Contents of trichothecenes in oats during official variety , organic cultivation and nitrogen fertilization trials in Finland

Natural toxins, such as mycotoxins, have emerged as a significant factor affecting the safety image of cereal grains as a raw material for the food and feed industry. The aim of the present study was to investigate the contents of trichothecenes in representative samples of oats during official variety, nitrogen fertilization and organic farming trials in Finland, 1997–1999. Further objectives were to promote industry and commerce by selection of high-quality oat varieties for various applications. The official variety trials conducted at 8–10 locations were managed following standard protocol. There were 2 types of agronomy trial, the first included comparison of oat cultivars grown in conventional and organic farming systems at 6 locations, and the second used 5 nitrogen rates (0, 40, 80, 120 and 160 kg N ha) at 2 locations. Regardless of wet cold summer occurring in Finland during 1998, the concentrations of Fusarium toxins were lowest during this 3 year monitoring period. More mycotoxins were produced during the warm, dry summers of 1997 and 1999 than in 1998. In all, 55% of the oat samples in the official variety trials contained deoxynivalenol (DON) within the range of 50– 896 μg kg. The differences in DON concentrations between organic and conventional cultivation were small. The results showed also that the use of various nitrogen fertilization levels only slightly affected the trichothecene concentrations. The contents of trichothecenes in Finnish grains appeared to be similar to or lower than those reported earlier in the Northern Hemisphere.


Introduction
The central goal of grain cultivation is the production of high-quality food or feed-related raw materials for the processing industry.Natural toxins, such as mycotoxins, have emerged as a significant factor affecting the safety image of cereal grains as a raw material for the food and feed industry.Many previous studies in Finland (Karppanen et al. 1985, Hietaniemi and Kumpulainen 1991, 1993, Rizzo 1993, Eskola et al. 2001, Rizzo et al. 2001) and other European countries (Tanaka et al. 1988, Langseth et al. 1989, 1999, Müller and Schwadorf 1993, Pettersson et al. 1995, Rizzo et al. 2001, Widestrand 2001, Döll et al. 2002, Schollenberger et al. 2002) as well as globally (Tanaka et al. 1988, Shephard et al. 1996, Groves et al. 1999, Janardhana et al. 1999, Salay and Mercadante 2002) have shown that there is reason to focus on Fusarium toxins and their appearance.A planned European Union directive will specify the maximum limits for trichothecenes such as 4-deoxynivalenol (DON), T-2 toxin and HT-2 toxin and for zearalenone (ZEN) and fumonisins.DON is the most frequently found contaminant of oats, barley, wheat and corn throughout the world (Scott 1989, WHO 1993, 2001).In addition to DON, T-2 toxin, HT-2 toxin and ZEN frequently occur in cereal crops cultivated in northern temperate regions (Hietaniemi and Kumpulainen 1991, Pettersson et al. 1995, Langseth et al. 1999, Rizzo et al. 2001, Thuvander et al. 2001, Eskola 2002, Lígia Martins and Martins 2002).On the other hand, fumonisins cause more extensive problems in the Southern Hemisphere than in the Northern (Shephard et al. 1996).The most frequently isolated Fusarium species are F. graminearum, F. culmorum, F. moniliforme, F. poae, F. equiseti and F. proliferatum (Ylimäki et al. 1979, Marasas et al. 1984, Hietaniemi and Kumpulainen 1991, Eriksen and Alexander 1998, Eskola et al. 2001, Creppy 2002).
Numerous studies on mould toxicoses have shown that type A trichothecenes such as T-2 toxin, HT-2 toxin, diacetoxyscirpenol (DAS) and neosolaniol are more acutely toxic, whereas type B trichothecenes such as DON and nivalenol (NIV) are less toxic and are therefore implicated in more chronic toxicoses (Atroshi et al. 2002, Creppy 2002, Gutleb et al. 2002, Vilà et al. 2002, de Vries et al. 2003, Sudakin 2003).For example, the most frequently found of these substances (DON) in cereal crops is known to affect symptoms of intoxication in humans such as abdominal pain or a feeling of fullness in the abdomen, dizziness, headache, throat irritation, nausea, vomiting, diarrhoea, and blood in the stool (Eriksen and Alexander 1998).In the case of intoxication in animals, symptoms such as loss of appetite, reduced weight gain, vomiting, heavy diarrhoea and swelling have been reported (Joffe 1978, Ueno 1987).In addition, DON is a very stable compound occurring during both storage and processing of food and does not degrade at high temperatures (Scott 1991).
Increased attention has been focused on mycotoxins in Finland, especially the late 1980s, Hietaniemi and Kumpulainen (1991) showed that almost all grain and feed samples studied in Finland during 1987-1988 contained from 7 to 300 µ g kg -1 of DON and smaller amounts (13-120 µ g kg -1 ) of 3-acetyldeoxynivalenol.The most toxic trichothecenes, T-2-and HT-2 toxin, as well as NIV and ZEN, were found at low concentrations in some samples.Based on the results of Rizzo (1993) trichothecene concentrations in Finnish grains (n = 292) have ranged 1-6300 µ g kg -1 for DON, 1-1000 µ g kg -1 for NIV and 5-238 µ g kg -1 for T-2.In the study of Eskola et al. (2002) DON was detected in 54 of 68 Finnish cereal samples in the concentration range 5-111 µ g kg -1 .NIV and HT-2 toxin were detected in 3 and 2 samples, respectively, in the concentration range 10-20 µ g kg -1 .In addition, Eskola et al. reported the contents of Fusarium toxins in 47 cereal grain samples collected in different parts of the country.The mean concentrations of DON, 3-AcDON, NIV, HT-2 and ZEN in contaminated samples were 60, 30, 42, 43 and 5 µ g kg -1 , respectively.T-2 was detected only once (23 µ g kg -1 ), while fusarenon X (FX) and diacetoxyscirpenol (DAS) were not detected.

