DON Occurrence in Grains: A North American Perspective

Introduction ABSTRACT Fungi associated with production and In agricultural commodities, the occurrence of deoxynivalenol (DON) has been reported all storage of grains in North America in- over the world, with levels varying among grain types and years of production. The grain supply clude the genera Fusarium, Penicillium, chain, including growers, buyers, and end users, have effectively managed DON with strategies and Aspergillus. Mycotoxins are second- to control this issue systematically. The safety of consumers is ensured through use of these man- ary metabolites of these filamentous fun- agement strategies. This is observed in this review of the North American systems. This article gi that can cause illness in humans and describes the occurrence and management of DON in North America, which is accomplished animals through absorption, ingestion, by 1) a review of the toxicological effects of DON; 2) a review of publically available data and or inhalation (2). Since grains, in general, introduction of new information regarding the occurrence of DON in wheat, maize, and barley in North America, including variability due to growing regions, grain varieties, and year of provide an ideal substratum (medium) production; 3) an overview of industry practices to reduce DON contamination from field for mold growth and mycotoxin pro- through milling when necessary; 4) a review of how all in the value chain, including growers, duction, close attention must be paid to buyers, and end users, have effectively managed DON for more than 20 years; 5) a description their food and feed safety. Additionally, of current maximum limits associated with DON; and 6) the economic impact of any potential grains and grain-based products repre- changes in international regulations. This article focuses on wheat, maize, and barley grown in sent one of the major sources of carbo- Canada and the USA, as these two countries are the major exporters of these grains in North hydrates for humans and livestock (3); America (1). therefore, their safety is of primary con- cern. The prevalence of mycotoxins in grains AUTHORS J. David Miller, Department of is usually associated with the occurrence Andreia Bianchini, The Food Chemistry, Carleton University, of the causal organisms and related causal Processing Center, Food Science and Ottawa, ON, Canada factors, such as temperature and moisture, Technology Department, University of W. Thomas Shier, Department of in both the field and in storage (2,3). De- Nebraska – Lincoln, NE, USA Medicinal Chemistry, University of oxynivalenol, also referred to as DON or vomitoxin, is one of a broad category of Richard Horsley, Department of Plant Minnesota, MN, USA mycotoxins known as trichothecenes. Sciences, North Dakota State Glen Weaver, Ardent Mills, Omaha, DON is produced mainly by F. grami- University, ND, USA NE, USA nearum and F. culmorum, especially in Maia M. Jack, CPGglobal, LLC, USA grains. Production of DON by F. pseudo- PUBLIC REVIEWERS Brent Kobielush, General Mills, MN, graminearum has also been reported in USA Dave Katzke, General Mills, MN, USA warmer climates, although with less fre- Dirk E. Maier, Kansas State University, quency than the major producers (2). Dojin Ryu, Bi-State School of Food F. graminearum is an important plant Science, University of Idaho/Washington Grain Science & Industry, Manhattan, KS, USA pathogen that causes Fusarium head State University, ID, USA blight (FHB) in wheat and barley and ear Sheryl Tittlemier, Grain Research Jim Pestka, Food Science and Human rot in maize. This leads to DON contami- Laboratory, Canadian Grain Nutrition/Microbiology & Molecular nation of these crops during crop growth, Commission, Winnipeg, MB, Canada Genetics, Michigan State University, prior to harvest. DON is found globally in East Lansing MI, USA these crops, as well as in rye, oats, and William W. Wilson, Department of rice (4–6). Agribusiness and Applied Economics, LIAISONS To minimize human exposure to DON North Dakota State University, ND, USA Benjamin Boroughs, NAMA Liaison, via the consumption of contaminated Hamed K. Abbas, USDA-ARS, NBCL, Director of Regulatory and Technical grain-based foods, regulatory organiza- Stoneville, MS, USA Affairs tions have established advisory levels, Susan Abel, Food & Consumer guidelines, and regulations for various Anne R. Bridges, AACCI Liaison, commodities and foods. In North Ameri- Products of Canada, Toronto, ON, Director of Technical Resources Canada ca, the US-FDA has imposed a restric- tion of 1 mg/kg on processed grains (7). Gordon Harrison, Canadian National Health Canada has set regulations of Millers’ Association, Ottawa, ON, 2 mg/kg in uncleaned soft wheat for use Canada in nonstaple foods and 1 mg/kg in un- cleaned soft wheat for use in baby foods; http://dx.doi.org/10.1094/CFW-60-1-0032 however, these regulations are currently ©2015 AACC International, Inc. under review.

32 / JANUARY–FEBRUARY 2015, VOL. 60, NO. 1 Toxicity and Deoxynivalenol (DON) when exposure occurs in single or mul- Although DON has been associated with DON was isolated from moldy barley tiple bolus doses (9). The latter does not a number of acute (8,9,13,25) and chronic by Japanese scientists in 1973 and rede- represent a likely scenario for the vast (8,9) health events, the likelihood of such scribed as “vomitoxin” by USDA research- majority of the global population today occasions remains relatively rare. In the ers in the same year (8–12). Since then, who can readily access minimally con- USA, for instance, advisory levels for grain toxicologists have sought to understand taminated sorted and cleaned grains that containing DON have been in place since the acute and chronic effects that may be very rarely result in acute incidences. 1982 (7). In fact, in 1993, an outbreak of associated with the ingestion of DON. In contrast, as previously mentioned, DON in wheat led to the elimination of a The principal health risks of DON are high chronic exposure has been shown to previously held limit of 2 mg/kg on un- associated with acute dietary exposure, cause growth retardation, alter immune processed raw wheat and wheat by-prod- which is caused by intake of large amounts function, and interfere with reproduc- ucts. After this incident the US-FDA de- of DON within a short time frame. Short- tion and development (8,9). In 2001, the cided to rely on purchasing and cleaning ly after ingestion of heavily contaminated Joint Food and Agriculture Organization/ practices to significantly reduce DON grains, DON causes vomiting and feed World Health Organization (FAO/WHO) levels to the 1 mg/kg maximum for fin- refusal in animals, especially swine, and Expert Committee on Food Additives ished wheat products, including the fol- causes gastroenteritis with vomiting in (JECFA) proposed a provisional maxi- lowing: flour, germ, and bran (7). Since humans (13–15). To date there is no evi- mum tolerable daily intake (PMTDI) for then, the US-FDA continues to recognize dence of occurrence of adverse human DON of 1 µg/kg body weight (bw) based an advisory limit of 1 mg/kg on finished health outcomes in North America as- on the growth effects associated with DON food products containing wheat and wheat sociated with acute dietary intake of DON. exposure (8,27,28). In the following de- by-products. A level that remains protec- The chronic effects of dietary intake of cade, investigations into the mechanisms tive for the USA population and USA ex- DON include weight gain suppression, of the effect of DON on weight gain were port markets. The same is true for advi- anorexia, and altered nutritional efficiency pursued and understood (8,9,29–32). In sory levels on finished grain established (16). It has also been shown to adversely 2010, JECFA extended the PMTDI to ap- by regulatory agencies elsewhere. affect immune systems (17). With any ply to both DON and its acetylated de- toxicological risk assessment, it is impor- rivatives and concluded that mean esti- DON Occurrence in North America tant to understand the absorption, distri- mates of national exposure to DON were The body of available research indicates bution, metabolism, and elimination of a below the PMTDI of 1 µg/kg bw (8,27,28). that acceptably low DON levels in unpro- given toxin or toxicant (18). Unlike other Likewise, recent national assessments de- cessed grain products can be achieved in toxins, such as dioxin and other fat-soluble termined that exposures to DON were exports and shipments for domestic use in compounds, DON is water soluble, which below levels of concern. Based on these most crop years. These levels are achieved allows it to be rapidly cleared in vivo. In collective assessments, the current global through the aggregation of grain stocks rodents and swine, which are frequently limits on DON are not only protective and blending that occurs along the supply used to study the adverse effects of DON from a chronic ingestion standpoint, but chain, monitoring of DON levels in deliv- (9), this toxin and its metabolites are ab- affirm that a level of 1 mg/kg on semipro- eries of grain shipments to processors, and sorbed and excreted quite rapidly (8). Spe- cessed grains is safe, and levels proposed a variety of grain sorting and cleaning cifically, studies have shown that 25% and for unprocessed raw grain would have no technologies available to processors. The 64% of radiolabeled, orally administered benefit from a public health safety per- North American grain industry’s experi- doses of DON and known metabolites in spective, as described further below. ence suggests that adoption of unduly rats were detected within 96 hr in urine and feces, respectively (8,19). Subsequent analyses of the tissues in these rats showed that no radioactivity was retained in the tissues after 96 hr (8,19). Similarly, in pigs DON is rapidly excreted, with a plasma half-life of 3.9 hr and no significant reten- tion in the tissues (8,19–23). Acute exposure to DON is most notably characterized by emesis (8,9). A few out- breaks of grossly contaminated grain in Japan and Korea (1940s–1960s) (24), China (1984–1991) (25), and India (1988) (26) have been described, with the symptoms of the disease being typical of illnesses associ- ated with DON exposure. However, current global regulations, better crop management techniques, improved resistance to FHB, and the advancement of milling practices have considerably reduced the number of acute incidences of DON-related illnesses. Additionally, animal studies have shown that DON most notably exerts its effects Fig. 1. Schematic of the North American grain-handling supply chain.

CEREAL FOODS WORLD / 33 restrictive maximum limits for DON in titation (LOQ) of the methods used standard deviations in Figure 2, DON unprocessed grains could actually hinder, (0.3–0.5 mg/kg). concentrations in individual samples in a rather than assist in, the management of To obtain each of the samples repre- given year do vary and can exceed 2 mg/kg. DON levels in foods and feeds, while con- sented in Figure 2, the sampling process For example, in 2014 the average level of tributing to significantly higher costs to suggested by the USDA Grain Inspection, DON was 0.85 mg/kg, and the levels of the entire supply chain. Packers and Stockyards Administration DON measured in individual samples Wheat. Due to its importance as a food- (GIPSA) for drawing a representative varied from LOD to 20.0 mg/kg. This stuff and as an exported grain, the bulk of sample from a lot of grain was followed. variation can be seen more clearly in Fig- DON monitoring and surveillance data Toxin levels were determined by enzyme- ure 3, where DON levels at the processor for grains pertains to wheat. DON is ob- linked immunoassay (ELISA). The meth- level in soft and hard wheat harvested in served in wheat at various stages of the od of choice (manufacturer and test type) the Midwestern and Northeastern regions North American grain-handling chain, a varied from one milling facility to an- of the USA from the 2003 to 2014 seasons general schematic of which is shown in other; however, all of them used GIPSA- are shown in more detail. Depending on Figure 1. approved methods (e.g., Neogen 2/3, the year and region, up to 14.3% of hard Wheat quality, including the occur- 5/5, Q+) with LOQ varying from 0.3 to wheat samples showed levels of DON rence of DON, is closely monitored 0.5 mg/kg. In this data set, reported val- above 2 mg/kg, while up to 62% of soft throughout the North American grain- ues below the limit of detection (LOD) wheat samples showed levels that high. handling chain. For example, at the early of the method used were assigned a nu- Overall, during the period surveyed 1.7% to middle stages of the handling chain, merical value of “zero” to calculate aver- of the hard wheat samples analyzed had DON is determined in soft and hard age values. For those reported values that levels of DON above 2 mg/kg, while more wheat delivered to USA wheat milling were below the LOQ but above the LOD, than 30% of the soft wheat samples had facilities by truck or railcar. In Figure 2, a numerical value of LOQ/2 was assigned levels of DON that exceeded 2 mg/kg. data from more than 42,100 samples to calculate average values. This data set These variations are discussed further representing wheat harvested throughout was used to produce Figures 2, 3, 7, 13, later in this article. 12 consecutive seasons (2003–2014) from and 14. The occurrence of DON in wheat taken various production areas in the USA and The annual average levels of DON at earlier stages in the handling chain has shipped to USA milling facilities are found in USA wheat delivered to milling also been observed in Canadian wheat summarized. Among the samples ana- facilities were low and clustered around (Fig. 4) and durum (Fig. 5). DON was lyzed, 81.1% were below the limit of quan- 0.5 mg/kg. However, as illustrated by the measured in 72% and 68% of wheat and durum samples, respectively, taken on farm and in 58% and 79% of wheat and durum samples, respectively, taken from primary elevator stocks before final blending of grain to meet quality speci- fications was performed. At later stages of the handling chain, <50% of wheat and durum samples contained quantifiable concentrations of DON. There were some instances of high concentrations of DON (i.e., >5 mg/kg) observed in individual Canadian samples, as with USA samples (Figs. 2 and 3). These instances occurred Fig. 2. Annual average deoxynivalenol (DON) concentrations in wheat across all USA growing re- in samples taken from on farm, primary gions and classes delivered to milling facilities from the 2003 through 2014 crop years (n = 42,131). Whiskers represent one standard deviation for the DON detected in each year. “Estimated Tonnes” elevator stocks, or elevator loadings but lists the amount of wheat represented by the samples analyzed each year. not at later stages in the grain-handling chain. The average DON concentrations in exported Canadian wheat and durum were both <1 mg/kg. The data incorporated into Figure 4 represents almost 13 million tonnes of wheat at the farm stage and more than 15 million tonnes exported at the end of the handling chain from 2009 through 2014, which encompasses years with higher and lower incidences of FHB (33). The data used to build Figure 4 included samples from primary elevator stock (n = 3,281), elevator loading (n = 572), Fig. 3. Levels of deoxynivalenol (DON), at the processor level, in soft and hard wheat harvested in and terminal stock (n = 323) stages that the Midwestern and Northeastern regions of the USA from the 2003 through 2014 seasons. Whis- were analyzed using ELISA or quantita- kers represent one standard deviation for the DON detected in each year. “Estimated Tonnes” lists tive lateral flow test methods, with an the amount of wheat represented by the samples analyzed each year. LOD of 0.5 mg/kg. A small number of

34 / JANUARY–FEBRUARY 2015, VOL. 60, NO. 1 on-farm samples (n = 16) were analyzed using ELISA with an LOD of 0.25 mg/kg; the remaining samples (n = 29,254) were analyzed using ELISA with an LOD of 0.5 mg/kg. Shipments were analyzed using ELISA (n = 130; LOD of 0.5 mg/kg) and gas chromatography–mass spectrom- etry (n = 499; LOQ of 0.05 mg/kg) (34). The data presented in Figure 5 repre- sent durum analyzed in 2010 through 2014, mostly using ELISA or quantitative lateral flow test methods with an LOD of 0.5 mg/kg to evaluate samples at the farm (n = 2,297), primary elevator stock (n = 120), elevator loading (n = 554), terminal stock (n = 496), and shipment (n = 407) stages. Some on-farm samples (n = 53) were analyzed using an ELISA with an LOD of 0.25 mg/kg. In addition, some shipment samples (n = 68) were ana- lyzed using gas chromatography–mass Fig. 4. Deoxynivalenol (DON) occurrence in Canadian wheat along the grain-handling chain. Amount of grain represented for Terminal Elevator Stocks is not available. spectrometry (LOQ of 0.05 mg/kg) (34). It is very difficult to directly compare the degree of DON occurrence in North American grains based on reports of DON in grains grown in other grain- producing regions, as different sampling schemes, sample preparation techniques, sample types, analytical methods, and production years all affect the concen- trations of DON observed. Unfortunate- ly, these details regarding samples, their provenance, and methods of analysis are often not provided in the scientific litera- ture or are not provided in enough detail. However, in general, average reported concentrations for DON in wheat, maize, and barley grown in North America are similar to those reported for these grains grown in other regions. A broad compilation of data on the global occurrence of DON in grains is Fig. 5. Deoxynivalenol (DON) occurrence in Canadian durum along the grain-handling chain. provided in Figure 6 (35–67). In this fig- Amount of grain represented for Terminal Stocks is not available. ure the average values for reported DON concentrations in different commodities throughout the world are presented, as reported by more than 30 scientific pa- pers published in the literature. These reports included samples from more than 30 countries in different regions of the world, such as Africa, the Americas, Asia, the Middle East, Australia, and Europe. The work in these reports ana- lyzed samples mostly described as “field samples” or recently harvested grain, with a small percentage of the samples being described as “commercial samples.” The majority of the studies also indicated that the sampling followed an opportunistic approach, with very few being part of a comprehensive monitoring program for Fig. 6. Reported occurrence of deoxynivalenol (DON) in various cereal grains grown around the DON in grains. This may result in data world (35–67).

