010). However, data on background and natural exposure levels of microorganisms which are, or potentially could be, the active organisms in MPCAs, are rarely available. Such data would help to estimate if there is any risk caused by application of MPCAs to the environment and in that way improve our understanding of what can be considered safe implementation of MPCAs in crop production. Plants host a broad range of fungal and bacterial epiphytic microorganisms, which play an important role in their growth, health and fitness (Kinkel, 1997; Andrews and Harris, 2000). The knowledge on the natural epiphytic microbiota on strawberries is limited. Fungal communities on mature strawberries have been shown to be dominated by the filamentous fungi B. cinerea, Penicillium spp., Alternaria spp., Cladosporium spp., Rhizopus spp., Aureobasidium pullulans and the

314

B. Jensen et al. / International Journal of Food Microbiology 160 (2013) 313–322

yeast Cryptococcus spp. (Dennis, 1976; Tronsmo, 1986; Weidenbörner et al., 1995; Tournas and Katsoudas, 2005). Bacterial populations on strawberry plants are dominated by Pseudomonas spp., Stenotrophomas spp., Bacillus spp. and Arthrobacter spp. (Krimm et al., 2005). Among the fungi isolated from strawberries Penicillium spp. and Alternaria spp. are known as potential mycotoxin producers (Andersen et al., 2002; Frisvad and Thrane, 2002; Andersen et al., 2004; Christensen et al., 2005), but we have found no reliable studies on mycotoxin production in strawberry. Additionally, Rhizopus spp. are frequently isolated from strawberries and some isolates can produce toxins originating from endosymbiont bacteria (Jennessen et al., 2005; Lackner et al., 2009). Trichoderma spp. and Ulocladium spp. used as MPCAs do not produce mycotoxins per se, but Trichoderma spp. are known to produce several biologically active compounds with the peptaibols (cyclic peptides where some of the amino acids have been substituted by hydroxy-organic acids) and semi-volatile pyrones considered to be the active antifungal compounds (Degenkolb et al., 2008b; Andersen and Hollensted, 2008). The objective of the present work was to elucidate whether naturally occurring microorganisms, with special emphasis on toxigenic fungi, and potential MPCAs pose a potential human health risk. The characterization of the background microbiota was done on mature, flawless strawberries grown in open fields sampled from four organically and four conventionally cultivated fields, respectively.

strawberries grown either conventionally or organically and to relate the microbiota to berry analyses of potential harmful metabolites. Berries from the eight growers mentioned above were collected on the same day mid July 2006. Approximately 700 high quality berries (flawless, with no blemish) were collected randomly from each grower. Berries were picked and placed directly into boxes and kept separated in order to avoid cross contamination between berries and growers. In addition 30 low quality berries (with disease symptoms, bruises or injuries) were selected from each grower and kept strictly separated from the high quality berries. Sampled berries were stored at 4 °C until isolation of microorganisms the following day. Furthermore sub-samples for metabolite analysis were stored at −20 °C. 2.3. Isolation of fungi, yeasts and bacteria from berries

Four organic and four conventional growers, all producing the strawberry cultivar Florence in open field, were included in the study. Even though the microbiota investigation from the eight growers was based on the same strawberry cultivar, several other factors varied between growers. To establish differences in production methods a questionnaire with supplementary information was devised in collaboration with The Horticultural Advisory Service (Aarhus, Denmark). Growers were asked for information on pest and disease management, fertilization, irrigation and other practices such as row distance and cover material and so on (Supplementary Tables 1 and 2).

Three samples each containing 200 g of whole berries were taken from each grower (Fig. 1). Each sample was mixed carefully with 400 ml sterile water (0.1% Triton-X, Sigma-Aldrich, St. Louis, MO) in a plastic bag. The bags were placed in an ultrasonic bath (Branson 3210, Danbury, CT) for 30 min. The subsequent isolation of microorganisms from berries followed a three-step procedure (Fig. 1). From each bag: i) the wash water was plated in 10-fold dilutions (10 −1 to 10 −5) onto the isolation media DG18 (Dichloran 18% Glycerol agar) (Samson et al., 2010), PDA (Potato Dextrose Agar) (Samson et al., 2010) amended with 50 ppm chlortetracycline and 50 ppm chloramphenicol, TSA (Tryptic Soy Agar; Difco Bacto, Kansas, USA), SDA (Sabouraud Dextrose Agar; Difco Bacto, Kansas, USA) and NA (Nutrient Agar; Difco, USA), ii) nine pieces of strawberries (approx. 3 × 3 mm) were plated directly onto each of the isolation media DG18, V8 (V8 juice agar) (Samson et al., 2010), SNA (Spezieller Nährstoffarmer Agar) (Samson et al., 2010) and PDA, all media with antibiotics as above, and iii) one hundred grams of strawberries was homogenized with 200 ml of sterile water (0.1% Triton-X) for 2 min in a Stomacher 400 (Seward Medical, London, UK) and plated in 10-fold dilutions (10−1 to 10 −2) onto DG18, V8, SNA, PDA with antibiotics as above, TSA, SDA and NA. Plates were incubated according to the conditions listed in Table 1. After 7 days of incubation fungi were identified and counted directly on the isolation plates when possible. Problematic fungi were identified by microscopy. Total numbers of bacterial and yeast colonies were counted after 3 and 2 days of incubation, respectively.

2.2. Sampling of berries for survey

2.4. Identification and metabolite profiling of fungi

Field grown strawberries were sampled in order to characterize the microbiota (quantification and strain identification) of edible

After seven days of incubation fungi were identified to genus level directly on the isolation plates using a stereomicroscope. Fungal

2. Materials and methods 2.1. Growers

Fig. 1. Outline of sampling of berries from eight strawberry growers and the three-step (i, ii and iii) procedure for isolation of fungi, yeasts, and bacteria from berries.

B. Jensen et al. / International Journal of Food Microbiology 160 (2013) 313–322 Table 1 Media and incubation conditions used for quantification, isolation and identification of fungi, yeasts and bacteria. Code

Mediuma

Target

Incubation

Step

DG18

W, H, Bd

SDA

Sabouraud dextrose agar

Fungi, yeasts Fungi, yeasts Yeast

25 °C, darkness

PDA

Dichloran 18% glycerol agarb Potato dextrose agarc

NA

Nutrient agar

TSA

Tryptic soy agar

V8

V8 juice agar

b

SNA

Spezieller Nährstoffarmer agarb CYA Czapek-Dox yeast autolysate agar DRYES Dichloran rose bengal yeast extract agar MEA Malt extract agar (Blakeslee) PCA Potato carrot agar SNA YEA

Spezieller Nährstoffarmer agar Yeast extract agar

21–23 °C, 12 h light A 20–23 °C, darkness Bacteria 20–23 °C, darkness Bacteria 20–23 °C, darkness Fungi, 21–23 °C, 8 h yeasts light B Fungi, 21–23 °C, 12 h yeasts light A Fungi 25 °C, darkness