Hietaniemi, V. et al. Contents of trichothecenes in oats
There are differences in susceptibility to Fusarium contamination amongst the various types of grain.Oats are regarded as being more susceptible to Fusarium fungi in Nordic countries, although studies done in Norway (Langseth et al. 2001) suggest that contamination in barley and wheat occurs just as frequently.The species prevalent here and in the other Nordic countries is F. avenaceum, which does not produce DON and becomes especially abundant during rainy, cool summers, such as in 1998.According to Marasas et al. (1984) and Eriksen and Alexander (1998) high temperatures during the growing season favour significant toxin formulators; in particular, F. culmorum, F. graminearum and F. poae proliferate in warm temperatures, which was quite clearly found in Finland in the year 2001 and 2002.A belated harvest increases the growth of Fusarium fungi and toxinrelated risk along with the rise in moisture.Fusarium fungi form toxins during the growing season, in warm-air drying, the fungi are preserved in grain, but trichothecene compounds no longer form under dry storage.
The aim of the present study during 1998-2000 was to investigate the contents of trichothecenes in representative samples of oats during official variety, nitrogen fertilization and organic farming trials in Finland, 1997Finland, -1999. .Further objectives were to enhance producers procedures for growing and harvesting oats and promote industry and commerce by selection of high-quality oat varieties for various applications.