CEREAL FOODS WORLD / 35 being presented that are not truly repre- varies based on the use of Good Agricul- of DON in wheat, based on the data set sentative of the occurrence of DON in tural Practices (GAPs), grain type and provided by milling facilities in the USA, grain from a particular region or time cultivar grown, toxigenicity of species are reported for two growing regions period. Nonetheless, even though the and fungal strains associated with the (Midwest and Northeast) and two wheat method of analysis varied from study to crop, and weather (temperature and pre- classes (soft and hard) for different years study and included both rapid screening cipitation). (2003–2014). As noted by the bars in the and comprehensive instrumental tech- chart, the Northeast is more prone to niques such as ELISA and liquid chroma- Effects of Growing Seasons, Regions, variation from year to year, regardless of tography–tandem mass spectrometry, with and Years the type of wheat grown, while the Mid- LODs varying from 0.01 to 180 µg/kg, The presence of DON in wheat is gov- west seems to show more variation for the averages presented in Figure 6 illus- erned by a number of factors that can certain classes. Because mycotoxin pro- trate that, among the commodities de- vary among growing years and regions. duction by molds is driven by intrinsic scribed, the levels of DON reported are These factors include environmental and extrinsic factors, the environmental similar. This applies for the different growing conditions such as temperature conditions encountered by the grain in areas of the world, with the exception of and precipitation, wheat type and cultivar the field have a great impact on the myco- some reports of high levels of DON in grown, as well as the species and chemo- toxin levels detected in the grain. Among wheat in Africa. Wheat samples from type (subpopulation) of F. graminearum the intrinsic factors is mold species (or Africa reported in this figure were col- infecting the grain. strains/chemotypes), while the extrinsic lected in Morocco, Tunisia, and Kenya Figure 2 illustrates how DON can vary factors include temperature, water activ- and reported in three independent stud- among growing years. Since wheat-grow- ity, nutrient availability, and chemical ies (40,44,67). Overall, 856 wheat samples ing regions are large in Canada and the agents (68). As a result, it is not uncom- were collected in these three countries, USA, there can also be considerable varia- mon to observe DON levels in grain vary- and most of the contamination seemed tion in the presence of DON within grow- ing greatly from year to year. Higher lev- to be associated with samples from Tuni- ing regions. The annual average DON els of DON in wheat are usually associ- sia (average DON levels of 17.9 mg/kg). concentrations in wheat across all USA ated with excessively wet periods close to More information regarding the occur- growing regions and classes delivered to the flowering stage, which is 6–8 weeks rence of DON in the Republic of South milling facilities peaked at 0.85 mg/kg in prior to harvest. Africa is presented in Box 1. The differ- 2014 and was at its lowest at 0.27 mg/kg Another example of such variation in ences among regions and grains are very in 2008. The effect of harvest year is fur- DON levels with harvesting year is illus- typical of mycotoxin occurrence, as it ther illustrated by Figure 3, where levels trated by the data presented by Martínez

Box 1 – Deoxynivalenol Incidence in Africa In developing nations with many low- to middle-income and high-density populations, grains represent a primary food source. If not controlled properly, grains are predisposed to contamination by several fungi that can produce toxins: for instance, deoxy- nivalenol (DON). Figure 1 in this box shows overall DON levels in maize and wheat as reported in a survey conducted in the Republic of South Africa (RSA) by the Southern African Grain Laboratory (SAGL) from 2009 to 2014. Five harvesting seasons were evaluated, including 36 grain production regions (Fig. 2). Altogether these provinces account for an approximate production of 1.85 million tonnes/season for wheat and 9.5 million tonnes/season for maize. Every growing season, SAGL randomly selects wheat and maize samples for mycotoxin analysis that represent different growing regions as well as different classes and grades. For the period described here, mycotoxin analysis was performed using a validated multimycotoxin screening method (UPLC- MS/MS). The highest incidence of DON in maize was reported in a sample from the 2009–2010 growing season at 1.8 mg/kg; the highest incidence reported in wheat was in a sample from the 2012–2013 growing season at 0.4 mg/kg. Acknowledgments: Information reported is courtesy of the Maize Trust and the Winter Cereal Trust (African industry organizations) via SAGL.

Fig. 1. Mean annual DON incidence (whiskers indicate standard devia- Fig. 2. RSA crop production regions. tion) in maize and wheat in the RSA.

36 / JANUARY–FEBRUARY 2015, VOL. 60, NO. 1 et al. (50). Data collected throughout Variation in DON content in Canadian reinforces the theory that annual varia- two harvesting seasons (2012–2013 and wheat shipments also occurs from year to tion in DON content occurs throughout 2013–2014) in the Pampas region of year (69). Variation in DON concentra- the Americas. Argentina showed that in some years tions in durum and wheat among years These variations observed among years DON levels were higher than others, on appears to be related to the quality of the may be exacerbated or more difficult to average by 99.7%. Researchers have attrib- grain in the shipments, with higher DON predict due to climate change. A shift in uted the differences to weather-related concentrations seen in grain downgraded environmental conditions could alter Fu- events (Box 2). due to Fusarium damage. This work also sarium populations. According to a re-

Box 2 – Deoxynivalenol Incidence in Argentinian Wheat The sporadic nature of Fusarium head blight and other diseases caused by Fusarium spp. affect different commodities throughout the world, and efforts have been made to reduce the presence of the causative agents. Such is the case for Argentina, where em- pirical forecasting models have been evaluated to estimate the spatial distribution of F. graminearum, a deoxynivalenol (DON) producing mold. These models predict the incidence of the disease based on environmental conditions such as maximum and minimum temperature, precipitation, and relative humidity. While validating the models in the Pampas region of Argentina, two harvesting periods were evaluated, including several wheat cultivars. The models were very accurate and correlated well with the mold incidence and levels of DON detected in the samples (Figs. 1–4). However, one interesting point to highlight from this study was a large decrease, on average (99.72%), in the levels of DON from 2012–2013 to 2013–2014, with no interventions applied by the researchers. Even though predictive models and other good agricultural practices have an impact on the quality of harvested grain, one cannot disregard many environmental factors unaccounted for that may impact DON levels. Research on the perfor- mance of the models was performed at the Agricultural Experimental Station Oliveros and the National Institute of Agricultural Technology in Argentina.

Fig. 1. DON incidence in the 2012–2013 harvest (Lat –32.55, Long –64.32). Fig. 2. Spatial distribution obtained by the Agricultural Experimental Station model for Fusarium incidence in the 2012–2013 harvest.

Fig. 3. DON incidence in the 2013–2014 harvest (Lat –32.55, Long –64.32). Fig. 4. Spatial distribution obtained by the Agricultural Experimental Station model for Fusarium incidence in the 2013–2014 harvest.

CEREAL FOODS WORLD / 37 view prepared by Wu et al. (70), the Fu- 2002 and up to 44% of isolates collected The effect of growing region, which is sarium populations in North America have in 2008. These population shifts may mostly associated with the climatic condi- changed greatly in a period of 10 years. impact the levels of DON found in the tions of a region, is clearly illustrated in Historically, the 15-ADON chemotype of grain at harvest, since greenhouse studies Figures 7 and 8. The map in Figure 7 is F. graminearum has been predominant in have shown that 3-ADON isolates pro- based on data gathered from milling facili- North America and 3-ADON in South duce higher levels of DON in FHB-sus- ties in the USA. In general, the samples America and Europe. However, the Fu- ceptible and -resistant wheat varieties than were collected and analyzed using GIPSA- sarium populations in North America have do 15-ADON isolates (71). In addition, approved methods. The map in Figure 8 been changing. For example in North Da- shifting environmental conditions could is based on samples collected from the kota, the 3-ADON chemotype comprised alter insect populations and impact fungal Canadian Prairies from 2008 through about 3% of Fusarium on wheat prior to infection and mycotoxin production. 2013 as part of the Canadian Grain Commission’s Harvest Sample Program. The Canadian samples were analyzed using gas chromatography–mass spec- trometry according to the method of Tittlemier et al. (69). The maps in Figures 7 and 8 illustrate the variation in DON occurrence in wheat from different growing regions in North America. In the USA, soft wheat from the Midwestern, Northeastern, and Southern regions had higher levels of DON than hard wheat from the same regions (P < 0.05), with hard wheat representing 65% and soft wheat 19% of total USA wheat pro- duction during the period surveyed (2003–2014). Among soft wheat samples, the highest levels of DON were found in the Midwestern and Northeastern re- gions (P < 0.05), with levels, on average, 2–3 times higher in those regions than the levels observed in the Southern part of the country. When hard wheat samples were compared, the Northeastern region showed the highest levels, the Midwestern and Southern regions showed similar levels of DON, and the Western region showed the lowest levels of DON (P < 0.05). For this analysis, the means and standard de- viations for each wheat class/region subset were calculated and compared pairwise using t-test statistics. In western Canada, higher average Fig. 7. Average deoxynivalenol (DON) concentrations at the processor level in soft and hard DON concentrations have been observed wheat harvested in different regions of the USA from the 2003 through 2014 harvest seasons. in hard red spring (HRS) wheat grown in Averages represent 2,818 soft wheat samples and 39,313 hard wheat samples. the eastern prairies of Manitoba (Fig. 8) compared with HRS wheat grown in west- ern Saskatchewan and Alberta. These re- gional differences are driven by environ- mental factors such as temperature and precipitation that predominate around the time of anthesis (and thus affect Fusarium infection), as well as the geographical dis- tribution of DON-producing Fusarium species. According to Cowger et al. (72) and Cowger and Arellano (73), DON lev- els, percentage of Fusarium-infected ker- nels, percentage of Fusarium-damaged kernels, and FHB symptoms at harvest- ripeness increased with an increasing num- Fig. 8. Average deoxynivalenol (DON) concentrations in Canada western red spring wheat grown ber of wet days following anthesis. Cana- in various crop regions of the Canadian Prairies. dian surveys have shown that the main

38 / JANUARY–FEBRUARY 2015, VOL. 60, NO. 1 DON producer, F. graminearum, is pres- used. The Canadian maize export data son of the results from the various studies. ent at higher levels in wheat from the east- (69) were based on random sampling of Nonetheless, the occurrence of DON in ern prairies, whereas other species such as shipment loads and gas chromatography– maize cannot be ignored, and the results F. poae and F. avenaceum are more impor- mass spectrometry analysis with an LOQ summarized in Table 2 provide an indica- tant in the western prairies (74). of 0.05 mg/kg. The Canadian field data tion of DON contamination observed Barley. Limited information exists (76–80) were obtained through random over the past 4 years in maize samples about levels of DON in North American sampling of 10 consecutive ears at 2 loca- from the beginning and end of the grain- barley. In one study, Schwarz et al. (75) tions per maize hybrid grown within a handling chain. Based on this data, the collected barley samples at harvest through- field. Another fundamental difference majority of DON in USA- and Canadian- out all barley-growing regions of North among the reports is the sampling stage. grown maize is present at concentrations Dakota and Minnesota. The samples used For example, the data relating to USA lower than 2 mg/kg, with a small number in their survey were part of regional crop and Canadian exports are from samples of instances where DON concentrations surveys and covered the period from 1993 taken at the end of the grain-handling exceeded 2 mg/kg. Additionally, as seen until 2003. Samples were cleaned using chain (after the grain has been blended in the Ontario field data in Table 2, the a Carter Day Dockage Tester (Seedburo and cleaned), whereas the Ontario field proportion of these higher concentrations Equipment Co.), but no further process- samples were taken at the beginning of will vary from year to year. ing or grading was done prior to DON the handling chain. This lack of consis- analysis. The levels of DON observed in tency with regard to the sampling point Management of DON in Wheat, Barley, the samples are summarized in Table 1. stage in the grain-handling chain, test and Other Small Grains The average DON levels detected in the methods used for DON analysis, and data The North American grain supply chain, samples varied from 0.5 to 10.3 mg/kg. reporting format preclude direct compari- which includes growers, grain buyers, and In most of the surveyed years as least one-third of the samples had levels in Table 1. Levels of deoxynivalenol (DON) in barley samples harvested in Minnesota and North excess of 3 mg/kg, with some years show- Dakota (adapted, with permission of the American Society of Brewing Chemists, from Schwarz ing as much as 59% of the samples with et al. [75]) those levels of contamination. Variation from year to year in DON levels and inci- dence is evident and illustrates once again the volatility of the natural occurrence of such toxins. Data on the occurrence of DON in samples of Canadian barley are present- ed in Figure 9. These samples were ana- lyzed in 2009 through 2014, mainly us- ing ELISA or quantitative lateral flow test methods, with LOD of 0.5 mg/kg, to analyze samples at the farm level (n = 2,237), primary elevator stock (n = 118), and shipment (n = 14) stages. One on- farm sample was analyzed using ELISA with an LOD of 0.25 mg/kg. In addition, some shipment samples (n = 26) were analyzed using gas chromatography–mass spectrometry (LOQ of 0.05 mg/kg) (69). The mean DON concentration for on-farm samples was 0.7 mg/kg, which is within the range of means reported for comparable harvest samples presented in Table 1. Maize. Data on the occurrence of DON in North American maize are presented in Table 2. The reports described in the table used different sampling strategies and methodologies to determine the level of contamination. For example, in the USA maize report on export data (76) a propor- tionate stratified sampling scheme was used to ensure samples taken for analysis were representative of USA yellow maize exports, the samples were prepared for analysis according to the GIPSA DON handbook, and an ELISA test kit (GIPSA Fig. 9. Deoxynivalenol (DON) occurrence in Canadian barley along the grain-handling chain. approved) with an LOD of 0.3 mg/kg was LOQ = limit of quantitation; LOD = limit of detection.