W, H, B

Fungi

25 °C, darkness

Identification

Fungi

25 °C, darkness

Identification

Fungi

21–23 °C, 8 h light B 21–23 °C, 12 h light A 25 °C, darkness

Identification

Fungi Fungi

W, H W, H W, H H, B H, B Identification

Identification Identification

a

Fungal medium recipes given in Samson et al. (2010), yeast medium (TSA) and the bacterial media (SDA and NA) were prepared according to the manufacture's directions (Difco, USA). b Addition of antibiotics: 50 ppm chloramphenicol and 50 ppm chlortetracycline. c Addition of antibiotics: 50 ppm chloramphenicol and 25 ppm chlortetracycline. d W = plate dilution of wash water, H = plate dilution of homogenate of washed berries and B = washed berry pieces.

isolates, 117 in total, were isolated for identification to species level. For morphological and chemical identification, Penicillium and Aspergillus were grown on CYA (Czapek-Dox Yeast Autolysate Agar) (Samson et al., 2010), MEA (Malt Extract Agar) (Samson et al., 2010) and YES (Yeast Extract Sucrose agar) (Samson et al., 2010), Alternaria on DRYES (Dichloran Rose bengal Yeast Extract Sucrose agar) (Samson et al., 2010) and PCA (Potato Carrot Agar) (Samson et al., 2010), and Trichoderma on MEA and SNA (Samson et al., 2010) were used. Incubation conditions are shown in Table 1. Secondary metabolites were extracted from three 6 mm plugs of 7 day old cultures using methanol/dichloromethane/ethyl acetate (1:2:3 v/v/v) as described by Smedsgaard (1997). Penicillium and Aspergillus extractions were done from cultures on YES; Alternaria from cultures on DRYES; and for Trichoderma the cultures on MEA were used. Extracts were analysed by HPLC–DAD with an acidic water–acetonitrile system on a Luna C18 column as described in details by Nielsen et al. (2009).

2.5. Identification of bacteria and yeasts From each of the eight growers ten bacterial and yeast colonies were randomly selected and isolated from each of the three replicates (30 bacterial and 30 yeast strains per grower). Both yeasts and bacteria were identified by using the Microbial Identification System (MIDI) based on fatty acid profiles from 24 h cultures using the four-step fatty acid extraction procedure (Sasser, 1990; Mansfeld-Giese et al., 2002): i) saponification, ii) methylation, iii) extraction and iv) base wash. Fatty acid methyl ester analyses were performed with an Agilent 6890 Plus Chromatograph and the Sherlock System Software 4.0 with the recommended libraries for aerobic heterotrophic bacteria and yeasts (Parsley, 1996).

315

2.6. Inoculation of berries with potentially toxic field fungi In order to study the worst-case probability of mycotoxin production in strawberries, fungal isolates with validated ability to produce mycotoxins were used for the inoculation experiments. Organic grown strawberries were purchased in two supermarkets, which supplied the cultivar Honeoye and an unspecified cultivar. The berries were surface disinfected in 0.4% hypochlorite solution for 1 min, washed in three sets of sterile water and allowed to dry for 15 min in a LAF bench. Berries were inoculated individually with seven different fungal strains from the IBT Culture Collection at DTU Systems Biology Penicillium brevicompactum IBT 28473, Penicillium bialowiezense IBT 25262, Penicillium polonicum IBT 28415, Penicillium expansum IBT 21525, Penicillium verrucosum IBT 28162, Aspergillus niger IBT 28144, Alternaria tenuissima IBT 41145, Alternaria arborescens IBT 41065 and Trichoderma sp. IBT 41139. Conidial suspensions were made by pouring sterile water containing 1% agar over the surface of the petri dish. The suspension was transferred to a beaker and adjusted to 103 or 105 conidia/ml (in duplicate for each berry supplier). Each berry was dipped in the suspension for ca. 2 s and then transferred to a small glass jar and loosely capped. After 4–8 days of incubation at 25 °C, the berries were visually inspected and clearly moulded berries were stored at −20 °C for later chemical analysis. For P. expansum a slightly different inoculation procedure was used. Berries of the cultivar Florence were rinsed three times in sterile water and then sprayed until run off with a suspension of 107 conidia/ml and incubated four days in plastic boxes at 4 °C and 20 °C, respectively. Berries for analysis were stored at −20 °C. 2.7. Metabolite analysis of berries Field samples as well as inoculated berries were analysed by LC-MS/MS for fumonisins B1, B2, B4 and B6, ochratoxins A, B, β and α, mycophenolic acid, citrinin, tenuazonic acid, malformines, infectopyrones, phomapyrone A, brevianamides A and B, cyclopenol, cyclopenin, viridicatin, AAL Toxins, as well as the major peptaibols observed in Trichoderma biocontrol strains (Supplementary Table 3). In brief, 5–15 g berries were homogenised with 6 volumes of acetonitrile in a stomacher (Seward Medical, London, UK) for 1 h, samples were then centrifuged at 10 000 g, and 1 ml transferred to vial fitting the auto-sampler. Sub-samples of 0.1 to 3 μl were injected on an Agilent 1100 LC system (Waldbronn, Germany), and separated on a Gemini C6-Phenyl 3 μm 2-mm ID× 50-mm column (Phenomenex, Torrance, CA, USA) using a acetonitrile–water gradient system buffered with 20 mM formic acid. The LC was coupled to a Quattro Ultima triple mass spectrometer (Waters-Micromass, Manchester, UK) with Z-spray ESI source as described in Sørensen et al. (2009). Chromatography and MS/MS were optimized on pure standards of the compounds listed in Supplementary Table 3. Crude extracts of four Trichoderma cf. harzianum biocontrol strains IBT 41406, IBT 41407, IBT 41408 and IBT 41409 from PDA cultures were used for optimising for the 9 major peptaibols. The method was quantitatively validated for fumonisin B2 and ochratoxin A in spiked non-infected berries (0, 20, 50, 100, 300, 1000, 5000, and 30 000 ng/g). Limit of detection (s/n 5) was approximately 20 ng/g for both. Patulin in berries was determined by HPLC with Diode array detection on an Agilent 1100 system equipped with a 15 cm, 2 mm ID, 3 μm, Luna PFP column (Phenomenex, Torrance, CA, USA). The HPLC was running an acidic (50 ppm trifluoroacetic acid)-acetonitrile gradient from 5 to 50% acetonitrile in 15 min using 276 nm as detection wavelength. Strawberries (subsample of 1.0 g) were homogenized with five 3.2 mm stainless steel balls in a 8-ml polypropylene tube with 3 ml water using Mini-beadbeater (Biospec Products Inc., Bartlesville, OK, USA) and the homogenate along with further 7 ml water was then extracted for 1 h on a rotary shaker at 200 rpm, centrifuged at 20 000 g, and the supernatant was acidified with 3 ml water containing 2%

B. Jensen et al. / International Journal of Food Microbiology 160 (2013) 313–322

Organic 4

Data were subjected to two-way analyses of variance. ANOVAs with production system and grower as class variables were performed. Bartlett's test for variance homogeneity was examined for all variables and when relevant, data were log10 transformed prior to ANOVA. Significant differences between treatment means were based on LSD values. In all analyses the level of significance was fixed at α = 0.05.

