Aflatoxins occurrence through the food chain in Costa Rica: Applying the One Health approach to mycotoxin surveillance

Accepted Manuscript Aflatoxins occurrence through the food chain in Costa Rica: Applying the One Health approach to mycotoxin surveillance

Fabio Granados-Chinchilla, Andrea Molina, Guadalupe Chavarría, Margarita Alfaro-Cascante, Diego Bogantes-Ledezma, Adriana Murillo-Williams PII:

S0956-7135(17)30322-5

DOI:

10.1016/j.foodcont.2017.06.023

Reference:

JFCO 5676

To appear in:

Food Control

Received Date:

20 March 2017

Revised Date:

23 May 2017

Accepted Date:

15 June 2017

Please cite this article as: Fabio Granados-Chinchilla, Andrea Molina, Guadalupe Chavarría, Margarita Alfaro-Cascante, Diego Bogantes-Ledezma, Adriana Murillo-Williams, Aflatoxins occurrence through the food chain in Costa Rica: Applying the One Health approach to mycotoxin surveillance, Food Control (2017), doi: 10.1016/j.foodcont.2017.06.023

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT 1 1

Aflatoxins occurrence through the food chain in Costa Rica: Applying the One Health

2

approach to mycotoxin surveillance

3 4

Fabio Granados-Chinchillaa, Andrea Molinaab, Guadalupe Chavarríaa, Margarita Alfaro-

5

Cascantea, Diego Bogantes-Ledezmac, Adriana Murillo-Williamsc

6 7 8 9 10 11 12 13 14 15

aCentro

de Investigación en Nutrición Animal (CINA), Universidad de Costa Rica, 11501-2060 Ciudad Universitaria Rodrigo Facio San José, Costa Rica. bEscuela de Zootecnia, Universidad de Costa Rica, 11501-2060 Ciudad Universitaria Rodrigo Facio San José, Costa Rica. cCentro para Investigaciones en Granos y Semillas (CIGRAS), Universidad de Costa Rica, 11501-2060 Ciudad Universitaria Rodrigo Facio, San José, Costa Rica. Corresponding author: Adriana Murillo-Williams, Centro para Investigaciones en Granos y

16

Semillas (CIGRAS), Universidad de Costa Rica, 11501-2060 Ciudad Universitaria Rodrigo

17

Facio, San José, Costa Rica; Tel: +506 2511 3517 Fax: +506 2511 4346. Email:

18

[email protected]

19 20 21 22

Highlights:  This study applied the One Health approach to the mycotoxin surveillance in Costa Rica

23

 A total of 970 samples of feedstuff and 5493 samples of foodstuff were evaluated

24

 Aflatoxin prevalence was higher for feedstuff (24.0%) than foodstuff (10.8%)

25

 The highest aflatoxin prevalence in feed occurred in corn ingredients, dog and dairy

26 27

cow feed  The highest aflatoxin prevalence in food occurred in corn, peanut, and red beans

ACCEPTED MANUSCRIPT 2 28

Abstract

29

Aflatoxins (AFs) are toxic metabolites produced by Aspergillus spp. and commonly found in

30

crops, grains, feedstuff, and forages. Exposure to AFs has been associated with increased risk of

31

liver cancer and growth retardation in humans, liver damage, immunosuppression,

32

embryotoxicity in both animals and humans, and decreased milk, egg and meat production in

33

animals. For the first time, the Costa Rican national mycotoxin surveillance programs for animal

34

feed and food are considered as a whole, applying the One Health approach to the mycotoxin

35

epidemiological research. Therefore, the aim of this study was to determine the occurrence of

36

AFs in cereals, nuts, grains intended for animal and human consumption in Costa Rica.

37

In animal feed and feed ingredients, 970 samples were analyzed for AFs from 2010 to 2016 with

38

an overall prevalence of positive samples of 24.0 % (ranging from 0.01 to 290 µg kg-1). Only

39

2.5% of the samples failed to comply the regulation for total AFs (20 µg kg-1 feed). From 5493

40

samples of agricultural commodities intended for human consumption analyzed from 2003 to

41

2015, there was an overall prevalence of AF positive samples of 10.8% (ranging from 0.48 to

42

500 µg kg-1), and 2.8% did not comply the regulation for AFs (20 µg kg-1). In both feed and

43

food, the highest AF prevalence corresponded to corn ingredients (27.8%) and white corn

44

(38.6%), respectively. Among the commodities intended for human consumption, red beans had

45

the highest aflatoxin concentrations (500 µg kg-1).

46 47 48

Keywords: Aflatoxins; Cereals; Foods; Feedstuffs; One Health Approach; Food Chain

ACCEPTED MANUSCRIPT 3 50

1. Introduction

51

Mycotoxins are secondary metabolites produced by fungi, mainly saprophytic, that can affect

52

crops in the field, during harvest, and storage. Aflatoxins (AFs) are mycotoxins classified as

53

furanocoumarins, produced by Aspergillus flavus and A. parasiticus (CAST, 2003). Aspergillus

54

flavus has a worldwide distribution and produces aflatoxin B1 (AFB1) and aflatoxin B2 (AFB2).

55

Aspergillus parasiticus produces aflatoxins B1, B2, G1 (AFG1) and G2 (AFG2). Aflatoxins have

56

been classified as human carcinogens (De Ruyck, De Boevre, Huybrechts, & De Saeger, 2015;

57

IARC, 2015), associated with children stunting (Wu, 2013), hepatotoxic for animals and humans

58

(Hgindu, Johnson, & Kenya, 1982), genotoxic, immunotoxic, and responsible for growth

59

retardation and decreased production in animals (Coulombe, Guarisco, Klein, & Hall, 2005;

60

Grace, 2013; Stoev, 2015). Within the aflatoxin group, AFB1 is the furthermost fraction found in

61

food and it has the highest genotoxic and carcinogenic potential (De Ruyck, De Boevre,

62

Huybrechts, & De Saeger, 2015). Furthermore, aflatoxin M1 (AFM1), the primary

63

monohydroxylated derivative of AFB1, may be present in milk from animals exposed to AFB1

64

contaminated feed (EFSA, 2007; Marín, Ramos, Cano-Sancho, & Sanchis, 2013).

65

Incidence of Aspergillus infection and the concomitant contamination with AFs can occur in a

66

wide variety of products and byproducts intended for animal and human consumption (Stoev,

67

2015). Such ingredients include corn, rice, peanut, sorghum, wheat, and soybean. Additional

68

feed ingredients commonly used in Costa Rica that could also serve as a substrate for the growth

69

of aflatoxigenic fungi include cassava, citrus pulp, banana peel, pineapple shells, and oil palm

70

seeds.

71

Crops can be contaminated with AFs in the field, at harvest or during the postharvest stages. In

72

the field, high-temperature stress and drought conditions after Aspergillus infection trigger AFs

ACCEPTED MANUSCRIPT 4 73

accumulation (Kebede, Abbas, Fisher, & Bellaloui, 2012). During storage, the rate and degree of

74

contamination depend on different factors such as temperature, humidity, water activity,

75

concurrent mycobiota, insect damage, and grain physical injury (EFSA, 2007).

76

The innocuity of cereal grain-based products for animals and humans should be ensured during

77

processing and throughout the entire food chain using the “farm to fork” models (Yazar &

78

Omurtag, 2008). Aflatoxins are very stable and may resist commonly used food processing

79

techniques like roasting, extrusion, baking, and cooking. For this reason, AFs represent a threat

80

to human and animal health worldwide, and maximum limits (ML) for AFs in food and feed

81

have been established in most countries (García & Heredia, 2014). In 1999, the Costa Rican

82

Ministry of Health set a ML of total AFs of 15 µg kg-1 for peanut; and for corn, rice, beans,

83

wheat, oilseeds, legumes, and other cereals a ML of 20 µg kg-1. The ML for AFs for feed and

84

feed ingredients was set at 20 µg kg-1 feed.

85

The concept of One Health recognizes the interconnections between, human, animal and

86

environmental health (Zinsstag, Waltner-Toews, & Tanner, 2015). Under the One Health

87

concept, this interdisciplinary epidemiological study brings together the national surveillance

88

program for animal feed coordinated by the Ministry of Agriculture and Livestock, and the

89

monitoring scheme for agricultural commodities intended for human consumption. Furthermore,

90

information on the incidence of AFs in feed and staple foods in Latin America is scarce, and it is

91

required to estimate the level of exposure of the population to AFs. Therefore, the aim of this

92

study was to determine the occurrence of AFs in agricultural commodities and products intended

93

for animals and for human consumption. Hence, the One Health approach will be applied for the

94

first time to the mycotoxin surveillance in Costa Rica. The results will improve our

ACCEPTED MANUSCRIPT 5 95

understanding of the mycotoxin problem in the country and can be used as a tool for decision-

96

making aimed to counteract mycotoxin exposure for both animals and humans.