CEREAL FOODS WORLD / 39 end users, has developed and refined man- of DON in wheat and barley that includes vars sown decreased from 76% in 1999 to agement practices and tools to control sowing resistant cultivars, use of fungi- 21% in 2011 (102). The sources of resis- FHB and DON with the benefit of three cides, choice of previous crop, method of tance in the improved cultivars includes decades of modern experience in pro- tillage, and disease forecasting (82–88). In exotic sources such as Sumai #3 from duction, handling, and processing. These maize, management of Gibberella ear rot China and native resistance already practices are effective in safeguarding the with similar cultural practices has had a present in existing breeding germplasm health of consumers. limited effect in reducing disease develop- (103,104). The incorporation of quantita- Results from monitoring and surveil- ment and mycotoxin accumulation (89). tive trait loci (QTL) conferring FHB resis- lance activities indicate that acceptably However, maize hybrids do vary in their tance by wheat breeding programs in low DON levels in unprocessed grain resistance to infection, and growers are Canada and the USA, including the QTL products can be achieved in exports and encouraged to compare these differences FHB1 from Sumai #3, has been enhanced shipments for domestic use in most crop when choosing a hybrid to plant (90). through marker-assisted selection (105). years following current advisory levels Following the severe epidemic of FHB Appropriate crop rotation to reduce and guidelines for DON. These levels are in wheat and barley in Canada and the DON accumulation is a requirement that achieved through the aggregation and USA in 1993, breeding for reduced DON is well-understood by grain growers in blending of grain stocks that occurs along became a higher priority (91). The use of high-risk areas (86). Additionally, com- the supply chain, monitoring of DON disease screening nurseries and collabora- bining improved cultivars with fungicides levels in deliveries of grain shipments to tive regional testing of advanced breeding and optimal spray technology at flowering processors, and a variety of grain sorting lines and cultivars became routine and provides the best management practice and cleaning technologies and methods still continues today. Since the early 2000s, for reducing the risk of DON in wheat that are available to processors. wheat and barley breeders have been suc- and barley, as demonstrated by Wegulo All participants in the grain value chain cessful in releasing cultivars with improved et al. (84) (Fig. 10) and others (106–108). in Canada and the USA (Fig. 1) effectively FHB resistance and reduced DON accu- Fungicides with proven effectiveness manage DON levels in wheat and other mulation to replace susceptible cultivars in reducing DON accumulation include grains. Coordinated research across years (92–101). FHB severity and/or DON ac- tebuconazole (Folicur), prothioconazole and dozens of locations by scientists in both cumulation in the improved cultivars has (Proline), metconazole (Caramba), and countries ensures that research occurs in been reduced by more than 50% compared a combination of prothioconazole and all market classes of wheat and barley and with the susceptible checks (controls), tebuconazole (Prosaro) (109). In a meta- for growers in all at-risk growing areas (81). and growers have been growing the new analysis of more than 100 uniform fun- The research also ensures that the recom- cultivars. For example, in the HRS wheat gicide trials conducted on wheat, Paul mendations made to growers are evidence- growing areas of Minnesota, North Da- et al. (110) found metconazole reduced based. Growers (producers) use an inte- kota, and South Dakota in the USA, the DON accumulation by 50% compared to grated approach to reduce the likelihood frequency of FHB-susceptible HRS culti- the check (control), followed by tebucon- azole (40%) and prothioconazole (32%) Table 2. Reported occurrences of deoxynivalenol (DON) in North American maize (69,76–80) (Fig. 11). Biological control agents are alterna- tives to synthetic fungicides. These agents may provide growers with an additional option for crop protection when the al- lowable window for applying synthetic fungicides has passed (111,112) or for organic production. Organisms that have been tested include the bacteria Bacillus amyloliquefaciens, B. subtilus, and Lyso- bacter enzymogenes and the yeasts Cryp- tococcus nodaensis and C. flavescens (81). Taegro (113) is a strain of B. subtilis that is marketed as a biological control agent. In uniform fungicide trials conducted by the U.S. Wheat and Barley Scab Initia- tive (USWBSI), none of the biological con- trol agents by themselves has been as efficacious as synthetic fungicides in re- ducing DON (77). Growers, crop consul- tants, extension educators, and others in the value chain can access information on best management practices and resistant cultivars online at Scab Smart Manage- ment (114) Important tools for alerting growers and others in the grain value chain of areas at risk for DON accumulation in-

40 / JANUARY–FEBRUARY 2015, VOL. 60, NO. 1 clude web-based disease-forecasting tools within a market class respond to the dis- truck and/or railcar loads for DON levels that utilize hundreds of automated weath- ease. The North American grain-handling to further segregate the grain or to reject a er stations located across Canada and the chain (Fig. 1) generally moves grain in load and have it sent back to the supplier. USA (82,115–117). The system used in bulk from individual farms to primary Steps are taken by processers to ensure Ontario, Manitoba, and Saskatchewan in elevators, where grain from a number of that DON concentrations are below US- Canada, Weather Central (118), is based individual farms is accepted for delivery FDA advisory levels or Health Canada on research at the University of Guelph and combined. Grain from primary eleva- guidelines. (82,115). A number of forecasting models tors is then moved via rail and/or domes- Another challenge in procuring grain are available in the USA; for example, the tic shipping to terminal elevators, where it each crop year is that processors and other Fusarium Risk Assessment Tool (119), is combined with grain from other pri- end users have limitations on the number which was developed using USDA-ARS mary elevators. This results in a large vol- and size of bins they have available for funds, and other forecasting models can ume of grain originating from a variety of segregating incoming grain. Thus, it is be accessed through the USWBSI website locations that should meet the quality and critical that procurers determine the risk (120). Both the Canadian and USA fore- safety specifications of end users. At each for unacceptable DON levels or other casting systems are widely used in the step in the value chain, processors test end-use quality traits in the origination spring and summer in all regions as an indicator of FHB likelihood. Each of the forecasting websites differ slightly, but in general, users can go to the websites and enter information such as their location, the market class of wheat they are produc- ing, the relative resistance of the cultivar they are growing, the date, and the num- ber of hours out they want to forecast potential risk. Based on the information provided, the websites will return maps of the specified region indicating the poten- tial risk of infection. The USWBSI also developed and maintains the FHB Alert System, which sends an e-mail or text mes- sage to subscribers in the USA to warn them that conditions in their region are favorable for FHB development. All of the forecasting tools and the FHB Alert Sys- tem warn participants in the grain value Fig. 10. Fungicide efficacy for Fusarium head blight (FHB) index, deoxynivalenol (DON), Fusarium- chain about regions with potentially un- damaged kernels (FDK), and yield in moderately resistant and susceptible wheat cultivars. Within acceptable levels of DON, growers of the each variable, least significance difference (LSD) values were calculated at P = 0.05. (Adapted, with permission of The American Phytopathological Society, from Wegulo et al. [84].) need for timely applications of fungicides, and grain buyers to develop strategies before harvest to manage potential prob- lems. This is the first phase of determin- ing the potential risk of unacceptable DON levels by producers, procurers, and processors and has a large impact on the sampling plan and management of crop segregation. In years and regions with a high risk of DON, surveillance includes sampling of individual fields before har- vest to ensure unacceptable grain is kept out of the food chain. Large processors and end users have staff or consultants who monitor the expected condition of the crop to avoid sourcing grain from areas with FHB. The overall strategy in North America for managing grain contamination by F. graminearum and DON is illustrated in Fig. 11. Efficacy of various fungicides for reducing Fusarium head blight (FHB) in the field and Figure 12. suppressing deoxynivalenol (DON) content in harvested grain, expressed as percent control com- The challenge of sourcing acceptable pared with untreated wheat. FHB = Fusarium head blight index; DON = deoxynivalenol; TEBU = grain is impacted by the year-to-year vari- tebuconazole; PROP = propiconazole; PROT = prothioconazole; TEBU + PROT = tebuconazole plus ations in DON levels in different sourcing prothioconazole; METC = metconazole. (Adapted, with permission of The American Phytopatho- areas and even by how different cultivars logical Society, from McMullen et al. [81].)

CEREAL FOODS WORLD / 41 areas for the different grains they wish to is based on 1) food safety; 2) functional system or local origins. Elaborate sys- procure. After this is completed, the pro- requirement; and 3) consideration of pro- tems are in place to assure consistency cessor can create the blend that will meet cessing performance. of grade, including minimizing DON lev- the customers’ requirements. The ranking Another consideration is whether the els in grains that are exported. Processors of components of the grain blend design processor is using grain from the export within Canada and the USA receive grain shipments originating from farms, pri- mary elevators, and terminal elevators; thus, the domestic processor often has much more variation to deal with than a processor procuring grain through the export channel. Having a target for fin- ished products is more appropriate to as- sure food safety compliance than target- ing unprocessed raw grain with a fixed level when DON distributions are not known. The relationships between FHB, FDK, and DON in grains are complex (121) but understood to result from both the plant response to the pathogen (122) and strain variation (123,124). Nonetheless, harvest surveys have demonstrated that DON in wheat can be managed by minimizing the amount of FDK present in grain and us- ing FDK as a grading factor (69). Growers can reduce FDK in the harvested grain by adjusting airflow settings on their com- Fig. 12. Overall strategy for managing deoxynivalenol (DON) in grains. bines (125). In addition to the blending that occurs along the Canadian and USA value chains, cleaning often occurs at larger primary and terminal elevators. Grain can be cleaned using a variety of techniques, such as optical sorting, sieving, and grav- ity tables. These handling steps can de- crease the level of FDK and, thereby, DON levels in a volume of grain (126). The ability to handle large volumes of grain of varying quality and meet desired specifications increases along the grain- handling chain. Figure 4 illustrates how higher concentrations (i.e., >2 mg/kg) of DON in wheat occurred at earlier stages Fig. 13. Levels of deoxynivalenol (DON) detected at the processor level in wheat samples from in the grain-handling chain, such as on the Northeastern, Midwestern, Western, and Southern regions of the USA from 2003 through farm or at primary elevators, but not fur- 2014 (n = 42,131). Samples include soft and hard wheat varieties. “Estimated Tonnes” lists the ther along the grain-handling chain. The amount of wheat represented by the samples analyzed each year. fraction of wheat samples containing DON at concentrations >2 mg/kg decreased from 19% on farm to 0.5% in shipments from terminal elevators when analyzed in nontargeted monitoring and surveillance programs (i.e., where the sampling scheme was not biased). A decrease in the occur- rence of DON concentrations along the handling chain similar to that described for wheat was observed for Canadian du- rum and barley and is illustrated in Fig- ures 5 and 9, respectively. Processors at intermediate stages of the grain-handling chain can be more limited Fig. 14. Annual average deoxynivalenol (DON) concentrations in exported soft wheat sublots in their ability to manage DON in deliver- from the USA. Values in parentheses indicate the number of sublot samples analyzed each year. ies coming directly from farms or primary

42 / JANUARY–FEBRUARY 2015, VOL. 60, NO. 1 elevators. Figure 13 shows the incidence of wheat leaving the USA as exports. Based can be physically indistinguishable from and levels of DON reaching USA milling on the data from the USA, as for Canada, healthy grains and not removed by some facilities from 2003 through 2014 and rep- in general as grain moves through the sorting methods; and 2) FHB level is not resents more than 1.2 million tonnes of handling chain to export the average re- always correlated with the concentration wheat. (Figures 13 and 14 use the same ported levels for DON are reduced, even of DON in the grains when secondary data set as was used for Figures 2 and 3.) more so for soft wheat than for hard Fusarium infection occurs (131). The DON levels shown in Figure 13 wheat in the last 10 years. Cleaning can also be used to remove follow a distribution, with the majority of infected grains. Nowicki et al. (132) the samples showing levels of contamina- Effect of Processing on DON Content reported that scouring was effective in tion below the LOQ of the method. Two in Wheat and Other Grains reducing the level of DON by 22% be- LOQs (0.3 and 0.5 mg/kg) are reported in Wheat Milling. Harvested grains are cause the toxin was unevenly distributed the figure because different milling facili- converted into flour and other fractions on the surface of grains; Fusarium was ties used different ELISA methods to eval- for human consumption or further manu- also removed from the surface of grains. uate the samples. Among the samples that facturing. Processing consists of cleaning According to Abbas et al. (133) the prep- showed quantifiable amounts of DON, (including sorting) and milling (Fig. 15). aration cleaning step in the milling pro- 3.1% had levels that exceeded 2 mg/kg. In The sorting step removes any impurities, cess reduced the concentration of DON Canada, data from a small targeted survey such as straw and dust, but can also be by 6–19% in infected wheat grains, with run from June 2011 through October 2013 used to remove broken or mold-damaged cleaning efficiency depending on the showed that up to 9% of wheat grown in grains. This selection or separation step is initial condition of the grain and the eastern Canada and delivered to milling performed based on the physical param- extent of the contamination (Table 3). facilities contained DON at concentra- eters of the kernels, such as shape, size, More recently, Lancova et al. (140) ob- tions ≥2.0 mg/kg (127). This is within the specific gravity, relative density, and color. served that the effective removal of range of 1–10% of samples containing Fusarium-infected grains become shriv- screenings and the outer layers of bran DON at concentrations ≥2.0 mg/kg noted eled and lighter than healthy grains (126). from the surface of grains during clean- by USA wheat processors (Fig. 13). Some studies observed that selection and ing steps reduced the concentration of Information regarding the levels of separation based on gravity was effective DON by 48%. DON in USA wheat destined for export in removing heavily infected grains with The international wheat milling indus- can be found in the U.S. Wheat Associates high concentrations of DON from healthy try has carried out several DON reduc- Crop Quality Reports (128). U.S. Wheat grains (126,130). Nevertheless, this man- tion trials in wheat mills over the last Associates provides annual reports on the agement approach has two potential prob- 5 years. These normally have shown the quality of the different classes of wheat lems: 1) many Fusarium-infected grains, need to both mechanically clean and to grown in the USA. Different USDA-ARS which may contain high levels of DON, pass the wheat through an optical sorter and USDA-GIPSA laboratories and other government institutions provide the data used in these reports. A historical data set made available by U.S. Wheat Associates provides information on DON levels in samples of hard (n = 6,278), soft (n = 1,869), and durum (n = 718) wheat di- rected to exporting from 1997 through 2013. This data set shows that, as discussed previously, the levels of DON in USA wheat being exported are maintained, through management, on average below 2.0 mg/kg. However, some years and/or wheat classes are more challenging to manage than others. For example, DON levels in samples representing shipment sublots of hard wheat for export have been reported as being below 1.0 mg/kg on average, with levels of up to 8.10 mg/kg found in individual samples in 2001. For durum wheat, averages also were below 1.0 mg/kg over the last 10 years, with maximum levels of 4.6 mg/kg being de- tected in 2003 in samples representing export sublots. For soft wheat directed to Fig. 15. Wheat cleaning and milling process diagram. Cleaning steps in green boxes indicate export, averages historically have been where reduction of deoxynivalenol (DON) typically occurs (129). * Wheat “Red Dog” consists of well below 1.5 mg/kg (Fig. 14); however, the offal from the “tail of the mill” together with some fine particles of wheat bran, germ, and levels of up to 5.9 mg/kg (in 2003) were flour. This product must be obtained in the usual process of commercial milling and must contain detected in samples representing sublots not more than 4% crude fiber.