3

2

1

3.2. Bacteria Bacteria made up the largest proportion of the total microbiota on strawberries with maximum densities of >10 4 CFU/g berry (Figs. 2 and 3). The density of cultivable bacteria was significantly higher on organic berries compared with conventional berries both in wash water (P b 0.0019) and in washed homogenized berries (P b 0.0128). For both quantification methods, growers #3 and #4 had significantly lower bacterial CFU/g berry (P b 0.05) than the other six growers (data not shown). Bacterial communities on strawberries from conventional and organic growers are shown in Table 2. The bacteria most commonly isolated belonged to the genera Curtobacterium, Serratia, Pseudomonas, Enterobacter and Rahnella. In total 34 different species from 23 different genera were identified. Fourteen species were only encountered from the organic growers and nine bacterial species were only isolated from conventional growers.

yeasts

bacteria

Cladosporium

Fig. 2. Quantification of the microbiota from healthy berries sampled from four organic and four conventional strawberry growers from washing water of the berries. * = P ≤ 0.05 or ** = P ≤ 0.001 means significant difference between conventional and organic growing system.

3.4. Filamentous fungi 3.4.1. Filamentous fungi loosely attached to the berries The most common fungal genera on the surface of the strawberries (found via the wash water) are given in Fig. 2. Cladosporium was the most abundant fungal genus isolated from wash water of both organic and conventional strawberries. Mean densities were >103 CFU/g berry (Fig. 2). Significantly higher CFU/g berry (P b 0.04) was isolated from the organically grown berries. Penicillium was the most abundant genus with mycotoxin producing potential. There was no significant difference between the two production systems (Fig. 2) and the mean density of Penicillium was approximately 200 CFU/g berry. However, there were significant differences in CFU densities especially among the organic growers (P b 0.001) as Penicillium CFU/g berry from growers #5 and #7 was at least 10-fold higher than from growers #6 and #8 (data not shown). Alternaria was isolated in low numbers and only in wash water from 6 out of 24 samples. On average b4 CFU/g berry was obtained from both conventional and 5 Conventional Organic

4

Log10 CFU/g berry

Based on a questionnaire of four conventional (1–4) and four organic growers (5–8) differences and similarities between the production systems were described and the results are summarized in Supplementary Tables 1 and 2. The grower production systems differed mainly in pest management and fertilization procedures, as the organic farming protocols do not allow pesticides and inorganic fertilization, which are common among conventional growers. All the conventional growers used pesticides, but with varying intensity ranging from 2 to 10 fungicide treatments. Only one of the organic growers (#8) used MPCAs, namely Trichoderma. However, no Trichoderma spp. were isolated from berries of this grower. Furthermore, one of the conventional growers (#4) also used biocontrol of spider mites. One of the conventional growers (#1) used organic fertilization and fewer pesticides than the other conventional growers. All growers used straw as cover material (Supplementary Tables 1 and 2).

other fungi

0

3. Results 3.1. Grower practice

Conventional

Alternaria

2.8. Statistics

5

Penicillium

formic acid. The extract was then applied to a 30 mg Strata-X-C column (Phenomenex). The column was washed with 3 ml 2% NaHCO3 and 3 ml 1% formic acid and eluted using 600 μL acetonitrile–water (3:7 v/v), and subsequently 2 μL of this sample was injected in the HPLC. Quantitative validation was done in spiked non-infected berries (0, 1, 3, 10, 15, 20, 50 and 100 μg/g). Limit of detection was 1 mg/kg (s/n 5 at 276 nm).

log10 CFU/g berry

316

3

2

1

yeasts

bacteria

other fungi

0

Penicillium

The majority of yeasts were isolated from wash water (Fig. 2) rather than the homogenate (Fig. 3). No effect of growing system on yeast density neither from wash water nor from homogenate was observed. Yeast communities on strawberries from conventional and organic growers are shown in Table 3. The yeasts most commonly isolated belonged to the genera Candida, Cryptococcus and Rhodotorula. In total 22 different species from 9 different genera were identified. Six yeast species were only encountered from the organic growers and four yeast species were only isolated from conventional growers.

Cladosporium

3.3. Yeasts

Fig. 3. Quantification of the microbiota from healthy berries sampled from four organic and four conventional strawberry growers from homogenate of washed berries. *=P≤0.05, or **=P≤0.001 means significant difference between conventional and organic growing system.

B. Jensen et al. / International Journal of Food Microbiology 160 (2013) 313–322 Table 2 Percent recovery of bacteria species of strawberries sampled from four conventional growers (1–4) and four organic growers (5–8) in 2006. Identification is based on signature fatty acids. Bacteria

Conventional growers 1

2

3

Organic growers 4

Pct. recovery Alcaligenes piechaudii Bacillus megaterium-GC subgroup B Chromobacterium violaceum Clavibacter michiganense nebraskense Curtobacterium flaccumfaciens flaccumfaciens Curtobacterium flaccumfaciens oortii Curtobacterium flaccumfaciens poinsettiae Enterobacter agglomerans Enterobacter intermedius Erwinia carotovora atroseptica Hafnia alvei Kluyveria ascorbata Kocuria kristinae Kocuria varians Microbacterium esteraromaticum Micrococcus lylae-GC subgroup A Morganella morganii Nesterenkonia halobia Paenibacillus macerans-GC subgroup B Pantoea agglomerans Pseudomonas chlororaphis Pseudomonas putida Pseudomonas syningae phaseolicola Pseudomonas syringae atrofaciens Rahnella aquatilis Rhodococcus luteus Serratia fonticola Serratia grimesii Serratia liquefaciens Staphylococcus epidermidis Staphylococcus warneri Staphylococcus xylosus Stenotrophomonas maltophilia Yersinia frederiksenii Number of bacteria species

– – – – – – 13 – – – 3 3 – 3 – – 3 – – – 7 7 3 – 23 – 7 13 13 – – – – – 12

– – 3 – – 7 10 14 – – – – – – – – – – – – 3 14 7 – 7 – 10 – 24 – – – – – 10

– – – – 43 10 10 23 – – – – 3 – – – – – – – – 3 – – – – 3 – 3 – – – – – 8

5

6

7

317

Table 3 Percent recovery of yeast species of strawberries sampled from four conventional growers (1–4) and four organic growers (5–8) in 2006. Identification is based on signature fatty acids. Yeast species identified

Conventional growers

8

1

Pct. recovery – – – 7 50 17 17 – – 3 – – – – – – – – – – – – – – – 3 – 3 – – – – – – 7

– – 4 – 8 – 4 8 – – – – – – – – – – – 4 8 31 4 4 – – 15 – 8 – – – – 4 12

7 13 – – – – – 13 7 – – – – – – – – – – 27 7 7 – – – – 7 – – 7 7 – – – 10

– 12 – – – – – 6 – – – – – – 6 6 – – – 12 – 6 – – – – 12 – 12 – 6 6 – 18 11

2

3

Organic growers 4

Pct. recovery – – – 8 – – – – – – – – – – – – – 17 17 – 33 – – – 8 – – – – – 8 – 8 – 7

organic growers. The other mycotoxin producing genera, Aspergillus and Fusarium, were only detected in one sample each while potential MPCAs such as Trichoderma were detected from four growers and Clonostachys from two growers.