97

2. Materials and methods

98

2.1. Samples

99

2.1.1. Animal feed and feed ingredient samples

100

Aflatoxin determinations were conducted in the Microbiology Laboratory of CINA, University

101

of Costa Rica. A total of 970 feedstuffs samples of ca. 5 kg were collected from hay (n =

102

322/970; 33.3%), dairy cow feed (n = 246/970; 25.4%), citrus pulp (n = 40/970; 4.1%), whole

103

corn (n = 36/970; 3.7%), dried distillers grains with solubles (DDGS; n = 36/970; 3.7%), calf

104

feed (n = 36/970; 3.7%), and different kinds of forages (n = 31/970; 3.2%), during 2010 to 2016

105

by government inspectors in Costa Rican feed manufacturers, as part of a countrywide

106

surveillance program. Selection of feed and feed ingredients to be tested, number of samples, and

107

sampling sites were chosen by feed control officials, taking into account the most common

108

feedstuff used in Costa Rica, import and export regulations, contamination risk factors,

109

productivity of the feed industry, and the risk for human and animal health associated with each

110

feed or feed ingredient. Sampling was performed following the Association of American Feed

111

Control Officials (AAFCO) recommendations for mycotoxin test object collection (AAFCO,

112

2017), and samples were taken from silos and storage reservoirs from feed manufacturing plants.

113

All samples were quartered and sieved (1 mm particle size). Fresh material (e.g. forages) was

114

dried at 60°C before it was processed.

115

2.1.2. Food commodities for human consumption

116

A total of 5493 food and agricultural commodities samples intended for human consumption

117

were analyzed during 2003-2015 in the Mycotoxin Laboratory of CIGRAS, University of Costa

ACCEPTED MANUSCRIPT 6 118

Rica. The majority of samples corresponded to the most commonly imported commodities for

119

human consumption in Costa Rica (i.e. rice, maize, peanuts, beans, wheat). Sampling was

120

conducted by the State Phytosanitary Service officials in grain shipments at the Pacific Seaport,

121

the Atlantic Seaport, and the Nicaragua border, and sent to CIGRAS for analysis. Other samples

122

analyzed corresponded to products sent by farmers, and the food industry to the Mycotoxin

123

Laboratory for quality control purposes.

124

2.2. Aflatoxin analysis

125

2.2.1 Animal feed and feed ingredients

126

From 2010 to 2011, samples were analyzed by Enzyme-Linked Immunosorbent Assay (ELISA),

127

and from 2012 to 2016 by High Performance Liquid Chromatography (HPLC).

128

Reagents. An analytical standard with a certified concentration of 2.0 μg mL-1, dissolved in

129

acetonitrile, was purchased from Trilogy® Analytical Laboratory Inc. Linear calibration curves

130

ranging from 0.004 to 0.04 µg mL-1 were prepared during quantification. Additionally, a

131

naturally contaminated reference material (TR-MT100, cornmeal, 17.4 µg kg-1 of total AFs) was

132

used as a quality control sample (TS-108, Washington, MO, USA). Potassium iodide and

133

metallic iodine (ACS grade), acetonitrile (ACN) and methanol (MeOH, chromatographic grade)

134

were purchased from J.T. Baker (Avantor Materials, PA, USA).

135

ELISA determinations. A (20.0 ± 0.1) g subsample was used for testing to which 100 mL of an

136

80:20 MeOH and H2O solution were added. Measurements were performed according to the

137

ELISA kit manufacturer (AgraQuant® Aflatoxin, Romer Labs®, Getzersdorf, Austria) which has

138

a quantitation range from 1 to 20 µg L-1. Briefly, 100 µL of the methanolic extract, dilution or

139

standard was mixed with 200 µL of conjugate directly in dilution microtitre wells. A 100 µL

140

aliquot of this mixture was added to antibody linked wells and incubated for 15 min. Afterward,

ACCEPTED MANUSCRIPT 7 141

100 µL of the substrate were incorporated, and the mixture was left to stand for 5 min at standar

142

temperature and pressure. Finally, 100 µL of stop solution was added to the mixture. Absorbance

143

measurements were performed immediately using two simultaneous wavelength (450 nm and

144

620 nm) using a SynergyTM Biotek HT microplate reader and the Gen 5TM software (BioTek

145

Instruments Inc., Winooski, VT, USA).

146

HPLC determinations. Aflatoxin analysis was performed using a modified ISO/IEC 17025

147

accredited version of the AOAC method 2003.02. Several modifications were included to span

148

the analysis for other feed and feed ingredients. Briefly, toxin fractions were obtained using an

149

isocratic high-performance liquid chromatography method. Equipment consisted of an Agilent 1

150

260 Infinity series HPLC with a quaternary pump (G1311B), a column compartment (G1316A)

151

kept at 42°C, a fluorescence detector (G1321B) and an autosampler system (G1329A) set to

152

inject 20 μL (Agilent Technologies, Santa Clara, CA). Peak separation was accomplished using a

153

5 μm Agilent Zorbax Eclipse C18 column (3.0 mm×150 mm). The mobile phase was set at a

154

flow rate of 0.8 mL min−1 and consisted of H2O (Type I, TOC 2 µg L-1, 0.055 µS cm-1), MeOH

155

and ACN 60:30:5. Fluorescent derivatives of AFB1, AFB2, AFG1, and AFG2 were generated with

156

an aqueous 1.2 mmol L-1 KI and 0.79 mmol L-1 I2 solution at a flow rate of 0.3 mL min−1 at 95°C

157

using a 0.15 mL reactor on a Pinnacle PCX system (Pickering Laboratories, Mountain View,

158

CA, USA). These derivatives emit light at 435 nm and after excitation at 365 nm.

159

A representative (25.0 ± 0.1) g subsample was used for extraction, 100 mL of an aqueous

160

acetonitrile solution (60 mL/100 mL ACN) was added to the sample. The mixture was forced

161

into contact and homogenized using a digital Ultra-turrax® at 18 000 rpm (T25, IKA® Werke

162

GmbH & Co. KG, Staufen im Breisgau, Germany) during 1-3 min. The supernatant was

163

removed and filtered by gravity through a Whatman® 541 ashless filters (GE Health Life

ACCEPTED MANUSCRIPT 8 164

Sciences Little Chalfont, Buckinghamshire, United Kingdom). A representative aliquot of 4 mL

165

was diluted to a total volume of 50 mL with phosphate saline buffer. The whole volume was

166

passed through an immunoaffinity column (IAC) (EASI EXTRACT Aflatoxin, R-biopharm,

167

Darmstadt, Germany) using a SPE 12 port vacuum manifold (Supelco, VisiprepTM, Bellefonte,

168

PA, USA) operating at 15 mm Hg (ca. 0.55 mL per minute). Finally, 3 mL of methanol were

169

used to elute analytes. The total volume recovered was concentrated ten fold under vacuum at 60

170

°C (Centrivap, LABCONCO, Kansas City, MO, USA) before injection.

171

2.2.2 Food and food commodities intended for human consumption

172

Data from samples destined for human consumption during 2003-2010 were obtained by the

173

American Association of Cereal Chemists (AACC) Method 45-15.01 (AACC, 2010) for total

174

AFs (sum of AFB1, AFB2, AFG1, and AFG2) with confirmation by thin layer chromatography

175

(TLC), and a limit of detection (LoD) of 2 µg kg-1 of AFs. From 2010 to 2015, total AFs were

176

determined by fluorometry using AflaTest® (VICAM®, Milford, MA, USA) IAC with a LoD of

177

0.48 µg kg-1 of total AFs, following the methods suggested by VICAM in the AflaTest®

178

Instruction Manual. Samples were ground and passed through a No. 20 sieve. For total aflatoxin

179

extraction, a 25 g sample was blended at high speed with a 70:30 MeOH: H2O solution. After

180

filtration with a 24 cm, Whatman N° 1-2V filter paper, an aliquot of the filtrate was diluted,

181

filtrated with an 11 cm Whatman No. 934-AH microfiber filter and passed through an AflaTest®

182

IAC. Aflatoxins were eluted from the column with methanol (HPLC grade), collected in a

183

cuvette, mixed with a bromine developer solution and placed in a Series 4 VICAM fluorometer

184

(VICAM, Milford, MA, USA) for total aflatoxin measurements.