CEREAL FOODS WORLD / 43 to maximize reduction of DON levels. achieved using only a visible light sort Break rolls do not crush the grains but They showed typical reductions from with a blue filter. In this case, reductions shear them open, separating the bran 3 mg/kg in wheat to about 1 mg/kg at the of 25% of the DON level or more have from endosperm. The sheared grains are output of the optical sorter placed after been achieved using a standard optical sieved into three main components: bran, mechanical cleaning, although there was sorter as found in the cleaning section of germ, and endosperm. In stage 2, the en- some variability depending on the in- a modern mill in Europe or the USA (per- dosperm particles are channeled to a set coming grain (personal communication). sonal communication). Optical sorting is a of smooth reduction rolls for final milling Mechanical cleaning normally consists of potential option for mills to use to man- into white flour. This separation of the a combination of a classifier, gravity sepa- age high or variable DON levels in wheat bran from the endosperm can signifi- ration, and aspiration, sometimes also us- and other grains; however, it is a signifi- cantly reduce DON levels, since most ing a scourer. cant investment for a mill. mycotoxins tend to be concentrated in The optical sorter, which is not standard During the milling process some grains the bran (outer layer) and germ fractions in all mills, normally uses a light sort in are polished, removing most of the bran (Table 4). the visible region and a light sort in the and germ and leaving the endosperm, In the case of the wheat milling process short-wave infrared (SWIR) region. The followed by the reduction of the endo- (Fig. 15), separation involves passing the largest reduction (relative) is normally sperm to a uniform particle size of flour. grain through 2–3 break rolls and 3–5 re- seen with the use of the SWIR region of This process involves a sequence of break- duction rolls, each followed by sieving. the sorter. Typically the optical sorter will ing (stage 1), reducing (stage 2), and sep- This results in five fractions: white flour, remove between 2 and 5% of the product arating steps. In stage 1 the grains are wheat germ, wheat feed (red dog), wheat stream to achieve these reductions. In passed through a set of rough or corru- shorts, and wheat bran. According to sev- smaller mills, some reduction can be gated break rolls rotating at different speeds. eral researchers, when milling wheat or maize mycotoxins tend to be concentrated Table 3. Effect of cleaning on removal of deoxynivalenol (DON) from grain (131,133–141) in the germ and bran fractions (133,141– 145). Similar to other mycotoxins, DON is a heat-stable compound (155) and may not be destroyed during most food pro- cessing operations, including milling (156). Therefore, the best way to reduce DON is to separate the highly contaminated ker- nels from the bulk during the preparation steps and subsequent sieving steps. Research has shown that the white flour has approximately half the level of DON found in the cleaned wheat, while the bran can have levels two or more times greater than the initial whole wheat (133,150). According to the research of Tanaka et al. (150), the flour fraction of barley con- tained 3.2% of the DON, while the bran fraction contained 96.8% of the DON found in the whole grain. In the case of wheat, the bran contained 81.6% of the DON. Milling wheat and barley is effec- tive in removing the DON contained in the outer layers of the kernels. It should also be noted that significant improve- ments have been made in modern mills to decrease DON in the flour fraction (Fig. 16). There have been varying reports of re- duction of DON levels using peeling or debranning of wheat. The assumption is that the mycotoxin is in the bran layer on the surface of the grain. Some trials have shown DON reductions of up to 50% af- ter debranning, but it has also been sug- gested that reductions will be lower when the whole grain has been affected by fun- gal attack. DON may also be distributed throughout the milling fractions indepen- dent of wheat variety (131,133–136,140, 157,158).

44 / JANUARY–FEBRUARY 2015, VOL. 60, NO. 1 Dry Milling of Maize (Corn). With for DON across multiple grains must sat- termination of MLs for DON across wheat, the milling process does not re- isfy the “significant contributor to total multiple grains must afford adequate duce the total amount of DON but rather dietary exposure for consumers” criterion health protections in bad climactic years, converts the wheat into fractions, segre- and provide health benefits. Setting levels provide additional health benefits at a gating the DON in lower value fractions for DON in an unprocessed raw commod- reasonably achievable level, and be prac- while the high-value fraction, white flour, ity that is not typically consumed in that tically achievable so that trade disruptions is relatively low in the toxin. Maize goes form and that contributes minimally to do not occur. through a different milling process (159). dietary exposure from the finished food Emesis is the critical effect in DON After cleaning, maize kernels are soaked product (after processing) may not sig- acute food toxicity. The acute reference twice in water, then passed through a nificantly change health outcomes. De- dose (ARfD) of 8 µg/kg bw/day of DON degerminator and dried again. The de- germed maize is then sifted and either Table 4. Effect of milling on removal and distribution of deoxynivalenol (DON) from grain (126, placed on gravity tables for production of 131–138,140,146–154) germ and coarse grits or rolled to make flour or fine grits. During the milling pro- cess, the DON segregates heavily into the bran, screenings, germ, and germ meal fractions, leaving the grits and flour frac- tions with 5–10% of the DON level con- tained in the bran (160). Most products for human consumption are prepared from flour and grits. Therefore, it is pos- sible to reduce DON in human foods during the milling process. Similar to wheat milling, these results suggest the milling process does not remove DON but redistributes it. This explains the con- sistent tendency for DON to segregate into milling fractions used for animal feed and out of fractions used for human foods; thus, it makes much more sense to regu- late milled products than unprocessed raw grain. Oat Milling. When oats are milled, they undergo a dehulling process to remove the husk from the grain before kilning (heat treatment to inactivate lipase). Most of the DON in oats is concentrated in the hull, so dehulling oats results in a reduction in the level of DON compared to the initial grain. This is evidenced by the oat flakes produced by the milling process, which contain 5–10% of the DON present in the oats before milling (161).

Risk Management Measures for DON Maximum levels (MLs), which are the maximum concentrations of a specific substance recommended to be legally per- mitted in a specific commodity, provide dietary guidance for consumers. MLs are not necessary for all contaminants in all foods. Establishing an ML for a given contaminant in a certain food would be necessary only if and when the contami- nant may be found in amounts that result in significant adverse public health impacts. In this case, the ML would afford adequate assurances of safety. The main criterion is that the contaminant is a significant con- tributor to total dietary exposure for con- Fig. 16. Deoxynivalenol (DON) distribution (%) in milled wheat fractions (total DON amount in sumers. Therefore, determination of MLs cleaned grains = 100%). (Derived from data presented in Tables 3 and 4.)

CEREAL FOODS WORLD / 45 used in international standards was estab- tion patterns for 13 regional clusters (164). unprocessed raw grain consumption as- lished by the 72nd JECFA in 2010 (162). The regions likely to contribute most to sumptions removed contributions from This ARfD is used in the derivation of DON exposure from different grains for primarily flour. proposed MLs that would adequately pro- the global population are represented by The WHO GEMS/Food cluster diets tect health (163). JECFA reported that clusters B and E (Table 5). Using this in- were updated in 2012 (165). The 2012 up- “dietary exposures to DON up to 50 µg/kg formation, the proposed MLs for unpro- date employed a new approach to assess bw/day are not likely to induce emesis” cessed raw wheat, maize, and barley were diets and consumption patterns, reorga- (162). This suggests that JECFA has estab- calculated. Based on the predicted con- nizing the regional dietary consumption lished an ARfD that affords at least a five- sumption of unprocessed raw grains, the patterns into 17 clusters (Table 6). MLs for fold margin of safety. calculated MLs that would adequately pro- consumption of unprocessed raw wheat, The 2006 WHO Global Environment tect health based on worst-case scenario maize, and barley that adequately protect Monitoring System/Food Contamination assumptions were 43, 4, and 11 mg/kg for health can be recalculated using the new Monitoring and Assessment Programme wheat, maize, and barley, respectively GEMS database information. These are (GEMS/Food) identified food consump- (details provided in Tables 5 and 7). The based on worst-case scenario assumptions (removing contributions primarily from Table 5. Grain consumption rates (g/day) derived from the 2006 WHO Global Environment flour). Based on the predicted consump- Monitoring System (GEMS) cluster diets tion of unprocessed raw grains, calculated MLs were 21, 19, and 9 mg/kg for wheat, maize, and barley, respectively (details pro- vided in Tables 6 and 7). In summary, understanding the differ- ences in consumption patterns across re- gional clusters of different unprocessed raw grains intended for sale and for direct human consumption is necessary to de- termine appropriate MLs for these grains. It is assumed that these grains will not undergo further processing. Addition- ally, contributions from different grains (durum wheat, soft wheat, maize, and barley), whether unprocessed raw, semi- processed, or processed, to the overall consumption of grain-based products should be tallied. This would allow for an Table 6. Grain consumption rates (g/day) derived from the 2012 WHO Global Environment Monitoring System (GEMS) cluster diets accurate assessment of appropriate MLs for various grains at each stage of the har- vesting and milling continuum. Further- more, MLs proposed should correspond to the point at which the grains are ana- lyzed and sampled for DON. This would ensure consistency of the reporting basis between MLs and the occurrence data. The point of sampling and analysis varies from country to country. For example, in the United Kingdom the sampling is a mixture of point-of-sale data and harvest data. To compare data and provide consen- sus on appropriate MLs, the international community must first attain consensus on stages of the harvesting and milling con- tinuum at which sampling and analysis Table 7. Comparison between corrected maximum levels (MLs) from 2006 WHO Global should be done. Environment Monitoring System/Food Contamination Monitoring and Assessment Programme (GEMS/Food) consumption data, 2012 WHO GEMS/Food consumption data, and proposed MLs Method for Estimating Intake of in the Codex Committee on Contaminants in Foods (CX/CF) 12/6/9 (166) Unprocessed Raw Grains Total grain consumption is not an ap- propriate substitute for unprocessed raw grain consumption when used to establish an ML for DON in unprocessed raw grains intended for direct human consumption. This assumption can result in an overesti- mation of the actual unprocessed raw

46 / JANUARY–FEBRUARY 2015, VOL. 60, NO. 1 grains consumed and would generate an traits being pursued. The presence of Cowger and Sutton (172) estimated the unrealistically low ML that would not DON also reduces the quantity of vul- impacts of the 2003 SRW wheat outbreak. provide additional public health benefits. nerable crops produced, raises production They interviewed researchers, extension To eliminate the inherent uncertainties costs, and increases premiums for wheat specialists, extension agents, millers, and associated with the GEMS/Food data rela- and other grains at risk for DON con- growers in the southeastern USA for opin- tive to the amount of unprocessed raw tamination. This leads to higher overall ions on the 2003 infestations. Lilleboe grains actually consumed globally, it may costs and risks and more complicated lo- (173,174) summarized the effects of FHB be prudent to seek actual unprocessed gistics for domestic processors and im- in 2010 and 2011 across states and crops raw grain (specifically durum wheat, soft porters, and, finally, it raises the costs of for the USA. McMullen et al. (81) sum- wheat, maize, and barley) consumption breeding. All of these effects would be marized past studies on the economic data from impacted nations to serve as exacerbated by proposals to measure and estimates of FHB outbreaks and the de- the basis for the proposed ML. Risk man- limit DON on unprocessed raw materials gree and location of FHB outbreaks by agement decisions should be based on the instead of products. class of wheat since 1991. most appropriate data when specifying Earlier studies by Johnson et al. (168) proposed mitigation measures for DON and Nganje et al. (169–171) have esti- Grower Risks and Responses to contamination of unprocessed raw grains mated the value of economic losses due to the Incidence of FHB intended for direct human consumption. DON to the USA industry. Johnson et al. One of the fundamental impacts of FHB It is an internationally accepted prin- (168) estimated production and price ef- on growers is that it increases risks, raises ciple that MLs for DON in unprocessed fects for HRS wheat, durum, and soft red costs, and ultimately increases the propen- raw grain should be set at levels that are as winter (SRW) wheat from 1993 to 1997. sity for growers to switch to other crops low as reasonably achievable (167) and They estimated relationships for yield as with less risk and/or an increase in the that MLs should be adequately protective a function of rainfall, temperature, and risk premiums for growing these crops. of health, yet also practically achievable so trend. To proportion losses due to FHB trade disruptions do not occur or are min- and determine yields without FHB, they FHB Risks imized to the extent possible. utilized the difference between the fore- There are two notable risks related to casted yields and actucal yields, together FHB for growers. One is the incidence of Aggregate Impact of DON on North with expert opinion. Acreage adjustments FHB during production, which reduces American Food Grains were included to compensate for higher yields and damages kernels, potentially to The long-term prospective impacts of acreage abandonment in FHB outbreaks. the point of being unmerchantable. Some imposing limits or further regulations Price effects were evaluated as the produc- crops are more susceptible to Fusarium regarding DON would affect growers, tion shortfall impact on market prices and infection and DON production due to millers, food processors, traders, han- the effect of crop quality premiums and dis- factors such as variety, growing regions, dlers, and ultimately consumers. Cur- counts. The economic impact of produc- and weather variations from year to year. rently the USA domestic industry con- tion losses ranged from 3.6 million tonnes Soft wheat varieties, especially soft white forms to advisory limits. Other countries in 1993 to a low of 1.7 million tonnes in winter (SWW) wheat, when grown in have limits on unprocessed raw and/or 1996. Price effects due to production short- temperate regions like eastern and mid- processed grains. Proposals have been falls resulted in higher prices than would western areas of the USA and eastern made that would change these limits to be have occurred without FHB outbreaks. Canada are vulnerable to DON produc- more restrictive, particularly for unpro- Combining the two effects reduced the tion. This is a result of the higher frequent cessed raw grains. If adopted, this would total impact of FHB. The largest produc- rainfall coupled with warm temperatures impact the entire value chain, which tion effect occurred in 1993; however, when that are typical for these areas, in addition could result in escalated costs, reduced considering both price and production to the emerging growth of other crops exports, expanded use of wheat as feed effects, the largest losses occurred for all (i.e., maize) in these regions. This cor- within North America, increased testing classes in 1995, due mostly to the large roborates discussions presented in previ- and segregation costs, and growers inevi- negative price effects on SRW wheat. For ous sections concerning regional vulner- tably further switching away from wheat HRS wheat, the year with the largest loss ability to DON production throughout and other affected grains in regions that was 1994 at $245 million. North America and elsewhere. While pub- are vulnerable to or at higher risk for ex- Nganje et al. (169) updated the results lished data for maize are more limited for cessive DON levels. from Johnson et al. (168) to cover the years North America, the presence of DON has The presence of DON has major impli- 1998–2000 and expanded the analysis to been demonstrated. Variability in occur- cations for the entire supply chain of af- include malting barley. Nganje et al. (169, rence of FHB also is seen from year to year. fected grains, including both for domestic 170) used generated losses attributable to For example, in 1996 the SWW wheat crop processors of grain-based products and FHB to estimate direct and secondary in Ontario and the eastern USA was dra- for exporters and importers. The presence yield losses that were then utilized within matically affected by FHB, cutting yields of FHB in North America and the resul- an input-output model of the economy to by a third, with much of the crop unus- tant incidence of DON raise costs and project direct and secondary economic im- able for human consumption. HRS wheat risks for growers, inducing them to use pacts on the larger economy. Direct eco- and durum crops were severely impacted more costly management practices and/or nomic losses in the USA were estimated in 1993, followed by several other years, shift to other crops. In addition, it raises at $870 million from 1998 to 2000. including 1998–2010 and 2014 (81). the cost of developing new cultivars, since More recent estimates of outbreaks The other critical risk related to FHB is all lines of germplasm must be screened have tended to focus on yield losses and price discounts, or more specifically, post- for Fusarium resistance regardless of other the extent of coverage of FHB outbreaks. harvest price discounts. These discounts

CEREAL FOODS WORLD / 47 are encountered in numerous forms. The counts. The former two have obvious, al- scab) for a representative country elevator most common current forms are simply though random, price adjustments. Table 8 in North Dakota for HRS wheat. At the rejecting the product, reclassifying it to a shows discounts that were applied in 2014 time these discounts were applied, wheat lower grade/class, or applying price dis- for Fusarium-damaged kernels (FDK or prices were $6/bu ($220/tonne). As an example, a discount of $1 is about 17% of Table 8. Example of wheat discounts for Fusarium-damaged wheat kernels (FDK) at a representative the price of the crop or a loss of $40/acre primary elevator in North Dakota ($99/ha), which is sufficient to shift plant- ings to other crops. Currently, it is common to either test a sample of the grain analytically at internal labs, send it to a grain inspection service for a certified test, or require growers to provide a certificate of a test result prior to unloading the grain. Similar proce- dures are used at export elevators. The presence of FDK results in financial losses to growers due to the penalties or reclassifications of wheat lots, as shown in Table 8. If the levels exceed grading stan- dards or contractual obligations, excessive FDK levels can rise to the point of a total crop loss for a grower.