Candida acidothermophilum Candida cacaoi Candida castrensis Candida famata Candida fragariorum Candida glaebosa Candida inconspicua Candida paludigena Candida silvae Candida utilis Candida zeylanoides Cryptococcus albidus var. albidus Cryptococcus neoformans–GC subgroup B Cryptococcus terreus Hanseniaspora uvarum Hansenula anomala Metschnikowia pulcherrima Rhodotorula minuta-GC subgroup B Rhodotorula rubra Sporobolomyces salmonicolor Trichosporon beigelii-GC subgroup B Zygosaccharomyces bisporus No match Number of species

– – – 15 – – – – – – 8 – 8 19 – – 4 12 19 – – – 15 7

– 4 – 13 – – – 9 – – 9 9 4 4 – 13 – 9 9 – – – 17 10

– – 3 14 – 7 3 3 – – – – 3 7 – – – 10 45 3 – – – 10

5

6

7

8

Pct. recovery – – – 24 – – – 5 – – 5 – 14 19 – – – 14 5 – – – 14 7

– – – 11 – – – 5 5 – 11 – – 5 – – – 16 21 5 – 5 16 9

4 – – 22 – – – 4 – 9 13 4 – 9 4 – – 9 9 – – 4 9 11

– 5 – 24 5 – – 10 – – 10 – – 5 – 5 – – 24 14 – – – 9

– – – 4 – – – 4 – – 4 – 4 11 – – – 4 54 7 4 – 7 9

for the wash water (Fig. 2). A significantly higher CFU density (P b 0.05) was isolated from organic berries compared to conventional. Penicillium spp. was still present after washing of berries from all growers. However, there was no significant effect of growing system or of grower on Penicillium CFU density of the homogenized berries and the mean density was b10 2 CFU/g berry (Fig. 3). Alternaria spp. were also occasionally isolated from the homogenized berries while no Aspergillus spp. and Fusarium spp. were detected. 3.4.4. Filamentous fungi identified The identification of potentially mycotoxin producing fungi (Table 4) revealed 14 species of Penicillium (P. brevicompactum, 100

3.4.3. Filamentous fungi strongly attached or internally present in berries The most common fungal genus of the washed strawberries (found in the homogenized strawberry pulp) was Cladosporium (Fig. 3). The CFU densities were approximately 10 fold lower than

Conventional

% berries with growth

Organic

80

60

40

20

Botrytis

Aspergillus

Fusarium

Alternaria

Penicillium

0

Cladosporium

3.4.2. Filamentous fungi strongly attached to the berries The most common fungal genus on the surface of the strawberries (found on the washed strawberries) was Penicillium and this was also the most frequently isolated genus growing from both conventional and organic grown berries. On average Penicillium grew from more than 70% of washed berry pieces (Fig. 4). The percentage of berries with Cladosporium spp. growth was also high on berries from both production systems but the percentage varied significantly between growers (P b 0.0001). Growth of Alternaria spp. from washed berry pieces was significantly higher (P b 0.0001) from conventional berries compared to organic berries (Fig. 4). Growth of Fusarium spp., although below 4%, was only detected in conventionally grown berries. Berries from two out of four growers in both production systems had growth of Aspergillus spp. There was no significant effect of growing system on the frequency of Botrytis spp. (Fig. 4). It was noted that growers #1, #2, #5 and #8 had significantly lower Botrytis spp. frequency (P b 0.0001) than the four other growers (data not shown).

Fig. 4. Frequency of fungal genera growing from washed strawberry pieces on semi-selective agar media. Berries were sampled from four organic and four conventional strawberry growers. *=significant difference between conventional and organic growing system at 0.05 level.

318

B. Jensen et al. / International Journal of Food Microbiology 160 (2013) 313–322

Table 4 Identified fungal strains from conventional (C) and organic (O) growers, their recovery from wash water, homogenate of washed berries or from washed berry pieces location on the berry and their metabolite production in pure culture. Identification based on morphology and metabolite profiles. Species identification

Cropping system

Number of strains

Recovery of the straina and metabolites producedb Wash water

Alternaria tenuissima species-group Alternaria infectoria species-group Penicillium bialowiezense Penicillium brevicompactum Aspergillus niger Aspergillus sp. Penicillium thomii Trichoderma harzianum Trichoderma sp. Penicillium freii Penicillium punicae Penicillium manginii Penicillium aurantiogriseum Penicillium olsonii Penicillium polonicum Penicillium carneum Penicillium pulvillorum Penicillium expansum Penicillium verrucosum Penicillium ‘buchwaldii’ Talaromyces aculeatus Talaromyces purpurogenus complex

C O C O C O C O C O C O C O C O C C C C C C C C C O O O O C

6 1 5 2 2 4 5 4 1 1 1 0 1 3 3 1 1 2 2 1 1 1 3 1 1 1 1 2 1 1

Homogenate c

Berry pieces –

AOHs, Als, TeA

–

N-Z, Infec, Alx

N-Z, Infec, Alx

–

Rai, MycA, Quin

–

Rai, MycA, Quin

Rai, MycA, Brev

–

Rai, MycA, Brev

–

–

Gamma-pyrone

–

–

Pyranonigrin A

Nothing known

–

Nothing known

Pachybasin

–

–

6-pentyl-alfa-pyrone Viridicatol Nothing known Citreoviridin Penicillic acid Asperphenamate Viri, cyclopenols MycA – – – Rai, asperentins Rai, asperentins Nothing known

– – Nothing known – – – – – – – – – – –

– Viridicatol – – – – Viri, cyclopenols – Pulvillonic acid Patulin Verrucolone – – –

a

The strain is isolated from either wash water, homogenate of washed berries of from washed berry pieces. Als: altersetin; Alx: altertoxins; AOHs: alternariols; Brev: brevianamide; Infec: infectopyrones; MycA: mycophenolic acid; Quin: quinolactacin N-Z: novo-zealandins; Rai: Raistrick phenols; TeA: tenuazonic acid; Viri: viridicatol. c “–” indicates no isolation of the species. b

P. bialowiezense, Penicillium aurantiogriseum, Penicillium olsonii, Penicillium carneum, Penicillium freii, P. polonicum, Penicillium punicae, Penicillium pulvillorum, Penicillium manginii, Penicillium thomii, Penicillium verrucosum, P. expansum and Penicillium ‘buchwaldii’ [nomen provisorium, Frisvad, pers. comm.]) and additional two Talaromyces species, Talaromyces aculeatus (formerly Penicillium aculeatum) and Talaromyces purpurogenus complex (formerly Penicillium purpurogenum complex) that have been transferred from Penicillium due the newly accepted single name nomenclature in mycology (Samson et al., 2011). Other strains were identified to be two species groups of Alternaria: A. tenuissima species-group and Alternaria infectoria species-group, as well as A. niger and an unidentified Aspergillus sp. Five Trichoderma strains, all isolated from wash water were also identified (Table 4). Other fungi occurring on several plate dilutions were identified to be B. cinerea, Phoma spp., unidentified fungal strains resembling Acremonium and Hormomyces and one strain of Stemphylium herbarum was also identified. Zygomycetes, especially Rhizopus spp. were frequently the dominating fungi growing from washed strawberry pieces on V8, and DG18. However, since Zygomycetes were not of major interest in the present study, these were not quantified. 3.5. Mycotoxins 3.5.1. Metabolite analysis of pure fungi cultures Known metabolites from representative fungal strains belonging to mycotoxin producing genera are listed in Table 4. The analyses showed that mycotoxins sensu stricto were not detected in any of the pure culture extracts except for P. expansum producing patulin. Chemical identification and metabolite profiling of the Penicillium strains showed that P. bialowiezense and the closely related P. brevicompactum were the most common species. Both were able to produce Raistrick phenols