185

Statistical analysis. Analyses of variance with posthoc Tukey tests were performed to

186

demonstrate differences in total aflatoxin prevalence across time. Statistical analysis performed

ACCEPTED MANUSCRIPT 9 187

using IBM PAWS Statistics 22 (SPSS, Inc., Armonk, NY). The coefficient of determination (r)

188

was used to corroborate association between aflatoxin concentrations and meteorological data. A

189

value of r ~ 0 was deemed as a lack of correlation. To assess a possible relationship between the

190

aflatoxin levels and a particular period of the year, Pearson’s product moment correlation was

191

performed. This data was evaluated using Sigmaplot 12.0 software (Systat Software Inc., San

192

Jose, CA).

193

For commodities and food ingredients intended for human consumption, a separate statistical

194

analysis was conducted with the most frequently analyzed grains: white and yellow corn, black

195

and red beans, and peanut. Although 2421 samples of milled rice were analyzed, this cereal was

196

not included in the analysis since most samples contained no measurable amounts of AFs. Data

197

analysis was conducted with PROC GLM of SAS Studio University Edition (SAS®). For all

198

statistical data, p values < 0.05 were considered significant.

199

3. Results and Discussion

200

3.1. Aflatoxins in animal feed and feed ingredients

201

Overall, relatively few samples exceeded the regulatory aflatoxin limit established by Costa

202

Rican and International authorities (prevalence of 2.5% [n = 24/970]). However, when other

203

concentrations (i.e. below 20 µg kg-1) were considered, incidence as high as 24.0% (n = 233/970)

204

was observed (Table 1, Figure 1D). In fact, 16.2% of the samples had aflatoxin concentrations

205

below the 5 µg kg-1 threshold (Table 1). That may represent a potential risk for animal health

206

depending on the animal species or the amount ingested. However, most samples exhibited AFs

207

levels below the detection limits (i.e. 0.01 and 3 µg kg-1 [76.0%]) for the methods. Interestingly,

208

from the four aflatoxin fractions, AFB1 was commonly found at concentrations that surpassed the

ACCEPTED MANUSCRIPT 10 209

FDA/EC regulatory limit (Figure 1 D). Studies suggest that a 20 µg AFB1 kg-1feed permissible

210

level prevents acute adverse health effects in dairy cattle and other ruminants (EFSA, 2004).

211

Since the HPLC-FLD was used to assess in-feed mycotoxin, maximum aflatoxin concentrations

212

ranged from 65.9 µg kg-1 in 2014 to 86.8 µg kg-1 in 2013, corresponding to the AFB1 fraction

213

(Table 1 and Figure 1D). In the case of the remaining shares, the highest concentrations observed

214

corresponded to 28.6, 26.0 and 47.6 µg kg-1 for AFB2, AFG1 and AFG2 respectively (Figure 1D),

215

all reached in 2016. However, as the regulatory threshold is set for total AFs, the sum of

216

individual fractions could exceed such limit. Noteworthy, after government officials set a strict

217

vigilance program in 2013, AFs incidence has significantly decreased from 64.5% in 2013 to

218

8.54% in 2015, and 10.1% in 2016 (Table 1), notwithstanding an increase in sampling frequency

219

(Table 1).

220

There was no evidence of a direct correlation between aflatoxin concentration in animal feedstuff

221

produced or stored (imported products) in Costa Rica, and the average rain precipitation (mx = -

222

1.72; r = 0.142), number of rainy days for a specific month (mx = -4.36; r = 0.142), mean

223

temperature (mx = 0.24; r = 0.112), and relative humidity (mx = -4.98; r = 0.164) during the same

224

period in Costa Rica. Accordingly, there is no clear trend on the time of the year in which AFs

225

levels may rise (Figure 1 A-D). Individual evaluation of the association between each of the

226

climate parameters and corn ingredients [r(106), imported], and dairy cow feed [r(244), of local

227

production], also indicated a lack of association; Pearson 0.231-0.296 and 0.450-0.488,

228

respectively with p < 0.001. Evidence suggests that just based on the overall weather data it may

229

be difficult to predict when the peaks of fungal contamination or toxin production take place

230

during production or storage (Medina, Rodríguez, & Magan, 2015), which, in turn makes the

231

application of possible control measures more difficult.

ACCEPTED MANUSCRIPT 11 232

Two important feed ingredient that showed relatively elevated AFs incidence and toxin

233

concentrations were ground and whole corn (n = 8/20; 40.0% and n = 9/36; 25.0%) (Table 2).

234

Additionally, when all sources of corn listed in Table 2 (i.e. corn gluten, white corn, corn meal,

235

DDGS, and whole corn), a total 27.8% (n = 30/108) prevalence was obtained. Prevalence values

236

from 20 to 25% in corn products have been reported in other studies in Latin America (Mendes

237

de Souza et al., 2013; Rodrigues & Naehrer, 2012). These values are relevant since just in 2014

238

Costa Rica imported over 172.4 million USD in corn products (PROCOMER, 2014), and most

239

of the animal feed produced in Costa Rica is corn-based. The use of corn germplasm not adapted

240

to local conditions (Fountain et al., 2014), open-pollinated varieties which may be more

241

susceptible to fungal contamination (Warburton & Williams, 2014), and physically damaged

242

kernels (Echandi, 1986) are factors that may lead to AFs contamination and may have been used

243

for feedstuffs. Furthermore, elevated levels of contamination can be achieved with inadequate

244

management of kernels during handling, transport, and storage. We contend that control efforts

245

must have a special focus on corn and corn products to minimize contamination along the food

246

chain and to be more cost-effective.

247

From the methodological standpoint, ELISA determinations suffered from a limited

248

responsiveness when compared with the more accurate and sensitive HPLC-FLD analysis (300

249

fold) which can quantify at the ng kg-1 level. Since LoD for the ELISA method was 3 µg kg-1,

250

some samples may have had AFs below this level, and therefore, these data must be used

251

conservatively.

252

Noteworthy, from 2010 to 2012, independently of the analytical method used, a stationary state

253

in toxin prevalence was attained (Table 1). Since no statistically significant differences (p < 0.05)

ACCEPTED MANUSCRIPT 12 254

were observed among these years, this could be an indication of improved management practices

255

in the country during this time frame.

256

On the other hand, though no unit operation is completely effective in decontaminating cereals,

257

such mechanical treatments such as conventional dry milling (Pietri, Zanetti, & Bertuzzi, 2009)

258

and dehulling (Siwela, Siwela, Matindi, Dube, & Nziramasanga, 2005) have been found to

259

reduce total aflatoxin concentrations to some extent. These facts may also explain more elevated

260

levels of AF in by-product based feeds (e.g. corn germ, DDGS) than in grains used for food

261

production. In fact, Pietri and coworkers found a significant percentage of AFs in corn germ after

262

milling (Pietri, Zanetti, & Bertuzzi, 2009). In this scenario, AFs in contaminated corn germ used

263

for animal feed may re-enter the food chain when AFB1 is metabolized to AFM1 and secreted

264

through milk, posing a health hazard to human consumers. The recurrent contamination found in

265

compound feed and feedingstuffs is reflected in the high prevalence of AFM1 (n = 44/70, 62.8%)

266

found in commercial milk sampled in Costa Rica in the years 2013 through 2014 (Chavarría,

267

Granados-Chinchilla, Alfaro-Cascante, & Molina, 2015). In this scenario, quantifiable data of

268

contaminants such as AFs is increasingly relevant because it allows an estimation of the possible

269

impact on several species along the food chain. For example, a Pearson’s Square used to

270

formulate a feed may result in 5 parts of soybean meal (64.8 kg; 29.2 kg crude protein) and 30

271

parts corn meal (388.7 kg; 38.9 kg crude protein), for a total of 453 kg (30.8 kg of crude protein)

272

feed requirement for a cow. In a worst-case scenario where corn meal contaminated with 290.4

273

µg kg-1 of aflatoxin is used, a 4 kg total daily feed intake for a single cow will result in an

274

undesired exposure of 996.7 µg of aflatoxin (i.e. 4 kg feed containing 85.8% corn meal times the

275

aflatoxin concentration in the raw ingredient). Considering an average body weight of an adult

276

cow of 589.7 kg, the daily exposure is 1.66 µg aflatoxin per live body weight. Furthermore, a

ACCEPTED MANUSCRIPT 13 277

carry-over from feed to milk of 6.2% (EFSA, 2004) would result under these conditions in 18.0

278

µg L-1 of AFM1. A calf with and average weigth of 27 kg would drink 3 L of milk per day,

279

divided among five feedings. The resulting AFM1 exposure would be of 2 µg kg-1 live body

280

weight. Although these concentrations may seem elevated, evidence shows that clinical signs in

281

cattle occur after exposure to concentrations as high as 1.5 mg kg-1 to 2.23 mg kg-1 feed, and in

282

small ruminants, > 50 mg kg-1 feed (Miller & Wilson, 1994). Still, long-term exposure to

283

relatively low concentrations of AFs may result in health issues for mammals, especially those

284

that are more susceptive to AFs (Zain, 2011). More importantly, this contamination may very

285

well continue downstream the rest of the food chain, particularly through processed milk and

286

milk products for human consumption.