Samples
The oat samples were collected after harvest during official and agronomy trials conducted by MTT Agrifood Research Finland in 1997-1999.The official variety trials conducted at 8-10 locations were managed following standard pro-tocol (Fig. 1).There were 2 types of agronomy trial, the first included comparison of oat cultivars grown in conventional and organic farming systems at 6 locations, and the second used 5 nitrogen rates (0, 40, 80, 120 and 160 kg N ha -1 ) with 4 oat cultivars at 2 locations (Fig. 1).More detailed information on the trials was published previously (Järvi et al. 2000, Eurola et al. 2003).After harvest, the grains were immediately dried with warm air in a flat bed grain drier to a moisture content of below 14%.The oat grains were sorted with a 2.0-mm sieve and hulled with a laboratory hulling machine BT 459 using air pressure.Oat groats were milled with a falling number hammer mill using a 1.0-mm sieve.The total number of oat samples analysed were 147, 147 and 99 in 1997, 1998 and 1999, respectively.The varieties studied were Leila, Kolbu, Salo, Belinda, Ve li, Roope, Aarre, Katri, Puhti and Yty.Kolbu and Roope have yellow husks, and the other varieties have white husks.Leila and Kolbu are cultivars developed in Norway, Salo and Vo l. 13 (2004): 54-67.Belinda in Sweden, and Ve li, Roope, Aarre, Katri, Puhti and Yty in Finland.The oat varieties selected for the project included the most popular cultivars as well as new varieties on the National list of cultivars in Finland.
For example, the variety Ve li was the most widely cultivated variety of oat in Finland from 1997 to 1999.Ve li gives high quantities, is early and possesses good quality.Cultivation-related reliability is favourable and the grain is whitehulled, mid-sized and rather thin-hulled.The protein content and the weight of the grain are high.Ve li is a thriving variety, reliable under cultivation, that also produces good harvests during poor years.Ve li is suitable as an oat for feed, food products and export.The year of National variety catalogue entry for this type, which was developed in Finland by Boreal Plant Breeding Ltd, was 1982.
On the other hand, Salo is a variety developed in Sweden by the Svalöf Weibull Company and emerged on the market in 1989.It has good lodging resistance and was the second most widely cultivated variety of oat in Finland from 1997 to 1999.Salo is a late variety with large grain size, low hull content and pure resistance against dryness in addition to being susceptible to oat leaf blotch.Due to its short, durable stem, Salo withstands 10-20 kg ha -1 of stronger nitrogen fertilization than normal.The main use of Salo has been as a food product.

Sample clean-up, identification and quantification
To determine the presence of mycotoxins, the ground-up samples (25 g) were extracted through a blend of acetonitrile-water (84:16) and then suction-filtered.Part of the filtrate (7.5 ml) was extracted for activated carbon cleaning (My-coSep#227, Romer Labs), after which the sample was transferred to a silylated test tube and evaporated to dryness under a stream of nitrogen.100 µ l N-trimethylsilylimidazole (TMSI) reagent was added to the residue evaporated to dryness.DON, DAS, 3-acetyldeoxynivalenol (3-AcDON), FX, NIV, T-2 and HT-2 toxins and 19nortestosterone (the internal standard) were identified and quantified as trimethylsilylether derivatives, employing GC-MS (Rizzo et al. 1986, Hietaniemi and Kumpulainen 1991, Saastamoinen and Saloniemi 1997).

Quality assurance
The reliability of the GC-MS method in trichothecene analysis was examined in repeatability, reproducibility, recovery and intercalibration studies by specifying the linearity of the method, the minimum detectable concentrations (quantification limit), utilizing certified reference materials and participating in FAPAS round tests.The recoveries for the compounds varied between 70-100% and the repeatability between 5% and 25%, while the minimum detectable concentrations were 50 µ g kg -1 with all trichothecenes.The various linearity areas with respect to trichothecenes were as follows: DON (25-8000 µ g kg -1 ), DAS (25-500 µ g kg -1 ), 3-AcDON (25-800 µ g kg -1 ), FX (25-250 µ g kg -1 ), NIV (25-400 µ g kg -1 ), T-2 (25-600 µ g kg -1 ) and HT-2 (25-1300 µ g kg -1 ).BCR wheat flour, CRM 379 was employed as the reference material (DON concentration 670 µ g kg -1 ).The Chemistry Laboratory of Agrifood Research Finland follows a quality control system in accordance with ISO 17025.In the case of samples whose concentration level was below the limit of quantification, the value reported in the mean value, median and standard deviation calculations was estimated as 50% of the quantification limit.
For statistical analyses Data Desk 6.1.1 (Data exploration and visualization) was used (Velleman and Hoaglin 1981).