and mycophenolic acid. Chemical identification of the Alternaria strains, mostly found in berries from conventional growers, revealed strains from both the A. tenuissima and the A. infectoria speciesgroups. Strains from the A. tenuissima species-group were able to produce alternariols, altertoxins and tenuazonic acid and those from the A. infectoria species-group could produce novo-zealandins, altertoxin derivatives, and infectopyrones. 3.5.2. Metabolite analysis of artificially infected strawberries A. niger grew very well when it was inoculated on strawberries, and the berries were very mouldy and would never be eaten as fresh berries by consumers. As seen in Table 5 very high levels of fumonisins B2 and B4 and ochratoxin A were detected in these strawberries. The amounts of biomass produced by the penicillia were much more modest than for A. niger and the amounts of metabolites were very low. Cyclopenol, cyclopenin, and viridicatin were found in all eight P. polonicum infected samples (Table 5), but could not be quantified due to very low amounts of the reference standard (not commercially available). P. brevicompactum grew on berries but mycophenolic acid was only detected in one out of four samples. In contrast, P. bialowiezense did not grow on berries. On the other hand P. expansum grew vigorously on berries at 20 °C and up to 20 mg/kg patulin was detected while no patulin was detected in inoculated berries incubated at 4 °C. Citrinin was not detected (b1 μg/g) in the strawberries. Finally peptaibols were detected in all berry samples artificially inoculated with Trichoderma spp. (Table 5). 3.5.3. Metabolite analysis of field samples The 24 samples from the field survey (8 growers × 3 replicas) and 30 worst case samples (berries with disease symptoms or otherwise low quality berries) all tested negative for fumonisins B2, B4 and B6;

B. Jensen et al. / International Journal of Food Microbiology 160 (2013) 313–322

319

Table 5 Detection of metabolites in berries inoculated with strains of fungi naturally occurring on strawberries. The metabolites were analyzed on inoculated berries incubated at 25 °C for 4–8 days. Species

Metabolite

Samples analyzed (n)

Samples with metabolite (n)

Quantification

Aspergillus niger

Ochratoxin A Fumonisin B2 Fumonisin B4 Cyclopenol Cyclopenin Viridicatin Mycophenolic acid Tenuazonic acid Altertoxin 1 Altenuene Peptaibol 1 Peptaibol 2

7 7 7 8 8 8 4 6 7 7 8 8

6 7 7 8 8 8 1 5 1 2 8 8

29–1076 ppb 38–2058 ppb 129–25,578 ppb Not quantifieda Not quantifieda Not quantifieda 2–5 ppb 5 ppb–1 ppm Trace amounts (ppb) Trace amounts (ppb) Not quantifieda Not quantifieda

Penicillium polonicum

Penicillium brevicompactum Alternaria arborescens Alternaria tenuissima Trichodema harzianum a

Detected but not quantified as reference standards are not available.

ochratoxins A, B, β, and α; mycophenolic acid, tenuazonic acid, the 9 Trichoderma peptaibols and the remaining metabolites are listed in Supplementary Table 3.

with spores; berry pieces = mainly mycelium growth) suggests that most of the Alternaria and Trichoderma species are located on the surface, while Penicillium species and Aspergillus species are found on the surface as well as in the interior of the berries.

4. Discussion 4.2. Yeasts Results from the present study of conventionally and organically grown strawberries revealed highly diverse microbial communities on berries including potential plant pathogens, opportunistic human pathogens, plant disease biocontrol agents and mycotoxin producers. To our knowledge this study is the first extensive characterization of microbial communities and metabolites on strawberries covering major groups of phyllosphere microorganisms including culturable filamentous fungi, yeasts and bacteria. 4.1. Filamentous fungi Cladosporium was the dominating filamentous fungus recovered from strawberries. However, potential mycotoxin producing genera such as Penicillium and in lower number Alternaria, Talaromyces and Fusarium also appeared frequently. Several Aspergillus spp. were also isolated from both washing and berry pieces. This was a surprise since the Aspergillus genus has not been mentioned in the literature among fungal genera isolated from high quality strawberries (Dennis, 1976; Tronsmo, 1986; McLean and Sutton, 1992; Tournas and Katsoudas, 2005). Trichoderma and Clonostachys strains were only isolated from few berry samples of the present field experiments. The isolated MPCA strains occurring in the background mycobiota were characterized as Trichoderma harzianum, Trichoderma sp. and Clonostachys rosea. This is in accordance with investigations from Norway (Tronsmo, 1986) and Canada (McLean and Sutton, 1992) finding none of the genera occurring naturally on berries. In contrast Parikka et al. (2009) reported abundant isolation of Trichoderma viride from Finish strawberries and Tournas and Katsoudas (2005) isolated Trichoderma spp. in 3% of strawberry samples purchased from supermarkets in USA. A wide range of Penicillium species were identified both from wash water and berry pieces. Several of these species are known as mycotoxin producers e.g. the predominant species P. brevicompactum as well as P. verrucosum and P. expansum. We also isolated and identified other potential mycotoxin producers namely two species groups of Alternaria: A. tenuissima species-group and A. infectoria species-group, as well as A. niger. The results of the survey show that only a limited number of culturable fungal species can be found on the surface or in the interior of the berries. Less than 30 fungal species from 20 fungal genera in total have been found on the berries compared to the hundreds of known food-borne fungal species. Analysis of the location of the different fungal species on the berries (wash water = surface contamination

In total 22 species from 9 genera were identified of which species from the genera Candida, Cryptococcus and Rhodotorula were dominant. Several yeasts have been reported as biocontrol agents mainly of post harvest diseases, such as blue, green and grey moulds (Sharma et al., 2009). In the present study 8 of the 22 species recovered have been reported to have biocontrol traits of which the species Candida famata (Arras, 1996), Rhodotorula minuta (Patiño-Vera et al., 2005) and Rhodotorula rubra (Dal Bello et al., 2008) were most frequently recovered. Among the yeast species isolated several are also known as opportunistic pathogens in humans causing infectious diseases. In total 6 of the 22 species recovered have been reported as human pathogens. Among the most frequently recovered species Cryptococcus neoformans has been associated with fungal meningoencephalitis (Lix, 2009) and the Candida species C. famata (Pfaller and Diekema, 2004) and Candida inconspicua (Loeffler et al., 2000) have been associated with infectious diseases. However, in general yeast other than Candida albicans are uncommon human pathogens and most commonly associated with immuno-compromised humans. 4.3. Bacteria Bacteria from the genera Curtobacterium, Serratia, Pseudomonas, Enterobacter and Rahnella were most frequently isolated in the present study. Krimm et al. (2005) reported Pseudomonas, Stenotrophomas, Bacillus and Arthrobacter as the dominating epiphytic bacteria of strawberry plants (leaves, flowers and berries). Methods used to isolate, quantify and identify bacterial phyllosphere populations differed between the two studies, which have been shown to play an important role in recovering of bacterial communities (Jacques and Morris, 1995). Krimm et al. (2005) isolated bacteria from agar imprints and wash water and the bacterial strains were identified employing morphotyping and DNA based PCR methods. Bacteria with known biocontrol traits against plant pathogens recovered from the present study included Bacillus megaterium (Abanda-Nkpwatt et al., 2006), Paenibacillus macerans (Li et al., 2007), Pantoea agglomerans (Francés et al., 2006), Pseudomonas chlororaphis (Johnson et al., 1998), Pseudomonas putida (Bora et al., 2004) and Serratia liquefaciens (Whiteman and Stewart, 1998). Potential opportunistic human pathogenic bacteria species isolated from strawberry in the present study included Rahnella aquatilis (Matsukura et al., 1996), Hafnia alvei (Ramos and Dámaso, 2000), Chromobacterium violaceum (Teoh et al., 2006) and different