287

However, in Costa Rica, cattle feeding is based in forage (n = 31, < 0.01 µg kg-1) and hay (n =

288

25/322, 7.8% prevalence) (Table 2). In this study, aflatoxin incidence was low in both feeding

289

types (Table 2), except for 2016, when AFs prevalence in hay increased to 12.4%. Interestingly,

290

samples with the highest aflatoxin concentration were collected in October [0.09 to 77.68 µg kg-

291

1]

292

= 1/25, 4.0% prevalence) showed similar values as those reported in a previous survey made in

293

South America were the authors found AFs in 8 % of the tested samples (Rodrigues & Naehrer,

294

2012). Aflatoxin levels below our method sensitivity in silage are significant since an adequate

295

ensiling process leads to anaerobic conditions and low pH, conditions that guarantee the non-

296

survival of aflatoxigenic fungi. However, other acid-tolerant and microaerophilic species (e.g.

297

Penicillum roqueforti) may be able to produce toxins such as mycophenolic acid or roquefortin C

298

(Malekinejad, Afzali, Mohammadi, & Sarir, 2015) under these conditions. In this scenario, a

299

possible silage contamination could stem from the seasonal scarcity or low supply of other feed

(Figure 1), including soybean meal. The highest prevalence AFs in soybean meal observed (n

ACCEPTED MANUSCRIPT 14 300

ingredients, thus prompting a hasty and meager silage production. Hence, unless deteriorated,

301

silage analysis can be focused on these aforementioned toxins. Another important result is the

302

low AF contamination level found in dog food, however, with high prevalence values (n = 7/14,

303

50.0%; Table 2). Although the number of dog food samples analyzed was low, and conclusions

304

cannot be drawn, aflatoxin contaminated corn-based products have been linked to poisoning in

305

dogs (Wouters et al., 2013).

306

3.2. Aflatoxins in food commodities for human consumption

307

A total of 5493 samples were analyzed for total AFs in the 13-year period. The vast majority of

308

samples corresponded to imported agricultural goods sent to CIGRAS by the phytosanitary

309

authorities as part of the national monitoring program. Since 2009, an increase in the total

310

number of samples analyzed per year was observed (Table 3). This trend can be explained by

311

the number of milled rice samples analyzed that went from 0.7% of the total number of samples

312

in 2009 (n=1/142), to 72.4 % (n=661/912), 71.6% (n=615/859), and 58.2% (n=438/753) for

313

2013, 2014, and 2015, respectively. The lowest number of samples analyzed occurred in 2007-

314

2008 (Table 3) during the World Food Crisis. However, it is not known if this was due to the

315

decrease in grain imports, a reduction in the quality-control measures, or both.

316

Rice and beans are a staple food in Costa Rica, with an average per capita consumption of 49 kg

317

year-1 and 10.5 kg year-1, respectively (FAO, 2016). However, to meet the domestic consumption

318

42% of the rice is imported, as well as 70% of the beans, 100% of the yellow corn, and 78% of

319

the white corn (SEPSA 2008; SEPSA, 2014). Accordingly, during the 13-year period, 49.8 % of

320

the total number of samples corresponded to rice (sum of milled and paddy rice, n=2740/5493),

321

followed by yellow corn (n=832/5493, 15%), peanut (n=572/5493, 10.5%), white corn

322

(n=453/5493, 8.2%), and beans (sum of red and black beans, n=274/5493, 5%) (Table 4).

ACCEPTED MANUSCRIPT 15 323

Regarding white corn and yellow corn, there was not a defined trend in the total number of

324

samples analyzed, however, for both red and black beans, there has been a reduction in the

325

number of samples analyzed since 2010. Most of the beans consumed in Costa Rica are imported

326

from other countries in Latin America and China (IICA, 2014). Therefore, the reduction in the

327

number of samples could be an indication of little or no aflatoxin monitoring in this important

328

staple food, a matter of concern since both red and black beans may have AFs contamination

329

above the national regulation (Table 4).

330

From the total number of samples, 10.8% had measurable amounts of AFs (Table 3), and just

331

2.8% had aflatoxin concentrations above the national ML (Table 3). In 2015, the average AFs

332

contamination of yellow and white corn samples that did not meet the national regulation was

333

199 µg kg-1 (n=8/80, [24-410] µg AFs kg-1), and 215 µg kg-1(n=7/43, [55-420] µg AFs kg-1),

334

respectively. These values are considered high since the maximum AFs contamination values

335

observed in white corn in 2014 and 2013 were 28 µg kg-1 and 62 µg kg-1, respectively; and 10 µg

336

kg-1 in 2014 and 16 µg kg-1 in 2013 for yellow corn. Reports of corn contamination with AFs are

337

commonly found in the literature, and recent reports from Zimbabwe (Hove et al. 2016), Brasil

338

(Oliveira, Rocha, Sulyok, Krska, & Mallmann, 2017), and Vietnam (Huong et al., 2016) confirm

339

that contamination levels can be variable and depend on environmental conditions during the

340

pre-harvest or post-harvest stages (Cotty & Jaime-García, 2007). However, high aflatoxin

341

concentrations in 2015 could not be associated with any climatic event in exporter countries.

342

A total of 44 different food and food ingredients were analyzed during the 13-year period. The

343

analyzed samples included grains, seeds, flour, grain byproducts, and condiments. However, as

344

shown in Table 4, 59.1% (n=26/44) of the goods had no measurable amounts of AFs, or very low

345

AFs prevalence, for example, wheat (1,3%), and milled rice (1.4%) (Table 4). Within the group

ACCEPTED MANUSCRIPT 16 346

of commonly analyzed grains, the highest AFs prevalence corresponded to white corn (38.6 %),

347

red beans (37%), and peanut (21.9%) (Table 4), and the highest number of samples with AFs

348

above the national ML corresponded to white corn (n=56), peanut (n=45), and red beans (n=35)

349

(Table 4).

350

Aflatoxins were more prevalent in white corn than in yellow corn and any other grain with

351

n≥100 (Table 4). Interestingly, the highest incidence of aflatoxin-contaminated white and yellow

352

corn was observed in 2013, with 69.6% (n=62/89) and 35% (n=19/54) of the samples with

353

measurable levels of this metabolite, respectively. Weather and host plant conditions are not

354

always optimal for aflatoxin contamination to occur. However, during 2012, the central United

355

States was affected by a severe drought (Umphlett, Timlin, & Fuchs, 2012), conditions that led

356

to higher incidence of aflatoxin contamination of the harvested crop in some states (Umphlett,

357

Timlin, & Fuchs, 2012). Therefore, the high frequency of contaminated white corn observed in

358

2013 could correspond to drought-hit corn from the 2012 US harvest.

359

Peanut samples were frequently contaminated with AFs, and 45 samples had AFs above the ML

360

(Table 4). Costa Rica depends on peanut imports since local production covers approximately

361

10% to 15% of the total demand (FAO, 2016). One of the most important problems concerning

362

peanut production worldwide is aflatoxin contamination. In this survey, results show the regular

363

occurrence of AFs in peanuts, in accordance with reports frequently found in the literature (Bhat

364

& Reddy, 2017; Chen, Liao, Lin, Chiueh, & Shih, 2013; Iqbal, Asi, Zuber, Akram, & Batool,

365

2013; Udomkun et al. 2017; Wu et al., 2016). Currently, aflatoxin contamination of peanut and

366

peanut products generates alerts in several countries that may lead to border rejections and the

367

removal of this product from the markets (RASFF, 2015; RASFF, 2016). The high frequency of

ACCEPTED MANUSCRIPT 17 368

contaminated peanut samples should be an indication for Costa Rican authorities to maintain the

369

monitoring programs.