Farm Management Practices and Costs As described earlier, growers now have a number of farm management practices that can be adopted with the goal of miti- gating some of the risks of DON. These include, but are not limited to, crop choices, choices among improved varieties regard- ing DON resistance, agronomic practices including crop rotation, application of fungicides, etc. Farm management prac- tices to control DON were analyzed re- cently by McGee et al. (175) and are pre- sented in Figure 12. As discussed earlier, fungicides are particularly important and, in response to the 1996 crop failure, were registered in Canada for mitigating FHB in soft wheat starting in 1999. In addition, Fig. 17. Area planted to selected crops in North Dakota. (Source: USDA-NASS [179]) varying forecasting tools now exist. DON- cast is one modeling tool that has been available to growers in Ontario since 2000; other similar models are available in Can- ada (118) and the USA production regions (119,176,177). To understand the quantitative impacts of these on farm management decisions, Wilson and Dahl (178) quantified the risk- iness of alternative grains in North Dakota. These include not only price risks but also risks related to yield, quality rejection, and post-harvest quality discounts. These re- sults illustrate the riskiness of HRS and durum wheat relative to competing crops.

Switching to Competing Less Risky Crops DON has resulted in growers shifting production to less risky crops/crop rota- Fig. 18. Area planted to selected crops in Canada. (Sources: USDA-FAS [180] and Agriculture and tions. Changes in cropping patterns have Agri-Food Canada [181], various years) been influenced by many factors, includ-

48 / JANUARY–FEBRUARY 2015, VOL. 60, NO. 1 ing government farm program changes; effectively manages its impact. However, thresholds for DON as a primary means the rising importance of ethanol; increased these adaptations have a cost, do not elim- of ensuring finished goods as consumed demand for maize, soybeans, and canola; inate DON completely, and have the effect will meet the <1 mg/kg advisory level set increased profitability of alternative crops; of reducing the product stream. by the US-FDA. DON level mitigation etc. Increased risk of FHB occurrence may As grain moves along the supply chain can be achieved through careful selection also impact a grower’s decision concern- (Fig. 1), its quality is improved by the ap- of lots of grain; blending, cleaning, and ing which crop to cultivate, as illustrated plication of several management strate- sorting; removal of the husk, hull, and bran in Figure 17 for North Dakota and Fig- gies, including segregation, blending, and layers; and dilution with other ingredients ure 18 for Canada. In North Dakota HRS cleaning (183–185). These strategies allow in a processed food. DON levels are not acres have declined from about nearly not only for the improvement of the qual- reduced through washing or cooking. 10 million acres (ma) (4.1 million hect- ity of the grain, but also its safety by man- Processors such as bakers or breakfast ares), to about 6 ma (2.4 million hectares); aging the presence of DON. This man- cereal manufacturers have few tools avail- durum has decreased from 2.5 ma to less agement, including the monitoring of able to reduce DON levels. This is par- than 1 ma; and malting barley acres have grain for DON along the handling chain, ticularly problematic for products that fallen similarly. Similar changes have also requires resources for proper sampling, contain whole grain and whose primary occurred in Canada for all wheat, decreas- analytical testing, and interpretation of component is a grain such as soft wheat. ing from 30 ma (12.3 million hectares) to test results. These activities add time, as Whole grain crackers and cereals are ex- 20 ma (8.2 million hectares), mostly switch- well as cost, to the handling of grain. amples of products for which ingredient ing to canola which is less risky and more At the mill level (or point of first process- dilution and bran removal are not avail- profitable. ing), the incidence of DON is also man- able options. For these processors, the Regardless of the cause, it is a fact that aged. The most effective way (Fig. 15) is primary tool available is the inclusion of grains that are vulnerable to Fusarium are to remove Fusarium-damaged kernels from DON limits in contractual agreements. more risky than the alternatives, and this the unprocessed raw grain during the ini- For primarily whole grain products to be provides a motivation to shift to less risky tial sorting and cleaning steps prior to sold in Canada or the USA, contracts lim- crops. Or, similarly, in order to retain the milling. Cleaning and scouring of wheat it DON to <1 mg/kg. For products des- same area planted with these crops, risk prior to milling reduces DON content, tined for Europe, contracts will often set premiums have to increase for these crops and the milling industry also uses optical the thresholds for DON at 0.5 mg/kg. To relative to competing ones, which has in- sorters that can effectively reduce DON illustrate, a list of countries and their con- deed been seen during this period. content. tract limits on DON in wheat is provided Ali and Vocke (182) analyzed the im- Further processors of grains such as in Table 9. It is clear from this list, that pact of DON on planting decisions by mills and food manufacturers specify importers can effectively limit their DON growers since the 1990s. The degree of disease incidence may be increasing due Table 9. Contract limits on deoxynivalenol (DON) (vomitoxin) specifications (Source: U.S. Wheat to larger maize plantings and the switch Associates) toward minimum or reduced tillage prac- tices, which increase host presence for DON development when environmental conditions are favorable. Similarly, much of the durum and malting barley produc- tion has shifted out of eastern, central, and northeastern counties in North Da- kota into more western counties in North Dakota and eastern counties in Montana. Growers have many options available to them when planning crop rotation pro- grams. Should the risks associated with growing a specific crop exceed the return, growers could cease to plant those crops. Soft wheat varieties, in particular, and dry-milled maize for food use are cur- rently at some risk, leading processors to haul grain greater distances at greater cost to meet their manufacturing requirements. This means significantly added costs for consumers.

Supply Chain Impacts: Current Practices The supply chain for these grains was described earlier. It is important to note that the supply chain has adapted to the incidence of DON and efficiently and

CEREAL FOODS WORLD / 49 content through maximum specifications. Mills would seek to purchase wheat with that have the ability to reduce DON con- It is worth mentioning, however, that some greater assurance of conformance to the tent in years when it is problematic. This grain suppliers are reluctant to contract MLs and/or add preprocessing functions is most commonly accomplished by tar- based on DON levels given that analytical to assure meeting MLs for DON. This geting origins and cleaning at the country tests add cost and delivery delays. Some could include a number of strategies: and export elevator (183,184) prior to handlers have been slow to adopt quanti- sourcing wheat from production regions shipment, ultimately to meet specifica- tative DON testing. that typically experience lower incidence tion limits set by importers. It is clear from Some further processors of grains, espe- of Fusarium and lower DON levels (where these strategies, that importers can effec- cially small- and medium-sized enterprises, possible); imposing stricter limits on tively limit their DON content through that have less leverage at the grower level wheat and/or larger discounts for ex- contract limits. Of course, tighter controls maintain that a ML set on grains could cessive DON; requiring more extensive would accrue greater costs due to testing enhance market pressures for growers to cleaning prior to shipment from the eleva- and potential rejections. Meeting these expand the use of tools such as fungi- tor; and segregation, testing, and cleaning specifications could impact the available cides. Others argue that sufficient safe- at the mill to assure ML conformance. supply of wheat. guards are already in place, which is evi- Ultimately, this would add costs to both Finally, importers would also be im- denced by targeted surveys of foods at the purchasing and processing. These added pacted indirectly. If stricter MLs are im- retail level that show a very high compli- costs would vary through time and geo- posed, the strategy of domestic mills in ance level for achieving DON levels of graphically. In-mill costs of cleaning, grain-exporting countries would change <1 mg/kg in the finished food. Both par- scouring, added storage, etc. would also and could reduce (slightly) the available ties agree, however, that setting any limit vary through time but, nevertheless, supply of DON-compliant wheat to the for DON on unprocessed raw grain would would be a real added cost to the milling export market. For this reason, there be highly problematic for North America sector. Since such MLs currently are not would be a minor component of crops given the vastly different agronomic prac- applied to unprocessed raw grains, major with unacceptable (even if precautionary) tices from those in Europe, such as farm processors have not yet quantified the DON levels that would be channeled else- size, transport, and storage, in addition to costs associated with such proposed regu- where in the market system. This would differences in climate and farm support lations. To do so would require detailed include the export market or markets for payments. simulation modeling of these functions nonfood uses. For importers, this could and risks. Ultimately, these costs would increase the risks of receiving shipments Prospective Impacts of Establishing vary by location and be determined by containing DON at unacceptable concen- International MLs for DON in logistics and the mill’s capability for seg- trations and, as a result, would require Unprocessed Raw Grain on Supply regation and cleaning. more scrutiny of contract limits and price Chain and Marketing Practices As previously discussed, growers cur- differentials. The response would be to Currently, MLs in the USA are applied rently seek to control DON through crop increase the intensity of cleaning and tar- on semiprocessed grains. However, some rotation strategies, adoption of improved geting of origins for grain prior to export- countries would like MLs to be applied cultivars, and use of fungicides and opti- ing, which also would affect costs. to unprocessed raw grains, which could mal spray technology at flowering. Tight- detect affected grain prior to entering the er MLs on DON would have the following Final Remarks primary elevators. The latter would be impacts: more intensive management, In agricultural commodities, the occur- more costly and restrictive. including variety selection and fungicide rence of DON has been reported all over There are different points in the supply use, all of which will add costs; larger dis- the world, with levels varying among grain chain (Fig. 1) at which MLs could be ap- counts for excessive DON; and greater types and years of production. The grain plied, including at the farm, at the point risk for wheat and other crops grown with supply chain, including growers, buyers, of entry into the marketing system (i.e., excessive DON, leading to a shift to other and end users, has effectively managed the primary elevator), at terminals, at the crops with less risk or greater returns. DON with strategies to control this issue points of first processing, or on finished These impacts would vary geographically systematically. The safety of consumers is products. Due to the density of transac- and through time. It is important to note ensured with use of these management tions and costs of executing such a pro- that as the risk associated with wheat strategies. gram, it would seem most practical that production increases and concurrently Based on the information reviewed and these regulations or standards would be competing crops become more viable introduced in this report, the occurrence evaluated at mills. This is a point with and profitable there would be a move of DON in North America does not ap- fewer locations and would likely be the away from production of wheat and pear to be different than that reported most cost-effective. For these reasons, the other affected crops in the USA, Canada, around the world. Sporadically, levels of direct impacts of establishing MLs on DON and elsewhere. This likely would reduce DON in grains (wheat, maize, and barley) would be on domestic mills, while there supplies and increase prices to all con- can be much higher than usual in certain would be indirect impacts on growers and sumers. years due to more severe Fusarium infec- exports and imports (discussed below). Importers of grains, at least from North tion. Cool and wet conditions during In response to this proposed regulation, America, also have tools available for con- specific developmental stages of grains domestic mills would be impacted most trolling DON content. In concept these (e.g., wheat flowering) promote Fusarium directly, although an increase in cost are similar to those of domestic mills and infection, although wet conditions at har- would impact other sectors in the supply include primarily contract limits for DON vest can lead to secondary Fusarium in- chain as well. Mills and first processors content. Additionally, these can include fection and DON production as well. would have to change their strategies. targeting locations, ports, and suppliers Other factors influencing the levels of

50 / JANUARY–FEBRUARY 2015, VOL. 60, NO. 1 DON in agricultural crops include the cluding contract limits and discounts for grain.pdf. USDA-FAS, Washington, DC, growing region and its climate, in addi- excessive DON; 3) segregating wheat in 2015. tion to wheat class. storage; 4) selectively testing and/or clean- 2. Frisvad, J. C., Thrane, U., and Samsom, As discussed in the report, manage- ing to assure contract performance; and R. A. Mycotoxin producers. In: Food Mycology: A Multifaceted Approach to ment of FHB and consequently DON in 5) where appropriate, use of cleaning and/ Fungi and Food. J. Dijksterhuis and R. A. grains is a complex problem with nu- or scouring to remove kernels with exces- Samson, eds. CRC Press Taylor & Francis merous component methods available. sive DON content. Through these pro- Group, Boca Raton, FL, 2007. Each method, such as crop rotation, till- cesses, North American domestic mills 3. Diaz, D. E. The Mycotoxin Blue Book. age, cultivar resistance, fungicides, bio- effectively and efficiently control the inci- Nottingham University Press, Notting- logical control agents, and optimized dence of excessive DON in their products. ham, U.K., 2005. spray technologies, provides some ben- Any additional restrictions would increase 4. Streit, E., Naehrer, K., Rodrigues, I., and efit, but using multiple approaches is the costs and risks to mills not only in North Schatzmayr, G. Mycotoxin occurrence in most efficient way for growers to manage America but also globally. feed and feed raw materials worldwide: this problem. More restrictive interventions on DON Long-term analysis with special focus on As far as the levels of DON found at content would have numerous impacts. It Europe and Asia. J. Sci. Food Agric. 93: 2892, 2013. different stages of the grain-handling is important to note that there would be 5. Schollenberger, M., Müller, H. M., Rüfle, chain, the data presented suggest that as increased costs and risks related to ex- M., Suchy, S., Plank, S., and Drochner, W. grain moves along the chain, its quality is ecuting any further restrictions. A more Natural occurrence of 16 Fusarium toxins improved by the application of manage- strict management, including the moni- in grains and feedstuffs of plant origin ment strategies such as blending, clean- toring of grain along the handling chain from Germany. Mycopathologia 161:43, ing, and selecting. These strategies allow for DON, would require resources for 2006. not only for the improvement of the qual- proper sampling and analytical testing, 6. Scott, P. M. Multi-year monitoring of ity of the grain, but also its safety, by man- lead to an increase in risks to growers, Canadian grains and grain-based foods aging the presence of DON. Indeed, nu- and cause changes at the processor level for trichothecenes and zearalenone. Food merous studies with the major grains, that also would impact grain to be ex- Addit. Contam. 14:333, 1997. 7. National Grain and Feed Association. including wheat, maize, and oats, have ported. All of these effects would add FDA’s advisory levels for deoxynivalenol shown that DON in whole grains is main- costs to the grain-handling chain. Addi- (vomitoxin). Page 6 in: FDA Mycotoxin ly redistributed during processing into the tionally, improvements in management Regulatory Guidance: A Guide for Grain bran and germ fractions rather than the of DON at the farm, elevator, and proces- Elevators, Feed Manufacturers, Grain Pro- endosperm. The flour fraction (from the sor levels have shown that any regulatory cessors and Exporters. National Grain and endosperm) destined for human con- limits would more efficiently safeguard Feed Association, Washington, DC, 2011. sumption typically contains DON levels public health if applied to the finished 8. Pestka, J. J. Toxicological mechanisms and that are 10–20 times lower than those product rather than to unprocessed raw potential health effects of deoxynivalenol observed in the bran or germ fractions, grains. and nivalenol. World Mycotoxin J. 3:323, which are mostly used for animal feed. 2010. 9. The technology or equipment used in the Acknowledgments Pestka, J. J. Deoxynivalenol: Mechanisms of action, human exposure, and toxico- mill affects the recovery and segregation Juan Carlos Batista (SENASA, Argentina); logical relevance. Arch. Toxicol. 84:663, of DON in each fraction. Therefore, it Benedict Deefholts (Buhler Sortex Limited, UK); Wiana Louw (SAGL, Southern African 2010. would be plausible to adopt different Grain Laboratory NPC, South Africa); Rodrigo 10. Yoshizawa, T., and Morooka, N. Deoxyni- standards or regulatory limits for differ- Mendoza (University of Nebraska – Lincoln, valenol and its monoacetate: New myco- ent milled fractions of grains instead of NE, USA); Donald Mennel (Mennel Milling toxins from Fusarium roseum and moldy applying a single level for the whole grain. Company, Fostoria, OH, USA); National Grain barley. Agric. Biol. Chem. 37:2933, 1973. That is, if action levels are to be endorsed, and Feed Association (Washington, DC, USA); 11. Tatsuno, T., Saito, M., Enomoto, M., and they should be endorsed for 1) flour and Terry Nelsen (AACCI Statistician, USA); Tsunoda, H. Nivalenol a toxic principle other milling fractions intended for direct North American Export Grain Association of Fusarium nivale. Chem. Pharm. Bull. human consumption; and 2) bran, germ, (Washington, DC, USA); North American (Tokyo) 16:2519, 1968. 12. Vesonder, R. F., Ciegler, A., and Jensen, and other fractions intended for feeding Millers Association; Lauren Robin (FDA/ A. H. Isolation of the emetic principle to various animal populations. CFSAN, USA); Luis Eduardo Sabillon (Uni- versity of Nebraska – Lincoln, NE, USA); Paul from Fusarium-infected corn. Appl. MLs should be adequately protective of Schwarz (North Dakota State University, ND, Microbiol. 26:1008, 1973. health yet also practically achievable so USA); U.S. Grains Council (Washington, DC, 13. Luo, X. Food poisoning caused by Fu- that trade disruptions do not occur. Since USA); U.S. Wheat Associates (Arlington, VA, sarium toxins. Page 129 in: Proceedings unprocessed raw grains are not typically USA); Felicia Wu (Michigan State University, of the Second Asian Conference on Food consumed in that form and contribute MI, USA). Safety. International Life Sciences Insti- minimally to dietary exposure from the Data were provided by Viterra Inc., Pioneer tute, Chatuchak, Thailand, 1994. 14. finished food product (after processing), Grain Company Operations, and Richardson Lun, A. K., Young, L. G., and Lumsden, J. H. The effects of vomitoxin and feed setting MLs for unprocessed raw grain International Corporate Quality Assurance and Food Safety. intake on the performance and blood would not significantly change health out- characteristics of young pigs. J. Anim. comes. References Sci. 61:1178, 1985. In the grain supply chain, current prac- 1. U.S. Department of Agriculture Foreign 15. Forsyth, D. M., Yoshizawa, T., Morooka, tices for managing risks of excessive DON Agricultural Service. Grain: World Mar- N., and Tuite, J. Emetic and refusal ac- content in grain include 1) targeting loca- kets and Trade. Published online at http:// tivity of deoxynivalenol to swine. Appl. tions with less incidence of DON; 2) in- apps.fas.usda.gov/psdonline/circulars/ Environ. Microbiol. 34:547, 1977.