320

B. Jensen et al. / International Journal of Food Microbiology 160 (2013) 313–322

Staphylococcus species (Otto, 2009). These bacteria are reported to be causing different infectious human diseases mainly related to immune-deficient hospitalized patients. In most cases the frequency was low (3–8%) and they were recovered only from 1 to 3 nurseries. One of the strains (R. aquatilis), was recovered in one nursery at a higher frequency (23%). Among the potential bacterial human pathogens, Staphylococcus strains can cause severe human infections (Otto, 2009). Opportunistic human pathogenic bacteria have also been reported associated with the rhizosphere of several crops (Berg et al., 2005).

production was detected in berries inoculated with a Trichoderma strain from the naturally occurring strawberry mycobiota. However, these metabolites are not considered harmful to higher animals. Since mouldy strawberries presumably are not consumed directly but discarded during cleaning, the risk of exposure to e.g. fumonisins and ochratoxin A is minimal. However, strawberries for processing, e.g. jams, may contain mouldy berries, as the cleaning step is not done by hand on a single berry level. In this case it is obvious that a few mouldy berries can contaminate a whole batch. 4.5. Cropping system

4.4. Mycotoxins Only a limited number of metabolites were produced by pure cultures of fungi isolated from the surface or the interior of the berries and none of these were mycotoxins sensu stricto except patulin produced by P. expansum. However, this species was seldom isolated. Therefore the Penicillium species, P. bialowiezense and P. brevicompactum, constitute the biggest mycotoxigenic potential amongst the fungal species detected in this survey, due to their potential production of the immunosuppressive mycophenolic acid and their frequent occurrence in berries from all growers. P. brevicompactum is very common on many fruits, berries and in soil and is able to produce mycophenolic acid on most substrates, including ginger, building materials and all agar substrates tested (Frisvad et al., 2007b). The other Penicillium and Talaromyces species isolated are not known to be producers of regulated mycotoxins (Samson and Frisvad, 2004), which were confirmed by the analyses. Two Alternaria species were found in strawberries, the surface growing A. tenuissima is able to produce the mutagenic alternariols, tenuazonic acid and altertoxins. The other species is A. infectoria, which may infect the berries and is known as potential producer of infectopyrones, novo-zealandins and many undescribed compounds of unknown toxicity (Andersen et al., 2002). A few strains of A. niger were found and this species and the closely related A. carbonarius are very common contaminant of grapes where they produce the carcinogenic mycotoxin, ochratoxin A. This mycotoxin is a significant problem in both wine and raisins and many other products. Recently it has been shown that A. niger also produces the carcinogenic compound fumonisin B2 (Frisvad et al., 2007a) and this species is the reason for the contamination of wine and raisins with fumonisins (Mogensen et al., 2010; Knudsen et al., 2011). As A. niger is very common in soil and capable of growing at low pH on media with high concentrations of organic acids, it is a potential problem in strawberries. However, the production of mycotoxins is regulated by environmental factors and furthermore also strain dependent. The genus of Trichoderma is generally not known for many mycotoxins, and the 3 known mycotoxins: harzianum A, trichodermin and gliotoxin are produced by Trichoderma species that are phylogenetic unrelated to the species used as biocontrol agents (Nielsen et al., 2005; Degenkolb et al., 2008a). Many species in Trichoderma produce peptaibols that make channels in membranes and thereby kill other microorganisms (Degenkolb et al., 2008b). So far, peptaibols have not been shown to be toxic to higher animals when introduced via a natural route. Our results from inoculation of berries with strawberry associated fungi show that A. niger can produce high amounts of fumonisins and ochratoxin A on strawberries – up to 2000 times the regulatory limit – highlighting that A. niger infected strawberries should be avoided. Furthermore, P. expansum also produced patulin on strawberries up to 2000 times the EU regulatory limit in baby food. For all the other artificial worst-case scenarios, very low levels of known fungal metabolites were detected. Surprisingly, mycophenolic acid was produced in very low levels by P. brevicompactum, and the other mycophenolic acid producer, P. bialowiezense, was unable to grow on strawberries, minimizing the risk for mycophenolic acid contamination of strawberries. Peptaibol

Only few differences between the strawberry microbiota of organic and conventional berries were identified. In theory the high use of fungicides in conventional systems may affect populations of non-target fungi. Higher populations of Cladosporium were recovered from organic berries compared to conventional berries, but on the other hand the populations of Alternaria and Fusarium were higher on conventional than on organic grown berries. The population density of yeasts did not differ between strawberries from organic and conventional production. In a study of the mango phyllosphere microbial community Jager et al. (2001) reported a reduction in the population density of yeasts due to fungicide applications. However, in the study of Jager et al. (2001) leaves directly exposed to fungicides were included, whereas in the present study mature strawberries were included, which may explain the contrasting results. The population density of bacteria was significantly higher in organically grown strawberries obtained from strawberry washing. Similarly, Jager et al. (2001) showed that pesticide applications reduced the bacterial population density in a mango phyllosphere. Interestingly, strains of Staphylococcus were only recovered from the organic nurseries, though in low frequency (6–8%). Staphylococcus has also been reported from partially processed conventional lettuce (Magnuson et al., 1990). The experiments performed in the present project do not provide any explanation for this difference in occurrence of Staphylococcus strains between conventional and organic strawberry growers. Bacteria with known biocontrol traits against plant pathogens were recovered from strawberries in the present study. Among these B. megaterium, P. macerans and P. agglomerans were exclusively recovered from organic nurseries, though only in 1–3 out of the four nurseries and in a moderate frequency (4–27%). The three other biocontrol agents; P. chlororaphis, P. putida and S. liquefaciens were recorded in 5–6 of the 8 nurseries in a moderate frequency (3–33%). Interestingly, for the conventional strawberry producer (grower #4) with high pesticide application levels none of the mentioned biocontrol agents were recovered. In general, it seems that organically produced strawberries harbour a higher amount of bacterial biocontrol agents against plant pathogens, than conventionally produced strawberries, which may be related to the high input of pesticides and/or mineral fertilizers in conventional production systems, but this hypothesis needs to be further tested. 4.6. Conclusions To our knowledge the present study is the first extensive characterization of microbial communities on strawberries covering major groups of phyllosphere microorganisms including culturable filamentous fungi, yeasts and bacteria. We showed that the indigenous communities on field grown mature strawberries are complex including potential plant pathogens, human pathogens, plant disease biocontrol agents and mycotoxin producers. The latter in terms of A. niger and P. expansum can colonise berries and produce high amounts of mycotoxins on berries artificially infected. However, since mouldy strawberries presumably are not consumed directly but discarded during cleaning, the risk of exposure to e.g. fumonisins, ochratoxin A and