370

Among the frequently contaminated grains, the maximum aflatoxin concentration observed

371

during the 13-year period corresponded to red beans (500 µg kg-1), followed by white corn (420

372

µg kg-1), and peanut (400 µg kg-1) (Table 4). Aflatoxin contamination of beans has not been

373

considered a significant problem. There are few reports about the presence of toxigenic (Freitas-

374

Costa & Scussel, 2002) or potentially toxigenic Aspergillus strains in beans (Domijan et al.,

375

2005; Tseng, Tu, & Tzean, 1995). Nevertheless, aflatoxin contamination has been reported at

376

variable levels in beans. For black beans, aflatoxin prevalence values of 95% and 75% have been

377

reported in Costa Rica (Echandi, 1986), and Brazil (Jager, Tedesco, Souto, & Oliveira, 2013),

378

respectively. High AFs concentrations have been documented also for black beans in Costa Rica

379

(Echandi, 1986), and Brazil (Scussel & Baratto, 1994), however, the reported AFs values were

380

below the 500 µg kg-1 of AFs level observed in red beans in the represent study. In addition to

381

the high AFs values observed in beans, the reduction in the number of samples analyzed should

382

be of concern. Beans could represent a primary source of AFs entering the food chain in Costa

383

Rica, Latin America, and other countries where this grain is a staple food, consequently,

384

aflatoxin monitoring in imported and locally grown beans should be compulsory.

385

Low aflatoxin contamination levels were observed in milled rice, with just one sample with AFs

386

concentration exceeding the ML (Table 4). The low aflatoxin contamination of milled rice has

387

been previously documented. Average AFB1 contamination levels reported by Reddy, Reddy, &

388

Muralidharan (2009) ranged from 0.5 to 3.5 µg kg-1. Mean levels below 5 µg kg-1 of AFs have

389

also been reported in China (Lai, Liu, Ruan, Zhang, & Liu, 2015; Liu, Gao, & Yu, 2006), Korea

390

(Park, Kim, & Kim, 2004), and Taiwan (Chen, Hsu, Wang, & Chien, 2016). Contrastingly, in

ACCEPTED MANUSCRIPT 18 391

this study, 8 samples of paddy rice had total aflatoxin concentration above 20 µg kg-1, with a

392

maximum concentration of 69 µg kg-1 (Table 4). High aflatoxin levels in paddy rice have been

393

documented and summarized in the review by Sempere Ferre (2016). The processes of rice

394

milling can lead to an aflatoxin contamination reduction in white rice (Castells, Ramos, Sanchis,

395

& Marín, 2007; Reddy, Reddy, & Muralidharan, 2009). However, high aflatoxin levels have

396

been reported in the bran fraction after the aflatoxin-contaminated rice has been milled (Castells,

397

Ramos, Sanchis, & Marín, 2007; Prietto et al., 2015; Trucksess, Abbas, Weaver, & Shier, 2011).

398

Accordingly, the two rice semolina samples analyzed during the 13 year period (Table 4)

399

originated from the paddy rice sample with the highest contamination (Table 4). The rice bran is

400

widely used by the food and animal feed industry (Friedman, 2013). Therefore, milling

401

contaminated rice samples could lead to highly contaminated byproducts entering the animal and

402

human food chain, in addition to an increased risk of milling equipment contamination, and

403

worker exposure to the toxins. The majority of the rice samples analyzed in this study

404

corresponded to imported milled rice, however, rice can be exposed to highly toxigenic

405

Aspergillus strains in the field (Abbas, Weaver, Zablotowicz, Horn, & Shier, 2005), and

406

aflatoxin contamination of the grain can occur. Therefore, locally grown rice should also be

407

assessed for AFs contamination to ensure it also meets the national ML.

408

3.3. Perspectives on Aflatoxin Prevalence in Feed and Food

409

When the overall results of AF prevalence in grain intended for animal feed are compared to

410

those for human consumption, it was observed that prevalence was consistently higher in feed

411

and feed products (24.0%) than in food and food products (10.8%). The latter might be an

412

indication of low-quality grain and grain byproducts being used for feed manufacture. Some

413

examples of the rice industry by-products used in animal feed production are rice bran, brewers

ACCEPTED MANUSCRIPT 19 414

rice, ground brown rice, and rice hulls (AAFCO, 2016). The latter highlights the importance of

415

the food industry to ensure the quality of the commodities used, and not relying on processing to

416

reduce AFs contamination, since the byproducts may enter the animal food chain. In the case of

417

agricultural products intended for human consumption, it is of great concern the high AFs

418

concentrations found in white corn, peanuts, and red beans. White corn and peanuts are common

419

substrates for Aspergillus growth and aflatoxin contamination. However, there are just a few

420

reports in the literature that indicate that beans can also have high aflatoxin concentrations.

421

Although the aflatoxin monitoring led to border rejections, it is unknown if the rejected food and

422

food commodities were used for animal feed production, or mixed with other grain lots to reduce

423

the AFs levels. Finally, though pre and postharvest measures are paramount to avoid the risk of

424

contamination in both feeds and foods, therefore, new trends in decontamination of AFs should

425

be considered as complete absence of such toxins is extremely challenging.

426

Although this report only includes information regarding aflatoxin contamination, simultaneous

427

contamination with other toxins should be assessed and contemplated as a whole. For example,

428

aflatoxin and fumonisin co-contamination of commodities have been previously documented. A

429

sample with toxin concentrations below those recommended by legal standards may not be

430

considered per se a risk, however, it could represent a health hazard if other toxins present in the

431

sample are taken into account, since interactions among toxins and synergistic effects may occur

432

(Šegvić Klarić, 2012; Alassane-Kpembi et al., 2013; Alassane-Kpembi, Puel, & Oswald, 2014).

433

4. Conclusions

434

Monitoring programs, that should include local and foreign products, ought to be subject to

435

constant maintenance and improvement. This type of epidemiological data and control is of high

436

importance for countries such as Costa Rica, especially since the mycotoxin working group of

ACCEPTED MANUSCRIPT 20 437

the Institute for Research on Cancer stated in their last report that surveillance data on exposure

438

to AFs in developing countries is scarce and monitoring programs should be a priority (IARC,

439

2015). Data that follows must be available for farmers, researchers and policy makers to take

440

serious steps towards crops, ingredients, and feed/food safeguard. Corn is very susceptible to

441

aflatoxin contamination, and therefore corn and corn products should be tested for AFs before

442

food processing. Aflatoxin-contaminated feed and food increases the burden of human exposure

443

or even affect newborn/young animals.

444

5. Acknowledgements

445

The authors would like to thank Marisol Jiménez, Astrid Leiva Gabriel, Andrea Porras, for their

446

excellent technical assistance during sample analysis. We also thank B.Q. Danilo Alvarado in the

447

Mycotoxin Laboratory at CIGRAS for his technical assistance and conducting the aflatoxin

448

determinations.

449 450

6. References

451 452 453 454 455 456

AACC (American Association of Cereal Chemists). (2010). International Approved Method of Analysis. St. Paul, Minnesota: AACC International. AAFCO (Association of American Feed Control Officials). (2016). Official Publication, Association of American Feed Control Officials. Oxfordshire: AAFCO. AAFCO (Association of American Feed Control Officials). (2017). Feed Inspector’s Manual.

457

Association of American Feed Control Officials. Oxfordshire: AAFCO.Abbas, H.K.,

458

Weaver M.A., Zablotowicz, R.M., Horn, B.W. & Shier, W.T. (2005). Relationships

459

between aflatoxin production and sclerotia formation among isolates of Aspergillus section

460

Flavi from the Mississippi Delta. European Journal of Plant Pathology, 112, 283-287.

ACCEPTED MANUSCRIPT 21 461

Alassane-Kpembi, I., Kolf-Clauw, M., Gauthier, T., Abrami, R., Abiola, F. A., Oswald, I. P., &

462

Puel, O. (2013). New insights into mycotoxin mixtures: The toxicity of low doses of Type B

463

trichothecenes on intestinal epithelial cells is synergistic. Toxicology and Applied

464

Pharmacology, 272(1), 191-198.

465

Alassane-Kpembi, I., Puel, O., & Oswald, I. P. (2014). Toxicological interactions between the

466

mycotoxins deoxynivalenol, nivalenol and their acetylated derivatives in intestinal epithelial

467

cells. Archives of Toxicology, 89(8), 1337-1346.

468

Bhat, R. & Reddy, K.R.N. (2017). Challenges and issues concerning mycotoxins contamination

469

in oil seeds and their edible oils: Updates from last decade. Food Chemistry, 215, 425-37.