CEREAL FOODS WORLD / 51 16. European Food Safety Authority. Deoxy- thecenes with a focus on DON: Sum- 41. Broggi, L. E., Pacin, A. M., Gasparovic, nivalenol in food and feed: Occurrence mary report. Toxicol. Lett. 153:1, 2004. A., Sacchi, C., Rothermel, A., Gallay, A., and exposure. EFSA J. 11:3379, 2013. 30. Amuzie, C. J., and Pestka, J. J. Suppres- and Resnik, S. Fumonisin occurrence in 17. Robbana-Barnat, S., Lafarge-Frayssinet, C., sion of insulin-like growth factor acid- naturally contaminated wheat grain har- Cohen, H., Neish, G. A., and Frayssinet, C. labile subunit expression—A novel mech- vested in Argentina. Food Control 23:59, Immunosuppressive properties of deoxy- anism for deoxynivalenol-induced growth 2007. nivalenol. Toxicology 48:155, 1988. retardation. Toxicol. Sci. 113:412, 2010. 42. Cendoya, E., Monge, M. P., Palacios, 18. Casarett, L. J., Doull, J., and Klaassen, C. D. 31. Bonnet, M. S., Roux, J., Mounien, L., Dal- S. A., Chiacchiera, S. M., Torres, A. M., Casarett and Doull’s Toxicology: The Basic laporta, M., and Troadec, J. D. Advances Farnochi, M. C., and Ramirez, M. L. Fu- Science of Poisons, 8th ed. C. D. Klaassen, in deoxynivalenol toxicity mechanisms: monisin occurrence in naturally contami- ed. McGraw-Hill Education, New York, The brain as a target. Toxins (Basel) nated wheat grain harvested in Argentina. 2013. 4:1120, 2012. Food Control 37:56, 2014. 19. Lake, B. G., Phillips, J. C., Walters, D. G., 32. Pinton, P., and Oswald, I. P. Effect of de- 43. Ediage, E. N., Hell, K., and De Saeger, S. Bayley, D. L., Cook, M. W., Thomas, L. V., oxynivalenol and other type B trichothe- A comprehensive study to explore differ- Gilbert, J., Startin, J. R., Baldwin, N. C., cenes on the intestine: A review. Toxins ences in mycotoxin patterns from agro- and Bycroft, B. W. Studies on the metabo- (Basel) 6:1615, 2014. ecological regions through maize, peanut, lism of deoxynivalenol in the rat. Food 33. Gräfenhan, T., Patrick, S. K., Roscoe, and cassava products: A case study, Cam- Chem. Toxicol. 25:589, 1987. M., Trelka, R., Gaba, D., Chan, J. M., eroon. J. Agric. Food Chem. 62:4789, 2014. 20. Prelusky, D. B., and Trenholm, H. L. Non- McKendry, T., Clear, R. M., and Tittlemier, 44. Ennouari, A., Sanchis, V., Marín, S., accumulation of residues in swine con- S. A. Fusarium damage in cereal grains Rahouti, M., and Zinedine, A. Occur- suming deoxynivalenol-contaminated from Western Canada. 1. Phylogenetic rence of deoxynivalenol in durum wheat diets. J. Food Sci. 57:801, 1991. analysis of moniliformin-producing Fu- from Morocco. Food Control 32:115, 21. Prelusky, D. B., and Trenholm, H. L. Tis- sarium species and their natural occur- 2013. sue distribution of deoxynivalenol in rence in mycotoxin-contaminated wheat, 45. Jajić, I., Jurić, V., and Abramović, B. First swine dose intravenously. J. Agric. Food oats, and rye. J. Agric. Food Chem. 61: survey of deoxynivalenol occurrence in Chem. 39:748, 1991. 5425, 2013. crops in Serbia. Food Control 19:545, 22. Prelusky, D. B., Hamilton, R. M., and 34. Tittlemier, S. A., Roscoe, M., Trelka, R., 2008. Trenholm, H. L. Transmission of residues Gaba, D., Chan, J. M., Patrick, S. K., 46. Jakšić, S., Abramović, B., Jajić, I., Baloš, to eggs following long-term administra- Sulyok, M., Krska, R., McKendry, T., and M. Ž., Mihaljev, Ž., Despotović, V., and tion of 14C-labelled deoxynivalenol to Gräfenhan, T. Fusarium damage in small Šojić, D. Co-occurrence of fumonisins laying hens. Poultry Sci. 68:744, 1989. cereal grains from Western Canada. and deoxynivalenol in wheat and maize 23. Rotter, B. A., Prelusky, D. B., and Pestka, 2. Occurrence of Fusarium toxins and harvested in Serbia. Bull. Environ. J. J. Toxicology of deoxynivalenol (vomi- their source organisms in durum wheat Contam. Toxicol. 89:615, 2012. toxin). J. Toxicol. Environ. Health Part A harvested in 2010. J. Agric. Food Chem. 47. Ji, F., Xu, J., Liu, X., Yin, X., and Shi, J. 48:1, 1996. 61:5438, 2013. Natural occurrence of deoxynivalenol 24. Yoshizawa, T., and Morooka, N. Deoxy- 35. Abia, W. A., Warth, B., Sulyok, M., Krska, and zearalenone in wheat from Jiangsu nivalenol and its monoacetate: New my- R., Tchana, A. N., Njobeh, P. B., Dutton, Province, China. Food Chem. 157:393, cotoxins from Fusarium roseum and M. F., and Moundipa, P. F. Determination 2014. moldy barley. Agric. Biol. Chem. 37:2933, of multi-mycotoxin occurrence in cereals, 48. Juan, C., Ritieni, A., and Mañes, J. Occur- 1973. nuts and their products in Cameroon by rence of Fusarium mycotoxins in Italian 25. Yoshizawa, T. Red-mold diseases and liquid chromatography tandem mass cereal and cereal products from organic natural occurrence in Japan. Page 195 in: spectrometry (LC-MS/MS). Food Con- farming. Food Chem. 141:1747, 2013. Trichothecenes, Chemical, Biological, and trol 31:438, 2013. 49. Kos, J., Hajnal, E. J., Škrinjar, M., Toxicological Aspects. Y. Ueno, ed. Kodan- 36. Alexa, E., Dehelean, C. A., Poiana, M.-A., Mišan, A., Mandić, A., Jovanov, P., and sha Ltd., Tokyo, 1983. Radulov, I., Cimpean, A.-M., Bordean, Milovanović, I. Presence of Fusarium 26. Miller, J. D. Mycotoxins in small grains D.-M., and Pop, G. The occurrence of toxins in maize from Autonomous Prov- and maize: Old problems, new challenges. mycotoxins in wheat from western ince of Vojvodina, Serbia. Food Control Food Addit. Contam. Part A Chem. Anal. Romania and histopathological impact 46:98, 2014. Control Expo. Risk. Assess. 25:219, 2008. as effect of feed intake. Chem. Central J. 50. Martínez, M. I., Moschini, R. C. Cazenave, 27. Canady, R. A., Coker, R. D., Rgan, S. K., 7:99, 2013. G., and Martin, N. H. Fusariosis de la Krska, R., Kuiper-Goodman, T., Olsen, 37. Al-Hazmi, N. A. Fungal flora and deoxy- espiga de trigo en la región pampeana en M., Pestka, J. J., Resnik, S., and Schlatter, nivalenol (DON) level in wheat from la campaña 2013/14, comparado con el J. Deoxynivalenol, safety evaluation of Jeddah market, Saudi Arabia. Afr. J. nivel epidémico estimado en 2012/14. certain mycotoxins in food. Page 420 in: Biotechnol. 10:168, 2011. Instituto Argentino de Sanidad y Calidad 56th Report of the Joint FAO/WHO Expert 38. Alkadri, D., Rubert, J., Prodi, A., Pisi, A., Vegetal (IASCAV) de Argentina, Buenos Committee on Food Additives. WHO Food Mañes, J., and Soler, C. Natural co-occur- Aires, 2014. Additives Series 47. World Health Organi- rence of mycotoxins in wheat grains from 51. Martinez, M., Castañares, E., Dinolfo, zation, Geneva, Switzerland, 2001. Italy and Syria. Food Chem. 157:111, 2014. M. I., Pacheco, W. G., Moreno, M. V., and 28. JECFA. Summary of toxicological evalua- 39. Barthel, J., Gottschalk, C., Rapp, M., Stenglein, S. A. Fusarium graminearum tions. Page 2 in: Summary Report of the Berger, M., Bauer, J., and Meyer, K. Oc- presence in wheat samples for human 72nd Meeting of the Joint FAO/WHO Ex- currence of type A, B and D trichothe- consumption. (In Spanish, with English pert Committee on Food Additives. Pub- cenes in barley and barley products from summary.) Rev. Argent. Microbiol. lished online at www.who.int/foodsafety/ the Bavarian market. Mycotoxin Res. 46(1):41, 2014. chem/summary72_rev.pdf. Food and 28:97, 2012. 52. Mirabolfathy, M. Deoxynivalenol and Agriculture Organization of the United 40. Bensassi, F., Zaied, C., Abid, S., Hajlaoui, DON-producing Fusarium graminearum Nations, Rome, 2010. M. R., and Bacha, H. Occurrence of de- isolates in wheat and barley crops in north 29. Larsen, J. C., Hunt, J., Perrin, I., and oxynivalenol in durum wheat in Tunisia. and northwest areas of Iran. Iran J. Plant Ruckenbauer, P. Workshop on tricho- Food Control 21:281, 2010. Pathol. 48:197, 2013.