B. Jensen et al. / International Journal of Food Microbiology 160 (2013) 313–322

patulin is minimal. On the other hand, strawberries for processing, e.g. jams, may contain mouldy berries as the cleaning step is not by hand on a single berry level. In this case it is obvious that a few mouldy berries can contaminate a whole batch. Furthermore differences in microbial communities on mature conventional and organic strawberries seem to be limited. Acknowledgements We greatly appreciate the involved strawberry growers for their cooperation. We also thank Karin Olesen (University of Copenhagen), Tina Tønnersen (Aarhus University), Lisette Knoth-Nielsen, Hanne Jakobsen and Jesper Mogensen (Technical University of Denmark) for their excellent technical assistance. This work was supported by the Danish Ministry of the Environment, Environmental Protection Agency. Appendix A. Supplementary data Supplementary data to this article can be found online at http:// dx.doi.org/10.1016/j.ijfoodmicro.2012.11.005. References Abanda-Nkpwatt, D., Krimm, U., Screiber, L., Schwab, W., 2006. Dual antagonism of aldehydes and epiphytic bacteria from strawberry leaf surfaces against the pathogenic fungus Botrytis cinerea in vitro. BioControl 51, 279–291. Andersen, B., Hollensted, M., 2008. Metabolite production by different Ulocladium species. International Journal of Food Microbiology 126, 172–179. Andersen, B., Krøger, E., Roberts, R.G., 2002. Chemical and morphological segregation of Alternaria arborescens, A. infectoria and A. tenuissima species-groups. Mycological Research 106, 170–182. Andersen, B., Smedsgaard, J., Frisvad, J.C., 2004. Penicillium expansum: Consistent production of patulin, chaetoglobosins, and other secondary metabolites in culture and their natural occurrence in fruit products. Journal of Agricultural and Food Chemistry 52, 2421–2428. Andrews, J.H., Harris, R.F., 2000. The ecology and biogeography of microorganisms on plant surfaces. Annual Review of Phytopathology 38, 145–180. Arras, G., 1996. Mode of action of an isolate of Candida famata in biological control of Penicillium digitatum in orange fruits. Postharvest Biology and Technology 8, 191–198. Berg, G., Eberl, L., Hartmann, A., 2005. The rhizosphere as a reservoir for opportunistic human pathogenic bacteria. Environmental Microbiology 7, 1673–1685. Bora, T., Özaktan, H., Göre, E., Aslan, E., 2004. Biological control of Fusarium oxysporum f. sp. melonis by wettable powder formulations of the two strains of Pseudomonas putida. Journal of Phytopathology 152, 471–475. Christensen, K.B., van Klink, J.W., Weavers, R.T., Larsen, T.O., Andersen, B., Phipps, R.K., 2005. Novel chemotaxonomic markers for the Alternaria infectoria species-group. Journal of Agricultural and Food Chemistry 53, 9431–9435. Dal Bello, G., Mónaco, C., Rollan, M.C., Lampugnani, G., Arteta, N., Abramoff, C., Ronco, L., Stocco, M., 2008. Biocontrol of postharvest grey mould in tomato by yeasts. Journal of Phytopathology 156, 257–263. Degenkolb, T., Dieckmann, R., Nielsen, K.F., Gräfenhan, T., Theis, C., Zafari, D., Chaverri, P., Ismaiel, A., Brückner, H., von Döhren, H., Thrane, U., Petrini, O., Samuels, G.J., 2008a. The Trichoderma brevicompactum clade: a separate lineage with new species, new peptaibiotics, and mycotoxins. Mycological Progress 7, 177–219. Degenkolb, T., von Döhren, H., Nielsen, K.F., Samuels, G.J., Brückner, H., 2008b. Recent advances and future prospects in peptaibiotics, hydrophobin, and mycotoxin research, and their importance for chemotaxonomy of Trichoderma and Hypocrea. Chemistry & Biodiversity 5, 671–680. Dennis, C., 1976. The microflora of the surface of soft fruits. In: Dickinson, C.H., Preece, T.F. (Eds.), Microbiology of Aerial Plant Surfaces. Academic Press, London - New York – San Francisco, pp. 419–439. Francés, J., Bonaterra, A., Moreno, M.C., Cabrefiga, J., Badosa, E., Montesinos, E., 2006. Pathogen aggressiveness and postharvest biocontrol efficacy in Pantoea agglomerans. Postharvest Biology and Technology 39, 299–307. Frisvad, J.C., Thrane, U., 2002. Mycotoxin production by common filamentous fungi, In: Samson, R.A., Hoekstra, E.S., Frisvad, J.C., Filtenborg, O. (Eds.), Introduction to Foodand Air Borne Fungi, 6th edition. Centraalbureau voor Schimmelcultures, Utrecht, pp. 321–330. Frisvad, J.C., Smedsgaard, J., Samson, R.A., Larsen, T.O., Thrane, U., 2007a. Fumonisin B2 production by Aspergillus niger. Journal of Agricultural and Food Chemistry 55, 9727–9732. Frisvad, J.C., Thrane, U., Samson, R.A., 2007b. Mycotoxin producers. In: Dijksterhuis, J., Samson, R.A. (Eds.), Food mycology. A Multifaceted Approach to Fungi and Food. CRC Press, Boca Raton, pp. 135–159. Jacques, M.A., Morris, C.E., 1995. A review of issues related to the quantification of bacteria from the phyllosphere. FEMS Microbiology Ecology 18, 1–14. Jager, E.S., Wehner, F.C., Korsten, L., 2001. Microbial ecology of the mango phylloplane. Microbial Ecology 42, 201–207.