470 471 472 473 474

CAST (Council for Agricultural Science and Technology). (2003). Mycotoxins: risks in plant, animal, and human systems. Ames, IA: Task Force Report No. 139. Castells, M., Ramos, A.J., Sanchis, V., & Marín, S. (2007). Distributions of total Aflatoxins in Milled Rice Fractions of Hulled Rice. J. Agric. Food Chemistry, 55, 2760-2764. Chavarría, G., Granados-Chinchilla, F., Alfaro-Cascante, M., & Molina, A. (2015). Detection of

475

aflatoxin M1 in milk, cheese and sour cream samples from Costa Rica using enzyme-

476

assisted extraction and HPLC. Food Additives & Contaminants. Part B, Surveillance, 3210,

477

1-8.

478 479 480

Chen, M.Z., Hsu, Y.H., Wang, T.S., & Chien, S.W. (2016). Mycotoxin monitoring for commercial foodstuffs in Taiwan. Journal of Food and Drug Analysis, 24, 147-156. Chen, Y.C., Liao, C.D., Lin, H.Y., Chiueh, L.C., & Shih, D.Y.C. (2013). Survey of aflatoxin

481

contamination in peanut products in Taiwan from 1997 to 2011. Journal of Food and Drug

482

Analysis, 21, 247-252.

ACCEPTED MANUSCRIPT 22 483 484 485

Cotty, P.J., & Jaime-García, R. (2007) Influences of climate on aflatoxin producing fungi and aflatoxin contamination. International Journal of Food Microbiology, 119, 109-115. Coulombe, R.A., Guarisco, J.A., Klein, P.J., & Hall, J.O. (2005). Chemoprevention of

486

aflatoxicosis in poultry by dietary butylated hydroxytoluene. Animal Feed Science and

487

Technology,121, 217-225.

488

De Ruyck, K., De Boevre, M., Huybrechts, I., & De Saeger, S. (2015). Dietary mycotoxins, co-

489

exposure, and carcinogenesis in humans: Short review. Mutation Research - Reviews in

490

Mutation Research, 766, 32–41.

491

Domijan, A.M., Peraica, M., Lender, V.Z., Cvjetkovic, B., Jurjevic, Z., Topolovec-Pintaric, S., &

492

Ivic, D. (2005). Seed-borne fungi and ochratoxin A contamination of dry beans (Phaseolus

493

vulgaris L.) in the Republic of Croatia. Food and Chemical Toxicology, 43, 427-432.

494

Echandi, R. (1986). The relationship between aflatoxin formation and kernel damage in Costa

495

Rica maize. In M.S. Zuber, E.B. Lillehoj, & B.L. Renfro (Eds.), Aflatoxins in maize: A

496

proceedings of the workshop (pp. 164-171). Mexico. D.F.: CIMMYT.

497

EFSA (European Food Safety Authority).(2004). Panel on Contaminants in the Food Chain.

498

Opinion of the Scientific Panel on contaminants in the food chain related to Aflatoxin B1 as

499

undesirable substance in animal feed. EFSA Journal, 39, 1-27.

500

EFSA (European Food Safety Authority). (2007). Opinion of the Scientific Panel on

501

contaminants in the food chain related to the potential increase of consumer health risk by a

502

possible increase of the existing maximum levels for aflatoxins in almonds, hazelnuts and

503

pistachios and derived products. EFSA Journal, 446, 1-127.

504 505

FAO (Food and Agriculture Organization of the United Nations). (2016). FAOSTAT, Statistic Division. Available online: http://www.fao.org/faostat/en/#data

ACCEPTED MANUSCRIPT 23 506

Fountain, J. C., Scully, B. T., Ni, X., Kemerait, R. C., Lee, R. D., Chen, Z. Y., & Guo, B. (2014).

507

Environmental influences on maize-Aspergillus flavus interactions and aflatoxin production.

508

Frontiers in Microbiology, 5, 40.

509

Freitas-Costa, L.L. & Scussel, V. M. (2002). Toxigenic fungi in beans (Phaseolus vulgaris L.)

510

classes black and color cultivated in the state of Santa Catarina, Brazil. Brazilian Journal

511

of Microbiology, 33, 138-144.

512

Friedman, M. (2013). Rice brans, rice bran oils, and rice hulls: composition, food and industrial

513

uses, and bioactivities in humans, animals and cells. Journal of Agricultural and Food

514

Chemistry, 61,10626-10641.

515

García, S. & Heredia, N.L. (2014). Aflatoxins: An overview. In J.B. Dixon, A.L. Barrientos

516

Velázquez, & Y. Deng (Eds.), Aflatoxin Control: Safeguarding Animal Feed with Calcium

517

Smectite (pp. 1-10). Madison: American Society of Agronomy and Soil Science Society of

518

America.

519

Grace, D. (2013). Animals and Aflatoxins. In L. Unnevehr, & D. Grace (Eds.), Finding Solutions

520

for Improved Food Safety. Focus brief 20 (5) (pp.1-2). Washington DC: International Food

521

Policy Research Institute.

522 523 524

Hgindu, A., Johnson, B.A., & Kenya, P.R. (1982). An outbreak of acute hepatitis by aflatoxin poisoning in Kenya. Lancet, 319, 1346-1348. Hove, M., De Boevre, M., Lachat, C., Jacxsens, L., Nyanga, L.K., & De Saege, S. (2016).

525

Occurrence and risk assessment of mycotoxins in subsistence farmed maize from

526

Zimbabwe. Food Control, 69, 36-44.

ACCEPTED MANUSCRIPT 24 527

Huong, B.T.M., Le, D.T., Tran, T.D., Madsen, H., Brimer, L., & Dalsgaard, A. (2016).

528

Aflatoxins and fumonisins in rice and maize staple cereals in Northern Vietnam and dietary

529

exposure in different ethnic groups. Food Control, 70, 191-200.

530

IARC (International Agency for Research on Cancer). (2015). Mycotoxin control in low- and

531

middle-income countries. In C. P. Wild, J. D. Miller, & J. D. Groopman (Eds.), IARC

532

Working Group Report No. 9. Lyon (FR): International Agency for Research on Cancer.

533

IARC Monographs on Evaluation of Carcinogenic Risks to Humans, 82, 1-556.

534

IICA (Instituto Interamericano de Cooperación para la Agricultura). (2014). Las cadenas de

535

valor de maíz blanco y frijol en Centroamérica: actores, problemas y acciones para su

536

competitividad. San José, Costa Rica: Red SICTA, Cooperación Suiza en América Central.

537

127 pp. Available online:

538

http://www.iica.int/sites/default/files/publications/files/2015/B3427e.pdf

539

Iqbal, S.Z., Asi, M.R., Zuber, M., Akram, N., & Batool, N. (2013). Aflatoxins contamination in

540

peanut and peanut products commercially available in retail markets of Punjab, Pakistan.

541

Food Control, 32, 83-86.

542 543

Jager, A.V., Tedesco, M.P., Souto, P.C.M.C. & Oliveira, C.A.F. (2013). Assessment of aflatoxin intake in Sao Paulo, Brazil. Food Control, 33, 87-92.

544

Kebede, H., Abbas, H. K., Fisher, D. K., & Bellaloui, N. (2012). Relationship between aflatoxin

545

contamination and physiological responses of corn plants under drought and heat stress.

546

Toxins, 4 (11), 1385–1403.

547 548

Lai, X., Liu, R., Ruan, C., Zhang, H., & Liu, C. (2015). Occurrence of aflatoxins and ochratoxin A in rice samples from six provinces in China. Food Control, 50, 401-404.

ACCEPTED MANUSCRIPT 25 549 550 551

Liu, Z., Gao, J., & Yu, J. 2006. Aflatoxins in stored maize and rice grains in Liaoning Province, China. Journal of Stored Products Research, 42, 468-479. Malekinejad, P., Afzali, N., Mohammadi, A., & Sarir, H. (2015). Effects of combination of

552

different levels sodium bentonite and silybummarinum seeds on performance and carcass

553

traits of broiler chicks fed diet contaminated with aflatoxin B1 in starter and grower period.

554

Journal of Applied Environmental and Biological Sciences, 5(12), 269-275.