52 / JANUARY–FEBRUARY 2015, VOL. 60, NO. 1 53. Mirabolfathy, M., and Aliakbari, F. Natu- toxin Res. 27:105, 2011. Fusarium head blight infected malting ral deoxynivalenol contamination of corn 64. Tran, S. T., and Smith, T. K. A survey of barley. J. Am. Soc. Brew. Chem. 64:1, produced in Golestan and Moqan areas free and conjugated deoxynivalenol in the 2006. in Iran. J. Agric. Sci. Technol. 12:233, 2009, 2010 and 2011 cereal crops in Aus- 76. U.S. Grains Council. Corn export cargo 2010. tralia. Anim. Prod. Sci. 53:407, 2013. quality report 2013/14. Published on- 54. Mishra, S., Ansari, K. M., Dwivedi, P. D., 65. Tutelyan, V. A., Zakharova, L. P., Sedova, line at www.grains.org/key-issues/corn- Pandey, H. P., and Das, M. Occurrence of I. B., Perederyaev, O. I., Aristarkhova, export-cargo-quality-report. U.S. Grains deoxynivalenol in cereals and exposure T. V., and Eller, K. I. Fusariotoxins in Council, Washington, DC, 2014. risk assessment in Indian population. Russian Federation 2005–2010 grain 77. Stewart, G., and Tenuta, A. 2014 Grain Food Control 30:549, 2013. harvests. Food Addit. Contam. Part B corn ear mould and vomitoxin survey. 55. Pleadin, J., Vahčić, N., Perši, N., Ševelj, D., Surveill. 6:139, 2013. Published online at http://fieldcropnews. Markov, K., and Frece, J. Fusarium myco- 66. Van der Fels-Klerx, H. J., de Rijk, T. C., com/2014/10/2014-grain-corn-ear-mould- toxins’ occurrence in cereals harvested Booij, C. J., Goedhart, P. W., Boers, E. A., and-vomitoxin-survey. Field Crop News, from Croatian fields. Food Control 32:49, Zhao, C., Waalwijk, C., Mol, H. G., and October 27, 2014. 2013. van der Lee, T. A. Occurrence of Fusari- 78. Stewart, G., and Tenuta, A. 2013 Grain 56. Probst, C., Bandyopadhyay, R., and Cotty, um head blight species and Fusarium corn ear mould and vomitoxin survey. P. J. Diversity of aflatoxin-producing fun- mycotoxins in winter wheat in the Neth- Published online at http://fieldcropnews. gi and their impact on food safety in sub- erlands in 2009. Food Addit. Contam. com/2013/10/grain-corn-ear-mould-and- Saharan Africa. Int. J. Food Microbiol. Part A Chem. Anal. Control, Expo. Risk vomitoxin-survey. Field Crop News, Oc- 174:113, 2014. Assess. 29:1716, 2012. tober 15, 2013. 57. Rubert, J., Dzuman, Z., Vaclavikova, M., 67. Wagacha, J. M., Steiner, U., Dehne, H.-W., 79. Stewart, G., and Tenuta, A. 2012 Grain Zachariasova, M., Soler, C., and Hajslova, Zuehlke, S., Spiteller, M., Muthomi, J., corn ear mould and vomitoxin survey. J. Analysis of mycotoxins in barley using and Oerke, E.-C. Diversity in mycotoxins Published online at http://fieldcropnews. ultra high liquid chromatography high and fungal species infecting wheat in com/2012/10/2012-grain-corn-ear-mould- resolution mass spectrometry: Compari- Nakuru District, Kenya. J. Phytopathol. and-vomitoxin-survey. Field Crop News, son of efficiency and efficacy of different 158:527, 2010. October 4, 2012. extraction procedures. Talanta 99:712, 68. Moss, M. O. The environmental factors 80. Stewart, G., and Tenuta, A. 2011 Grain 2012. controlling mycotoxin formation. In: My- corn ear mould and vomitoxin survey. 58. Savi, G. D., Piacentini, K. C., Tibola, C. S., cotoxins and Animal Foods. J. E. Smith CROP-PEST Ontario 16(15), 2011. and Scussel, V. M. Mycoflora and deoxy- and R. S. Henderson, eds. CRC Press, 81. McMullen, M., Bergstrom, G., De Wolfe, nivalenol in whole wheat grains (Triticum Boca Raton, FL, 1991. E., Dill-Macky, R., Hershman, D., Shaner, aestivum L.) from southern Brazil. Food 69. Tittlemier, S. A., Gaba, D., and Chan, J. M. G., and Van Sanford, D. A united effort to Addit. Contam. Part B Surveill. 7:232, Monitoring of Fusarium trichothecenes fight an enemy of wheat and barley: Fu- 2014. in Canadian cereal grain shipments from sarium head blight. Plant Dis. 96:1712, 59. Schollenberger, M., Müller, H. M., Rüfle, 2010 to 2012. J. Agric. Food Chem. 61: 2012. M., Suchy, S., Plank, S., and Drochner, W. 7412, 2013. 82. Hooker, D. C., Schaafsma, A. W., and Natural occurrence of 16 Fusarium toxins 70. Wu, F., Bhatnagar, D., Bui-Klimke, T., Tamburic-Illinic, L. Using weather vari- in grains and feedstuffs of plant origin Carbone, I., Hellmich, R., Munkvold, G., ables pre- and post-heading to predict from Germany. Mycopathologia 161:43, Paul, P., Payne, G., and Takle, E. Climate deoxynivalenol content in winter wheat. 2006. change impacts on mycotoxin risks in U.S. Plant Dis. 86:611, 2002. 60. Sifuentes dos Santos, J., Souza, T. M., maize. World Mycotoxin J. 4:79, 2011. 83. Biandino, M., Haidukowski, M., Pascale, Ono, E. Y. S., Hashimoto, E. H., Bassoi, 71. Puri, K. D., and Zhong, S. The 3ADON M., Plizzan, L., Scudellan, D., and Reyneri, M. C., Zavariz de Miranda, M., Itano, population of Fusarium graminearum A. Integrated strategies for the control of E. N., and Hirooka, E. Y. Natural occur- found in North Dakota is more aggressive Fusarium head blight and deoxynivalenol rence of deoxynivalenol in wheat from and produces a higher level of DON than contamination in winter wheat. Field Paraná State, Brazil and estimated daily the prevalent 15ADON population in Crops Res. 133:139, 2012. intake by wheat products. Food Chem. spring wheat. Phytopathology 100:1007, 84. Wegulo, S. N., Bockus, W. W., Nopsa, J. H., 138:90, 2013. 2010. De Wolfe, E. D., Eskridge, K. M., Peiris, 61. Stryk, J., Křížová, L., Pozdíšek, J., and 72. Cowger, C., Patton-Ozkurt, J., Brown- K. H. S., and Dowell, F. E. Effects of inte- Pavlok, S. Occurrence of deoxynivalenol Guedira, G., and Perugini, L. Post-anthe- grating cultivar resistance and fungicide and zearalenone in maize silages in the sis moisture increased Fusarium head application on Fusarium head blight and Czech Republic—A year survey. Page 127 blight and deoxynivalenol levels in North deoxynivalenol in winter wheat. Plant in: 15th International Conference, Forage Carolina winter wheat field experiment. Dis. 95:554, 2011. Conservation, High Tatras, Slavakia. L. Phytopathology 99:320, 2009. 85. Koch, H.-J., Pringas, C., and Maerlaender, Rajčáková, ed. Published online at www. 73. Cowger, C., and Arellano, C. Fusarium B. Evaluation of environmental and man- cabdirect.org/abstracts/20143195008. graminearum infection and deoxynivale- agement effects on Fusarium head blight html. CABI, Wallingford, U.K., 2013. nol concentrations during development infection and deoxynivalenol concentra- 62. Bilal, T., Aksakal, D. H., Sünnetci, S., of wheat spikes. Phytopathology 103:460, tion in the grain of winter wheat. Eur. J. Keser, O., and Eseceli, H. Detection of 2013. Agron. 24:357, 2006. aflatoxin, zearalenone and deoxynivale- 74. Canadian Grain Commission. Fusarium 86. Pirgozliev, S. R., Edwards, S. G., Hare, nol in some feed and feedstuffs in Turkey. head blight in Canada, maps and charts. M. C., and Jenkinson, P. Strategies for Pak. Vet. J. 34:459, 2014. Published online at www.grainscanada. the control of Fusarium head blight in 63. Tangni, E. K., Motte, J.-C., Callebaut, A., gc.ca/str-rst/fusarium/fhbmc-feccg-en. cereals. Eur. J. Plant Pathol. 109:731, Chandelier, A., De Schrijver, M., and htm. Canadian Grain Commission, Win- 2003. Pussemier, L. Deoxynivalenol loads in nipeg, MB, Canada, 2014. 87. Dill-Macky, R., and Jones, R. K. The ef- matched pair wheat samples in Belgium: 75. Schwarz, P. B., Horsley, R. D., Steffenson, fect of previous crop residues and tillage Comparison of ELISA VERATOX kit B. J., Salas, B., and Barr, J. M. Quality on Fusarium head blight of wheat. Plant against liquid chromatography. Myco- risks associated with the utilization of Dis. 84:71, 2000.

CEREAL FOODS WORLD / 53 88. Miller, J. D., Culley, J., Fraser, K., Hubbard, FHB by growers in the north central re- 113. Novozymes. Taegro product description. S., Meloche, F., Quellet, T., Seaman, W. L., gion. Page 3 in: Proceedings from the 2011 Published online at www.bioag.novozymes. Seifer, K. A., Turkington, K., and Voldeng, National Fusarium Head Blight Forum. com/en/products/unitedstates/biocontrol/ H. Effect of tillage practice on Fusarium S. Canty, A. Clark, A. Anderson-Scully, Pages/Taegro.aspx. Novozymes, Franklin- head blight of wheat. Can. J. Plant Pathol. and D. Van Sanford, eds. University of ton, NC, 2015. 20:95, 1998. Kentucky, Lexington, KY, 2011. 114. ScabUSA.org. Scab Smart Management. 89. Munkvold, G. P. Cultural and genetic 103. McKendry, A. Native resistance: An es- Published online at www.scabsmart.org. approaches to managing mycotoxins in sential building block for accelerating the ScabUSA.org, 2009. maize. Annu. Rev. Phytopathol. 41:99, development of scab resistant soft red 115. Schaafsma, A. W., and Hooker, D. C. Cli- 2003. winter wheat. Cereal Res. Commun. matic models to predict occurrence of 90. Woloshuk, C., and Wise, K. Diseases of 36(Suppl. B):135, 2008. Fusarium toxins in wheat and maize. Int. corn Gibberella ear rot. Purdue Ext. Bull. 104. Stack, R. W., Frohberg, R. C., and Casper, J. Food Microbiol. 119:116, 2007. BP-77-W, 2010. H. H. Reaction of spring wheats incorpo- 116. Molineros, J. E., De Wolf, E. D., Madden, 91. Rudd, J. C., Horsley, R. D., McKendry, rating Sumai #3-derived resistance to in- L. V., Paul, P., and Lipps, P. E. Incorpora- A. L., and Elias, E. Host plant resistance oculation with seven Fusarium species. tion of host reaction and crop residue genes for Fusarium head blight: Sources, Cereal Res. Commun. 25:667, 1997. level into prediction models for Fusari- mechanisms, and utility in conventional 105. McCartney, C. A., Somers, D. J., Fedak, um head blight. Page 119 in: Proceedings breeding systems. Crop Sci. 41:620, 2000. G., DePauw, R. M., Thomas, J., et al. The of the 2005 National Fusarium Head 92. Fox, S. L., Humphreys, G., Brown, P. D., evaluation of FHB resistance QTLs intro- Blight Forum. S. M. Canty, T. Boring, McCallum, B. D., Fetch, T. G., Menzies, gressed into elite Canadian spring wheat J. Wardwell, L. Siler, and R. W. Ward, eds. J. G., Gilbert, J. A., Fernandez, M. R., germplasm. Mol. Breed. 20:209, 2007. Michigan State University, East Lansing, Despin, T., and Niziol, D. Cardale hard 106. Mesterházy, Á., Tóth, B., Varga, M., MI, 2005. red spring wheat. Can. J. Plant Sci. 93: Bartók, T., Szabó-Hevér, Á., Farády, L., 117. De Wolf, E. D., Madden, L. V., and Lipps, 307, 2013. and Lehoczki-Krsjak, S. Role of fungi- P. E. Risk assessment models for wheat 93. Smith, K. P., Budde, R., Dill-Macky, R., cides, application of nozzle types, and Fusarium head blight epidemics based on Rasmusson, D. C., Schiefelbein, E., the resistance level of wheat varieties in within-season weather data. Phytopathol- Steffenson, B., Wiersma, J. J., Wiersma, the control of Fusarium head blight and ogy 93:428, 2003. J. V., and Zhang. B. Registration of ‘Quest’ deoxynivalenol. Toxins (Basel) 3:1453, 118. Weather INnovations. Weather Central. spring malting barley with improved re- 2011. Published online at www.weathercentral. sistance to Fusarium head blight. J. Plant 107. Horsley, R. D., Pederson, J. D., Schwarz, ca. Weather INnovations, Chatham, ON, Reg. 7:125, 2012. P. B., McKay, K., Hochhalter, M. R., and Canada, 2014. 94. Mergoum, M., Simsek, S., Frohberg, R., McMullen, M. P. Integrated use of tebu- 119. Fusarium Head Blight Prediction Center. Rasmussen, J., Friesen, T., and Adhikari, conazole and Fusarium head blight-resis- Fusarium Risk Assessment Tool. Pub- T. ‘Barlow’: A high-quality and high- tant barley genotypes. Agron. J. 98:194, lished online at www.wheatscab.psu.edu/ yielding hard red spring wheat cultivar 2006. riskTool.html. 2012. adapted to the North Central Plains of 108. Mesterhazy, A., Bartok, T., and Lamper, 120. U.S. Wheat and Barley Scab Initiative. the USA. J. Plant Reg. 5:62, 2011. C. Influence of wheat cultivar, species of FHB forecasting models. Published on- 95. Fowler, D. B. CDC Buteo hard red winter Fusarium, and isolate aggressiveness on line at http://scabusa.org/research_ wheat. Can. J. Plant Sci. 90:707, 2010. the efficacy of fungicides for control of mgmt#mgmt_pred-models. USWBSI, 96. Griffey, C., Thomason, W., Pitman, R., Fusarium head blight. Plant Dis. 87:1107, 2015. Beahm, B., Paling, J., et al. Registration of 2003. 121. Miller, J. D., Young, J. C., and Sampson, Jamestown wheat. J. Plant Reg. 4:28, 2010. 109. Bradley, C. A., and McMullen, M. P. Fun- D. C. Deoxynivalenol and Fusarium 97. Hatchett, J., Graybosch, R., Baenziger, P., gicides for FHB management: Past, pres- head blight resistance in spring cereals. Baltensperger, D., Watkins, J., et al. Regis- ent and future. Page 12 in: Proceedings from Phytopathol. Z. 113:359, 1985. tration of NE01643 wheat. J. Plant Reg. the 2008 National Fusarium Head Blight 122. Boutigny, A. L., Richard-Forget, F., and 2:36, 2008. Forum. S. Canty, A. Clark, E. Walton, Barreau, C. Natural mechanisms for 98. Mergoum, M., Frohberg, R. C., Stack, D. Ellis, J. Mundell, and D. Van Sanford, cereal resistance to the accumulation of R. W., Rasmussen, J. W., and Friesen, T. L. eds. University of Kentucky, Lexington, Fusarium trichothecenes. Eur. J. Plant Registration of Faller spring wheat. J. Plant KY, 2008. Pathol. 121:411, 2008. Reg. 2:224, 2008. 110. Paul, P. A., Lipps, P. E., Hershman, D. E., 123. Foroud, N. A., McCormick, S. P., Mac- 99. McKendry, A. L., Tague, D. N., Wright, McMullen, M. P., Draper, M. A., and Millan, T., and Badea, A. Greenhouse R. L., Tremain, J. A., and Conley, S. P. Madden, L. V. Efficacy of triazole-based studies reveal increased aggressiveness Registration of Truman wheat. Crop Sci. fungicides for Fusarium head blight and of emergent Canadian Fusarium grami- 45:421, 2005. deoxynivalenol control in wheat: A multi- nearum chemotypes in wheat. Plant Dis. 100. Ibrahim, A. M. H., Haley, S. D., Jin, Y., variate meta-analysis. Phytopathology 92: 96:1271, 2012. Langham, M. A. C., Stymiest, C., et al. 800, 2008. 124. Gilbert, J., Clear, R. M., Ward, T. J., Gaba, Registration of Expedition wheat. Crop 111. Schisler, D. A., Core, A. B., Boehm, M. J., D., Tekauz, A., Turkington, T. K., Woods, Sci. 44:1470, 2004. Horst, L., Krause, C., Dunlap, C. A., and S. M., Nowicki, T., and O’Donnell, K. 101. Tamburic-Ilincic, L., Falk, D. E., and Rooney, A. P. Population dynamics of the Relative aggressiveness and production Schaafsma, A. Fusarium ratings in On- Fusarium head blight biocontrol agent of 3- or 15-acetyl deoxynivalenol and tario Winter Wheat Performance Trial Cryptococcus flavescens OH 182.9 on deoxynivalenol by Fusarium grami- (OWWPT) using an index that combines wheat anthers and heads. Biol. Control nearum in spring wheat. Can. J. Plant Fusarium head blight symptoms and de- 70:17, 2014. Pathol. 32:146, 2010. oxynivalenol levels. Czech J. Genet. Plant 112. Yuen, G. Y., and Schoneweis, S. D. Strate- 125. Salgado, J. D., Wallhead, M., Madden, Breed. 47:S115, 2011. gies for managing Fusarium head blight L. V., and Paul, P. A. Grain harvesting 102. Anderson, J. A., Glover, K., and Mergoum, and deoxynivalenol accumulation in strategies to minimize grain quality losses M. Successful adoption of spring wheat wheat. Int. J. Food Microbiol. 119:126, due to Fusarium head blight in wheat. cultivars with moderate resistance to 2007. Plant Dis. 95:1448, 2011.