321

Jennessen, J., Nielsen, K.F., Houbraken, J., Schnürer, J., Lyhne, E.K., Frisvad, J.C., Samson, R.A., 2005. Secondary metabolite and mycotoxin production by the Rhizopus microsporus group. Journal of Agricultural and Food Chemistry 53, 1833–1840. Jensen, B.H., Petersen, A., Christensen, H.B., Andersen, J.H., Hermann, S.S., Poulsen, M.E., Hilbert, G., Grossmann, A., Holm, M., 2010. Pesticidrester i fødevarer 2009 – Resultater fra den danske pesticidkontrol. 978-87-7083-941-9 . (Web) http://www. fodevarestyrelsen.dk/Publikationer/Alle_publikationer/2010/003.htm. (12-22-2011). Johnson, L., Hökeberg, M., Gerhardson, B., 1998. Performance of Pseudomonas chloraraphis biocontrol agent MA 342 against seed-borne diseases in field experiments. European Journal of Plant Pathology 104, 701–711. Kinkel, L.L., 1997. Microbial population dynamics on leaves. Annual Review of Phytopathology 35, 327–347. Knudsen, P.B., Mogensen, J.M., Larsen, T.O., Nielsen, K.F., 2011. Occurrence of fumonisins B2 and B4 in retail raisins. Journal of Agricultural and Food Chemistry 59, 772–776. Krimm, U., Abanda-Nkpwatt, D., Schwab, W., Schreiber, L., 2005. Epiphytic microorganisms on strawberry plants (Fragaria ananassa cv. Elsanta): Identification of bacterial isolates and analysis of their interaction with leaf surfaces. FEMS Microbiology Ecology 53, 483–492. Lackner, G., Möbius, N., Scherlach, K., Partida-Martinez, L.P., Winkler, R., Schmitt, I., Hertweck, C., 2009. Global distribution and evolution of a toxinogenic BurkholderiaRhizopus symbiosis. Applied and Environmental Microbiology 75, 2982–2986. Laengle, T., Strasser, H., 2010. Developing a risk indicator to comparatively assess environmental risks posed by microbial and conventional pest control agents. Biocontrol Science and Technology 20, 659–681. Li, B., Ravnskov, S., Guanlin, X., Larsen, J., 2007. Biocontrol of Pythium damping-off in cucumber by arbuscular mycorrhiza-associated bacteria from the genus Paenibacillus. Biocontrol 52, 863–875. Lix, X.R., 2009. Cryptococcus neoformans: Morphogenesis, infection and evolution. Infection, Genetics and Evolution 9, 401–416. Loeffler, J., Hebart, H., Magga, S., Schmidt, D., Klingspor, L., Tollemar, J., Schumacher, U., Einsele, H., 2000. Identification of rare Candida species and other yeasts by polymerase chain reaction and slot blot hybridization. Diagnostic Microbiology and Infectious Diseases 38, 207–212. Magnuson, J.A., King Jr., A.D., Türük, T., 1990. Microflora of partially processed lettuce. Applied and Environmental Microbiology 56, 3851–3854. Mansfeld-Giese, K., Larsen, J., Bødker, L., 2002. Bacterial populations associated with mycelium of the arbuscular mycorhrhizal fungus Glomus intraradices. FEMS Microbiology Ecology 41, 133–140. Matsukura, H., Katayama, K., Kitano, N., Kobayashi, C., Kanegane, A., Higuchi, A., Kyotani, S., 1996. Infective endocarditis caused by an unusual negative rod, Rahnella aquatilis. Pediatric Cardiology 17, 108–111. McLean, M.A., Sutton, J.C., 1992. Mycoflora on strawberry in Ontario. Canadian Journal of Botany 79, 846–852. Mogensen, J.M., Larsen, T.O., Nielsen, K.F., 2010. Widespread occurrence of the mycotoxin fumonisin B2 in wine. Journal of Agricultural and Food Chemistry 58, 4853–4857. Nielsen, K.F., Gräfenhan, T., Zafari, D., Thrane, U., 2005. Trichothecene production by Trichoderma brevicompactum. Journal of Agricultural and Food Chemistry 53, 8190–8196. Nielsen, K.F., Mogensen, J.M., Johansen, M., Larsen, T.O., Frisvad, J.C., 2009. Review of secondary metabolites and mycotoxins from the Aspergillus niger group. Analytical and Bioanalytical Chemistry 395, 1225–1242. Otto, M., 2009. Staphylococcus epidermis – the accidental pathogen. Nature Review Microbiology 7, 555–567. Parikka, P., Kivijärvi, P., Prokkola, S., Kemppainen, R., 2009. Microbiological quality of organic strawberry fruit. Acta Horticulturae 842, 377–380. Parsley, R., 1996. Operation manual, version 6. Microbial Identification System, MIDI Inc., Newark, Delaware, USA . (219 pp.). Patiño-Vera, M., Jiménez, B., Ortiz, M., Carillo, A., Galindo, E., 2005. Pilot-scale production and liquid formulation of Rhodotorula minuta, a potential biocontrol agent of mango anthracnose. Journal of Applied Microbiology 99, 540–550. Pfaller, M.A., Diekema, D.J., 2004. Twelve years of fluconazole in clinical practice: global trends in species distribution and fluconazole susceptibility of bloodstream isolates of Candida. Clinical Microbiology and Infection 10, 11–23. Ramos, A., Dámaso, D., 2000. Extraintestinal infection due to Hafnia alvei. European Journal of Microbiological Infectious Diseases 19, 708–710. Samson, R.A., Frisvad, J.C., 2004. Penicillium subgenus Penicillium: new taxonomic schemes, mycotoxins and other extrolites. Studies in Mycology 49, 1–257. Samson, R.A., Houbraken, J., Frisvad, J.C., Thrane, U., Andersen, B., 2010. Food and Indoor fungi. CBS-KNAW Fungal Biodiversity Centre, Utrecht . (390 pp.). Samson, R.A., Yilmaz, N., Houbraken, J., Spierenburg, H., Seifert, K.A., Peterson, S.W., Varga, J., Frisvad, J.C., 2011. Phylogeny and nomenclature of the genus Talaromyces and taxa accommodated in Penicillium subgenus Biverticillium. Studies in Mycology 70, 159–183. Sasser, M., 1990. Identification of bacteria through fatty acid analysis. In: Clement, Z., Rudolph, K., Sands, D.C. (Eds.), Methods in Phytobacteriology. Akadémiai Kiadó, Budapest, pp. 199–204. Sharma, R.R., Singh, D., Singh, R., 2009. Biological control of postharvest diseases of fruits and vegetables by microbial antagonists: A review. Biological Control 50, 205–211. Smedsgaard, J., 1997. Micro-scale extraction procedure for standardized screening of fungal metabolite production in cultures. Journal of Chromatography A 760, 264–270. Sørensen, J.L., Phipps, R.K., Nielsen, K.F., Schroers, H.J., Frank, J., Thrane, U., 2009. Analysis of Fusarium avenaceum metabolites produced during wet apple core rot. Journal of Agricultural and Food Chemistry 57, 1632–1639.

322

B. Jensen et al. / International Journal of Food Microbiology 160 (2013) 313–322

Teoh, A.Y.B., Hui, M., Ngo, K.Y., Wong, J., Lee, K.F., Lai, P.B.S., 2006. Fatal septicaemia from Chromobacterium violaceum reports and review of the literature. Hong Kong Medical Journal 12, 228–231. Tournas, V.H., Katsoudas, E., 2005. Mould and yeast flora in fresh berries, grapes and citrus fruits. International Journal of Food Microbiology 105, 11–17. Tronsmo, A., 1986. Trichoderma used as a biocontrol agent against Botrytis cinerea rots on strawberry and apple. Meldinger fra Norges Landbrukshøgskole 65, 1–22.

Weidenbörner, M., Wieczorek, C., Kunz, B., 1995. Mold spectra of various foods in relation to plating medium. Journal of Food Protection 58, 661–665. Whiteman, S.A., Stewart, A., 1998. Suppression of Botrytis cinerea sporulation on irradiated grape leaf tissue by an antagonistic bacterium Serratia liquefaciens. New Zealand Journal of Crop and Horticultural Science 26, 325–330.