555 556 557 558 559

Marín, S., Ramos, A. J., Cano-Sancho, G., & Sanchis, V. (2013). Mycotoxins: Occurrence, toxicology, and exposure assessment. Food and Chemical Toxicology, 60, 218-237. Medina, A., Rodríguez, A., & Magan, N. (2015). Effect of climate change on Aspergillus flavus and aflatoxin B1 production. Frontiers in Microbiology, 5, 1-7. Mendes de Souza, M.L., Sulyok, M., Freitas-Silva, O., Soares Costa, S., Brabet, C., Machinski

560

Junior, M., Leiko Sekiyama, B., Azevedo Vargas, E., Krska, R., & Schuhmacher, R. (2013).

561

Cooccurrence of Mycotoxins in Maize and Poultry Feeds from Brazil by Liquid

562

Chromatography/Tandem Mass Spectrometry. The Scientific World Journal, 2013:427369

563

(9 pp). http://dx.doi.org/10.1155/2013/427369

564

Miller, D.M., & Wilson, D.M. (1994). Veterinary diseases related to aflatoxins. In D.L. Eaton,

565

and J.D. Groopman (Eds.), The Toxicology of Aflatoxins: Human Health, Veterinary and

566

Agricultural Significance (pp. 347-364). New York: Academic Press.

567

Oliveira, M. S., Rocha, A., Sulyok, M., Krska, R., & Mallmann, C.A. (2017). Natural mycotoxin

568

contamination of maize (Zea mays L.) in the South region of Brazil, Food Control, Part B,

569

73, 127-132.

570 571

Park, J.W., Kim, E.K., & Kim, Y.B. (2004). Estimation of the daily exposure of Koreans to aflatoxin B1 through food consumption. Food Additives and Contaminants, 21, 70-75.

ACCEPTED MANUSCRIPT 26 572

Pietri, A., Zanetti, M., & Bertuzzi, T. (2009). Distribution of aflatoxins and fumonisins in dry-

573

milled maize fractions. Food Additives & Contaminants: Part A, 26(3), 372-380.

574

Prietto, L., Silveira Moraes, P., Basso Kraus, R., Meneghetti, V., Alves Fagundes, C.A., &

575

Badiale Furlong, E. (2015). Post-harvest operations and aflatoxin levels in rice (Oryza

576

sativa). Crop Protection, 78, 172-177.

577

PROCOMER (Promotora del Comercio Exterior de Costa Rica). (2014). Estadísticas de

578

Comercio Exterior de Costa Rica. San José, Costa Rica: PROCOMER, 258. Available

579

online: http://www.procomer.com/uploads/downloads/anuario-estadistico-2014.pdf

580

RASFF (The Rapid Alert System for Food and Feed). (2015). RASFF for safer food. The Rapid

581

Alert System for Food and Feed 2014. Directorate general for health and consumer

582

protection. Brussels: European Commission. 49 p.

583

RASFF (The Rapid Alert System for Food and Feed). (2016). The Rapid Alert System for Food

584

and Feed 2015 Annual Report. Directorate general for health and consumer protection.

585

Brussels: European Commission. 45 p.

586 587 588 589 590

Reddy, K.R.N., Reddy, C.S., & Muralidharan, K. 2009. Detection of Aspergillus spp. and aflatoxins B1 in rice in India. Food Microbiology, 26, 27-31. Rodrigues, I., & Naehrer, K. (2012). A three-year survey on the worldwide occurrence of mycotoxins in feedstuffs and feed. Toxins, 4(9):663-675. Scussel, V.M. & Baratto, W.R. (1994). Levels of aflatoxins in grains from Santa Catarina State,

591

Southern Brazil. In Proceedings of the 6th international working conference on stored-

592

product protection (p. 1051-1053). Wallingford, UK.

593 594

Šegvić Klarić, M. (2012). Adverse effects of combined mycotoxins. Arhiv Za Higijenu Rada I Toksikologiju, 63(4), 519-530.

ACCEPTED MANUSCRIPT 27 595 596

Sempere Ferre, F. (2016). Worldwide occurrence of mycotoxins in rice. Food Control, 62, 291298.

597

SEPSA (Secretaría Ejecutiva de Planificación Sectorial Agropecuaria). (2008). Plan Nacional

598

de Alimentos Costa Rica: oportunidad para la agricultura nacional. San José, Costa Rica:

599

SEPSA. 84 pp.

600 601 602

SEPSA (Secretaría Ejecutiva de Planificación Sectorial Agropecuaria). (2014). Situación y desafíos de la agricultura costarricense 2009-2012. San José, Costa Rica: SEPSA. 84 pp. Siwela, A. H., Siwela, M., Matindi, G., Dube, S., & Nziramasanga, N. (2005). Decontamination

603

of aflatoxin-contaminated maize by dehulling. Journal of the Science of Food and

604

Agriculture,85(15), 2535-2538.

605

Stoev, S. D. (2015). Foodborne mycotoxicoses, risk assessment and underestimated hazard of

606

masked mycotoxins and joint mycotoxin effects or interaction. Environmental Toxicology

607

and Pharmacology, 39(2), 794-809.

608

Trucksess, M.W., Abbas, H.K., Weaver, C.M., & Shier, W.T. (2011). Distribution of aflatoxins

609

in shelling and milling fraction of naturally contaminated rice. Food Additives and

610

Contaminants: Part A, 28, 1076-1082.

611

Tseng, T.C., Tu, J.C, & Tzean, S.S. (1995). Mycoflora and mycotoxins in dry bean (Phaseolus

612

vulgaris) produced in Taiwan and in Ontario, Canada. Botanical Bulletin of Academia

613

Sinica 36, 229-234.

614

Udomkun, P., Nimo Wiredu, A., Nagle, M., Bandyopadhyay, R., Müller, J., & Vanlauwe, B.

615

(2017). Mycotoxins in Sub-Saharan Africa: Present situation, socio-economic impact,

616

awareness, and outlook. Food Control, Part A, 72, 110-122.

ACCEPTED MANUSCRIPT 28 617

Umphlett, N.A., Timlin, M.S., & Fuchs, B.A. (2012). Regional Drought Perspective. In B.A.

618

Fuchs, D.A. Wood, & D. Ebbeka (Eds.), From Too Much to Too Little: How the central

619

U.S. drought of 2012 evolved out of one of the most devastating floods on record in 2011

620

(p. 2). Central U.S. 2012 Drought Assessment. Available online:

621

http://drought.unl.edu/Portals/0/docs/CentralUSDroughtAssessment2012.pdf

622 623 624

Warburton, M. L., & Williams, W. P. (2014). Aflatoxin resistance in maize: what have we learned lately? Advances in Botany, Article ID 352831, 10 pp. doi:10.1155/2014/352831 Wouters, A. T., Casagrande, R. A., Wouters, F., Watanabe, T. T., Boabaid, F. M., Cruz, C. E., &

625

Driemeier, D. (2013). An outbreak of aflatoxin poisoning in dogs associated with aflatoxin

626

B1-contaminated maize products. Journal of Veterinary and Diagnostic Investigation, 25,

627

282-287.

628

Wu, F. (2013). Aflatoxin Exposure and Chronic Human Disease: Estimates of Burden Disease.

629

In L. Unnevehr, & D. Grace (Eds.), Finding Solutions for Improved Food Safety. Focus

630

brief 20 (3) (pp.1-2). Washington DC: International Food Policy Research Institute.

631

Wu, L.X., Ding, X.X., Li, P.W., Du, X.H., Zhou, H.Y., Bai, Y. Z., & Zhang, L.X. (2016).

632

Aflatoxin contamination of peanuts at harvest in China from 2010 to 2013 and its

633

relationship with climatic conditions. Food Control, 60, 117-123.

634 635 636 637

Yazar, S. & Omurtag, G.Z. (2008). Fumonisins, trichothecenes and zearalenone in cereals. International Journal of Molecular Science, 9, 2062-2090. Zain, M.E. (2011). Impact of mycotoxins on humans and animals. Journal of Saudi Chemical Society, 15, 129-144.

ACCEPTED MANUSCRIPT 29 638

Zinsstag, J., Waltner-Toews, D., & Tanner, M. (2015). Theoretical Issues of One Health. In J.

639

Zinsstag et al. (Eds), One Health: the theory and practice of integrated health approaches.

640

(pp. 16-25). Wallingford: CAB International.