54 / JANUARY–FEBRUARY 2015, VOL. 60, NO. 1 126. Tkachuk, R., Dexter, J. E., Tipples, K. H., 140. Lancova, K., Hajslova, J., Kostelanska, M., Food Addit. Contam. Part A Chem. Anal. and Nowicki, T. W. Removal by specific Kohoutkova, J., Nedelnik, J., Moravcova, Control Expo. Risk Assess. 28:1694, 2011. gravity table of tombstone kernels and H., and Vanova, M. Fate of trichothecene 153. Thammawong, M., Okadome, H., Shiina, associated trichothecenes from wheat mycotoxins during the processing: Mill- T., Nakagawa, H., Nagashima, H., infected with Fusarium head blight. ing and baking. Food Addit. Contam. Nakajima, T., and Kushiro, M. Distinct Cereal Chem. 68:428, 1991. Part A Chem. Anal. Control, Expo. Risk distribution of deoxynivalenol, nivalenol, 127. Harrison, G., Nowicki, T. W., Slate, A. B., Assess. 25:650, 2008. and ergosterol in Fusarium-infected Japa- and Whitaker, T. B. Final report: Canada 141. Scudamore, K., and Patel, S. The fate of nese soft red winter wheat milling frac- Grains Council multi-year study of inci- deoxynivalenol and fumonisins in wheat tions. Mycopathologia 172:323, 2011. dence and levels of ochratoxin A and de- and maize during commercial breakfast 154. Zheng, Y., Hossen, S., Sago, Y., Yoshida, oxynivalenol in wheat and oats milled in cereal production. World Mycotoxin J. M., Nakagawa, H., Nagashima, H., Canada. January 2014. 1:437, 2008. Okadome, H., Nakajima, T., and Kushiro, 128. U.S. Wheat Associates. 2014 Crop Quality 142. Brera, C., Debegnach, F., Grossi, S., and M. Effect of milling on the content of de- Report. Published online at www.uswheat. Miraglia, M. Effect of industrial process- oxynivalenol, nivalenol, and zearalenone org/cropQuality. USW, Arlington, VA, ing on the distribution of fumonisin B1 in in Japanese wheat. Food Control 40:193, 2014. dry milling corn fractions. J. Food Prot. 2014. 129. Hazel, C. M., and Patel, S. Influence of 67:1261, 2004. 155. Wolf, C. E., and Bullerman, L. B. Heat and processing on trichothecene levels. Toxic. 143. Katta, S., Cagampang, A., Jackson, L., and pH alter the concentration of deoxyniva- Lett. 153:51, 2004. Bullerman, L. Distribution of Fusarium lenol in an aqueous environment. J. Food 130. Huff, W., and Hagler, W., Jr. Density seg- molds and fumonisins in dry-milled corn Prot. 61:365, 1998. regation of corn and wheat naturally con- fractions. Cereal Chem. 74:858, 1997. 156. Bullerman, L. B., and Bianchini, A. Stabil- taminated with aflatoxin, deoxynivalenol 144. Park, D. L. Effect of processing on afla- ity of mycotoxins during food processing. and zearalenone. J. Food Prot. 48:416, toxin. Page 173 in: Mycotoxins and Food Int. J. Food Microbiol. 119:140, 2007. 1985. Safety. J. W. DeVries, M. W. Trucksess, 157. Hart, L. P., and Braselton, W. E., Jr. Distri- 131. Seitz, L. M., and Bechtel, D. B. Chemical, and L. S. Jackson, eds. Springer Science+ bution of vomitoxin in dry milled frac- physical, and microscopical studies of scab- Business Media, New York, 2002. tions of wheat infected with Gibberella infected hard red winter wheat. J. Agric. 145. Scudamore, K., Banks, J., and MacDonald, zeae. J. Agric. Food Chem. 31:657, 1983. Food Chem. 33:373, 1985. S. Fate of ochratoxin A in the processing 158. Cheli, F., Pinotti, L., Rossi, L., and D e l l’or t o, 132. Nowicki, T., Gaba, D., Dexter, J., Matsuo, of whole wheat grains during milling and V. Effect of milling procedures on myco- R., and Clear, R. Retention of the Fusari- bread production. Food Addit. Contam. toxin distribution in wheat fractions: A um mycotoxin deoxynivalenol in wheat 20:1153, 2003. review. LWT Food Sci. Technol. 54:307, during processing and cooking of spa- 146. Scott, P., Kanhere, S., Lau, P., Dexter, J., 2013. ghetti and noodles. J. Cereal Sci. 8:189, and Greenhalgh, R. Effects of experimen- 159. Pacin, A. M., and Resnik, S. L. Reduction 1988. tal flour milling and breadbaking on re- of mycotoxin contamination by segrega- 133. Abbas, H., Mirocha, C., Pawlosky, R., and tention of deoxynivalenol (vomitoxin) in tion with sieves prior to maize milling. Pusch, D. Effect of cleaning, milling, and hard red spring wheat. Cereal Chem. 60: Page 219 in: Novel Technologies in Food baking on deoxynivalenol in wheat. Appl. 421, 1983. Science. Integrating Food Science and En- Environ. Microbiol. 50:482, 1985. 147. Tanaka, T., Hasegawa, A., Yamamoto, S., gineering Knowledge into the Food Chain, 134. Scott, P., Kanhere, S., Dexter, J., Brennan, Matsuki, Y., and Ueno, Y. Residues of Fu- vol. 7. Springer Science+Business Media, P., and Trenholm, H. Distribution of the sarium mycotoxins, nivalenol, deoxyni- New York, 2012. trichothecene mycotoxin deoxynivalenol valenol and zearalenone, in wheat and 160. Schollenberger, M., Muller, H. M., Rufle, (vomitoxin) during the milling of natu- processed food after milling and baking. M., Suchy, S., and Drochner, W. Redis- rally contaminated hard red spring wheat J. Food Hyg. Soc. Jpn. 27:653, 1986. tribution of 16 Fusarium toxins during and its fate in baked products. Food Addit. 148. Lee, U. S., Jang, H. S., Tanaka, T., Oh, Y. J., commercial dry milling of maize. Cereal Contam. 1:313, 1984. Cho, C. M., and Ueno, Y. Effect of milling Chem. 85:557, 2008. 135. Young, J. C., Fulcher, R. G., Hayhoe, J. H., on decontamination of Fusarium myco- 161. Scudamore, K. A., Baillie, H., Patel, S., Scott, P. M., and Dexter, J. E. Effect of mill- toxins nivalenol, deoxynivalenol and and Edwards, S. G. Occurrence and fate ing and baking on deoxynivalenol (vomi- zearalenone in Korean wheat. J. Agric. of Fusarium mycotoxins during commer- toxin) content of eastern Canadian wheats. Food Chem. 35:126, 1987. cial processing of oats in the UK. Food J. Agric. Food Chem. 32:659, 1984. 149. Trigo-Stockli, D., Deyoe, C., Satumbaga, Addit. Contam. 24:1374, 2007. 136. Seitz, L., Eustace, W., Mohr, H., Shogren, R., and Pedersen, J. Distribution of deoxy- 162. JECFA. Deoxynivalenol. Page 47 in: Eval- M., and Yamazaki, W. Cleaning, milling, nivalenol and zearalenone in milled frac- uation of Certain Contaminants in Foods. and baking tests with hard red winter tions of wheat. Cereal Chem. 73:388, 1996. WHO Tech. Rep. Ser. 959. Published on- wheat containing deoxynivalenol. Cereal 150. Tanaka, K., Hara, N., Goto, T., and Manabe, line at http://whqlibdoc.who.int/trs/WHO_ Chem. 63:146, 1986. M. Reduction of mycotoxins contamina- TRS_959_eng.pdf. World Health Organi- 137. Dexter, J., Clear, R., and Preston, K. Fu- tion by processing grain. Mycotoxins zation, Geneva, 2011. sarium head blight: Effect on the milling 1999(Suppl. 2):95, 1999. 163. FAO/WHO. Principles and Methods for and baking of some Canadian wheats. 151. Kostelanska, M., Dzuman, Z., Malachova, the Risk Assessment of Chemicals in Food: Cereal Chem. 73:695, 1996. A., Capouchova, I., Prokinova, E., Skeri- Environmental Health Criteria, No. 240. 138. Visconti, A., Haidukowski, E. M., Pascale, kova, A., and Hajslova, J. Effects of mill- Published online at www.who.int/food M., and Silvestri, M. Reduction of deoxy- ing and baking technologies on levels of safety/publications/chemical-food/en. nivalenol during durum wheat processing deoxynivalenol and its masked form de- World Health Organization, Geneva, 2009. and spaghetti cooking. Toxicol. Lett. 153: oxynivalenol-3-glucoside. J. Agric. Food 164. WHO Global Environment Monitoring 181, 2004. Chem. 59:9303, 2011. System (GEMS). GEMS 2006 Cluster 139. Delwiche, S. R., Pearson, T. C., and Brabec, 152. Edwards, S. G., Dickin, E., MacDonald, Diets Database (www.who.int/foodsafety/ D. L. High-speed optical sorting of soft S., Buttler, D., Hazel, C., Patel, S., and chem/ClusterDietsAug06.xls); WHO wheat for reduction of deoxynivalenol. Scudamore, K. Distribution of Fusarium Country Assignments to the 13 Proposed Plant Dis. 89:1214, 2005. mycotoxins in UK wheat mill fractions. GEMS/Food Consumption Cluster Diets

CEREAL FOODS WORLD / 55 (www.who.int/foodsafety/chem/countries. W. R. Bushnell, eds. American Phytopath- in wheat. J. Agribus. 31:181, 2013. pdf). World Health Organization, Geneva, ological Society, St. Paul, MN, 2003. 176. North Dakota State University. Fusari- 2006. 169. Nganje, W. E., Johnson, D. D., Wilson, um head blight (scab) forecasting mod- 165. WHO Global Environment Monitoring W. W., Leistritz, F. L., Bangsund, D. A., els. Published online at www.ag.ndsu. System (GEMS). GEMS 2012 Cluster and Tiapo, N. M. Economic impacts of edu/cpr/plant-pathology/fusarium-head- Diets Database (www.who.int/foodsafety/ Fusarium head blight in wheat and bar- blight-scab-forecasting-models-07-10-14. areas_work/chemical-risks/IEDIcalcula ley: 1988–2000. Agribus. Appl. Econ. Rep. Crop & Pest Report. NDSU, Fargo, ND, tion0217clustersfinal.xlsm); GEMS/ No. 464. North Dakota State University, 2014. Food Cluster Diets 2012 (www.who.int/ Fargo, ND, 2001. 177. North Dakota State University. Small foodsafety/chem/cluster_diets_2012.pdf); 170. Nganje, W. E., Bangsund, D. A., Leistritz, grain disease forecasting model. Pub- GEMS/Food: Food Contamination Moni- F. L., Wilson, W. W., and Tiapo, N. M. lished online at www.ag.ndsu.edu/crop- toring and Assessment Programme Regional economic impacts of Fusarium disease. NDSU, Fargo, ND, 2015. (www.who.int/foodsafety/areas_work/ head blight in wheat and barley. Rev. 178. Wilson, W., and Dahl, B. Grain contract- chemical-risks/gems-food/en). World Agric. Econ. 26:332, 2004. ing strategies: The case of durum wheat. Health Organization, Geneva, 2012. 171. Nganje, W. E., Kaitibie, S., Wilson, W. W., Agribusiness 27:344, 2010. 166. Codex Alimentarius Commission. Pro- Leistritz, F. L., and Bangsund, D. A. Eco- 179. U.S. Department of Agriculture National posed draft: Maximum levels for deoxy- nomic impacts of Fusarium head blight Agricultural Statistics Service. Quick nivalenol (DON) in cereals and cereal- in wheat and barley: 1993–2001. Agribus. Stats Database. Published online at http:// based products and associated sampling Appl. Econ. Rep. No. 528. North Dakota quickstats.nass.usda.gov. USDA-NASS, plans. CX/CF 12/6/9. Published online at State University, Fargo, ND, 2004. Washington, DC, 2015. ftp://ftp.fao.org/codex/meetings/cccf/ 172. Cowger, C., and Sutton, A. L. The South- 180. U.S. Department of Agriculture Foreign cccf6/cf06_09e.pdf. Joint FAO/WHO eastern U.S. Fusarium head blight epi- Agricultural Service. PS&D Database. Pub- Food Standard Programme, Rome, 2012. demic of 2003. Plant Health Prog. DOI: lished online at http://apps.fas.usda.gov/ 167. Codex Alimentarius Commission. Codex 10.1094/PHP-2005-1026-01-RS. 2005. psdonline/psdQuery.aspx. USDA-FAS, Standard 193-1995, Codex general stan- 173. Lilleboe, D. FHB in 2010: An overiew. Washington, DC, 2015. dard for contaminants and toxins in food Published online at http://scabusa.org/ 181. Agriculture and Agri-Food Canada. Can- and feed. Section 1.3.1: Principles re- pdfs/fus-focus_09-10_newsletter.pdf. ada: Outlook for Principal Field Crops. garding contaminants in food and feed: Fusarium Focus 10(2):2, 2010. [Various years.] Published online at www. General. Published online at www.codex 174. Lilleboe, D. Fusarium head blight in 2011: agr.gc.ca/eng/industry-markets-and- alimentarius.org/download/standards/17/ An overview. Published online at http:// trade/statistics-and-market-information/ CXS_193e.pdf. FAO/WHO, Rome, 2013. scabusa.org/pdfs/USWBSI-Article_2011- by-product-sector/crops/crops-market- 168. Johnson, D. D., Flaskerud, G. K., Taylor, Update_9-29-11.pdf. U.S. Wheat U.S. information-canadian-industry/canada- R. D., and Satyanarayana, V. Quantifying Wheat and Barley Scab Initiative, 2011. outlook-for-principal-field-crops/?id= economic impacts of Fusarium head blight 175. McKee, G., Ransom, J., and McMullen, 1378743094676. Agriculture and Agri- in wheat. Page 461 in: Fusarium Head Blight M. Determinants of adoption of Fusari- Food Canada, Ottawa, ON, 2015. of Wheat and Barley. K. J. Leonard and um head blight management techniques 182. Ali, M., and Vocke, G. Consequences for higher input costs and wheat prices for U.S. wheat producers. Outlook No. WHS- 09c-01, March 2009. Published online at AACC International Scientific Initiatives http://www.ers.usda.gov/publications/ whs-wheat-outlook/whs09c01.aspx. U.S. Food Safety Worldwide Department of Agriculture Economic Research Service, Washington, DC, 2009. 183. Johnson, D. D., and Wilson, W. Wheat To learn more about DON in cereals and the keys to its successful global cleaning decisions at country elevators. Am. J. Agric. Econ. 74:1307, 1992. management, visit the AACCI Scientific Initiatives web page on 184. Johnson, D. D., and Wilson, W. Wheat “Food Safety Worldwide” and review the linked presentations. cleaning decisions at country elevators. J. Agric. Resour. Econ. 18:198, 1993. http://www.aaccnet.org/initiatives/Pages/FoodSafetyWorldwide.aspx 185. Wilson, W., Johnson, D. D., and Dahl, B. The economics of grain cleaning on the prairies. Can. J. Agric. Econ. 48:278, 2000.

56 / JANUARY–FEBRUARY 2015, VOL. 60, NO. 1