641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660

ACCEPTED MANUSCRIPT 30 661 662 663

Table 1. Prevalence of aflatoxin contamination in agricultural commodities intended animal consumption from 2010 to 2016 in Costa Rica. Year*

664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682

Number of samples

Prevalence (%)**

Maximum aflatoxin concentration (µg kg-1)

Concentration range (µg kg-1) Total x < LoD x<5 5 ≤ x < 10 10 ≤ x < x ≥ 20 20 2010 55 38 7 2 5 3 31.0x 36.7 2011 77 48 5 11 10 3 37.7x 290.4 2012 103 72 26 0 0 5 30.1x 72.1 2013 110 39 53 13 2 3 64.5y 86.8 2014 174 132 36 4 0 2 24.1x 65.9 2015 164 150 12 0 0 2 8.5z 69.3 2016 287 258 18 3 2 6 10.1z 77.7 Overall*** 970 737 157 33 19 24 x=24.0 (76.0%)u (16.2%)v (3.4)w (2.0%)w (2.5%)w *Samples from 2010 and 2011 were analyzed with ELISA assays with a LoD of 3 µg kg-1. The rest of the results were obtained by HPLC analysis with a LoD of 0.01 µg kg-1. **Prevalence is defined as the number of samples with aflatoxin concentration above the LoD for the method. ***Rows/columns with the same superscript do not differ significantly, p < 0.05.

ACCEPTED MANUSCRIPT 31 683 684 685

Table 2. Prevalence of aflatoxin contamination in agricultural commodities intended for animal consumption from 2010-2016 in Costa Rica. Matrix

Number of samples Total 108 2 1 20

686

Concentration range (µg kg-1) < LoD > 20 78 10 12 3

Maximum concentration (µg kg-1)

Corn ingredients Corn. gluten meal Corn. whole white Corn. ground Dried Distillers Grains with 31 22 5 Solubles (DDGS) Corn. whole 36 27 2 Corn. grits 6 5 0 Corn. cracked 12 12 0 Compound feed 370 286 5 Dog food 14 7 0 Fish feed (snapper [n = 2]. trout/salmon [n = 1] and tilapia 22 14 0 [n = 19]) Swine feed 15 11 0 Horse feed 8 6 0 Dairy cow feed 246 188 5 Calf feed 30 26 0 Fiber supplement feed 9 8 0 Rodent/Laboratory Animals 4 4 0 feed Cattle feed 6 6 0 Shrimp feed 3 3 0 Goat feed 2 2 0 Rabbit feed 2 2 0 Poultry feed 9 9 0 Other feed ingredients 150 142 2 Cocoa beans 2 0 0 Rice bran 3 2 0 Palm kernel cake meal 13 11 2 Soybean meal 25 24 0 Citrus pulp 40 38 0 Chamomile flowers 22 22 0 (Matricaria chamomilla ) Wheat middlings 11 11 0 Cassava meal 13 13 0 Pineapple by-products 12 12 0 Banana peels 5 5 0 Orange seeds and peels 1 1 0 Pineapple peels 1 1 0 Soybean hulls 1 1 0 Rice by-product fractions 1 1 0 Silages and Hay 330 305 6 Hay 322 297 6 African bermuda grass silage 3 3 0 Corn silage 2 2 0 Citrus pulp silage 1 1 0 Digitgrass silage 1 1 0 Sorghum silage 1 1 0 Forages 31 31 0 Digit grass (Digitaria eriantha) 15 15 0 King grass (Pennisetum 8 8 0 purpureum /P. typhoides) African bermuda grass (Cynodon 3 3 0 nlemfuensis ) Mombasa Guinea Grass 2 2 0 (Panicum maximum) Sorghum (Sorghum bicolor) 2 2 0 Cassava 1 1 0 *Prevalence is defined as the number of samples with aflatoxin concentration above the LoD for the method.

Prevalence* (%)

290.4 7.1 1.34 290.4

27.8 100.0 100.0 40.0

86.8

29.0

72.14 5.94 0 86.7 3.9

25.0 16.7 0.0 22.7 50.0

6.41

36.4

1.43 17.8 86.7 5.9 2.1

26.7 25.0 23.6 13.3 11.1

0

0.0

0 0 0 0 0 20.7 0.8 12.7 20.7 0.9 0.1

0.0 0.0 0.0 0.0 0.0 5.3 100 33.3 15.4 4.0 5.0

0

0.0

0 0 0 0 0 0 0 0 77.7 77.7 0 0 0 0 0 0 0

0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 7.6 7.8 0.0 0.0 0.0 0.0 0.0 0.0 0.0

0

0.0

0

0.0

0

0.0

0 0

0.0 0.0

ACCEPTED MANUSCRIPT 32 687 688 689

Table 3. Prevalence of aflatoxin contamination in agricultural products intended for human consumption from 2003 to 2015 in Costa Rica. Year*

2003* 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Overall

690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705

Number of samples

Total

x < LoD

435 437 364 316 134 90 142 186 295 570 912 859 753 5493

414 424 351 310 106 83 127 128 250 494 800 805 675 4967 (90.4%)

Concentration range (µg kg-1) x<5 5 ≤ x < 10 10 ≤ x < 20 0 0 2 0 1 0 1 5 20 35 53 40 46 203 (3.7%)

6 2 2 1 4 1 8 9 10 13 23 6 12 97 (1.8%)

5 4 1 2 5 3 1 11 4 12 17 3 3 71 (1.3%)

Prevalence (%)**

Maximum aflatoxin concentration (µg kg-1)

4.8 3.0 3.6 1.9 20.1 7.7 10.5 31.1 15.2 13.3 12.3 6.3 10.3

400 54 350 46 500 20 100 150 230 360 350 150 420

x ≥ 20 10 7 8 3 17 3 5 33 11 16 19 5 17 154 (2.8%)

x =10.8

Samples from 2003 through 2010 were analyzed using the AACC method 45-15.01 with a LoD of 2 µg kg-1. From 2011 and forward, samples were analyzed using AflaTest immunoaffinity columns and a fluorometric method with a LoD of 0.48 µg kg-1. **Prevalence is defined as the number of samples with aflatoxin concentration above the LoD for the method. *

ACCEPTED MANUSCRIPT 33 706 707 708

Table 4. Prevalence of aflatoxin contamination in agricultural commodities intended for human consumption from 2003 to 2015 in Costa Rica. Matrix

709 710 711 712 713 714 715 716 717

Number of samples

Maximum concentration (µg kg-1)

Concentration range (µg kg-1) Total < LoD > 20 Semolina, corn 2 0 1 Semolina, rice 2 0 2 Cassava peel 1 0 0 Broken peanut 2 1 1 White corn 453 278 56 Red beans 164 102 35 Corn. unidentified 44 32 0 Citrus pulp 4 3 0 Pistachio 9 7 1 Peanut 572 447 39 (45**) Corn flour 6 5 1 Black beans 110 103 2 Paddy rice 319 294 8 Macadamia nuts 78 72 0 Yellow corn 832 770 9 Almonds 65 62 0 Milled rice 2421 2386 1 Wheat 234 231 0 Nutmeg 40 40 0 Sunflower seed 25 25 0 Cocoa 14 14 0 Cashews 11 11 0 Oats 10 10 0 Soybeans 10 10 0 Sorghum 9 9 0 Nutmeg mace 9 9 0 Chili peppers 8 8 0 Hazelnuts 7 7 0 Mixed seeds 8 8 0 Polenta 3 3 0 Rice flour 2 2 0 Powdered cinnamon 2 2 0 Soybean flour 2 2 0 Malt 2 2 0 Peanut butter 2 2 0 Linseed seed 2 2 0 Nutmeg seed shells 2 2 0 Cocoa butter 1 1 0 Oil palm kernel flour 1 1 0 Sesame seeds 1 1 0 Chia seed 1 1 0 Pumpkin seed 1 1 0 Cocoa liquor 1 1 0 Dried plums 1 1 0 *Prevalence is defined as the number of samples with aflatoxin concentration above the LoD for the method. ** Number of samples with total AFs above 15 µg kg-1, the maximum limit established for peanut.

370 140 10 80 420 500 8.9 2.4 230 400 110 80 69 11 410 8.9 28.5 1.7 -

Prevalence* (%)

100 100 100 50 38.6 37 27.3 25 22.2 21.8 16.6 8 7.8 7.7 7.5 4.6 1.4 1.3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

ACCEPTED MANUSCRIPT 34

718 719 720 721 722 723 724 725

Figure 1. Bubble plots representing climatic parameters for Costa Rica from 2012 to 2016, and aflatoxins concentrations in animal feed. A. Precipitation; B. Number of rainy days; C. Mean temperatures; D. Seasonal distribution and levels of the four different fractions of aflatoxins present in animal feed on a yearly basis. Every sample, from 2012 to 2016, was considered. Red line represents FDA/EC ML. Symbology: AFB1 , AFB2 , AFG1 , AFG2 .