Official variety trials
The results of the official variety trials showed that the mycotoxin DON was found most fre-

Hietaniemi, V. et al. Contents of trichothecenes in oats
quently in Finnish oats during 1997-1999.In 1997, the procent of positive toxin findings out of all oat samples studied (51) were as follows: DON 69%, DAS not detected, 3-AcDON 6%, FX not detected, NIV 14%, T-2 toxin 4% and HT-2 toxin 4%.The mean DON concentration of all samples was µ g kg -1 (median 79 µ g kg -1 , standard deviation 206 µ g kg -1 ).A few individual, rather high contents of NIV were found within a range of < 50-575 µ g kg -1 .The mean T-2-and HT-2 toxin concentrations were below the quantification limit of 50 µ g kg -1 .The highest mean DON concentrations were determined in the areas of Hyrylä, Pälkäne, Mikkeli and Laukaa (Fig. 1).The mean contents of DON, amount of precipitation, effective temperature sum values, and pH and the type of soil as obtained for May -August are presented in Table 1.None of the above-mentioned background factors did not explain directly the higher toxin concentrations in these areas.The amounts of precipitation and effective temperature sums were similar to the average values observed in other areas, and not even the soil types could indicate a direct connection with elevated toxin concentrations (Table 1).The varieties, Leila, Salo, Ve li, Belinda, Kolbu and Roope, were employed in the official trials.Figure 2 shows DON contents of six oat

Hietaniemi, V. et al. Contents of trichothecenes in oats
cultivars across all sites of official variety trials during 1997-1999.The highest individual DON concentrations were found in the varieties Ve li and Salo.According to the statistical analyses no distinct differences were found in DON contents of various varieties.It is most likely that the climate had the greatest impact by fixing the time of heavy rainfalls in the right period of the growing season, for example during heading and harvesting time.
During these 3 research years, the lowest mycotoxin concentrations were determined from the oat harvest of 1998 (Table 2).The mean DON content from the 51 samples studied was 99 µ g kg -1 (median < 50 µ g kg -1 , standard deviation 171 µ g kg -1 ).Other toxins were found in only a few samples as small concentrations, e.g., no T-2 toxin content exceeding the quantification limit was found and only one HT-2 toxin concentration in excess of the quantification limit was determined.Summer 1998 was cool and rainy throughout the Finland.Despite this, conditions for the formation of toxins were not favourable.Nevertheless, fusarium head blight visibly appeared in Finnish grain, especially in rye.In the investigations of Eskola et al. (2000 and2001) a lot of mouldiness of the Finnish grain in 1998 was found.Despite of the heavy contamination of the moulds, contents of trichothecenes were low.The findings in 1998 also indicated that Ve li and Salo had the highest individual DON concentrations.The highest DON concentrations were detected in Laukaa, central Finland, where rainfall was also high and the effective temperature sum was low.However, similar amounts of rain were also observed in other areas, and there was no association with elevated toxin concentrations within the same area (Table 1).
As a research year 1999 was quite similar to 1997 (Table 2).In several areas, 1999 was clearly less subject to rain, but the effective temperature sums were similar to those in 1997.The average DON concentrations for the oat samples were, 154 µ g kg -1 , and the number of positive findings out of the 59 samples studied was 37. The largest DON concentrations were detected in the variety Ve li.As in 1998, no positive T-2 toxin results were measured, but 4 positive HT-2 toxin contents were found, the highest value being 240 µ g kg -1 .NIV was found from 14 samples out of a total number of 59, with the concentrations ranging < 50-423 µ g kg -1 .From the regional stand point, the most contaminated oat samples were found in southern, southwestern and eastern Finland (Fig. 1).The most precipitation occurred in eastern Finland within a range of 183-243 mm, and the least was measured in southwestern Finland at 92 mm.The year 1999 also indicated that a direct connection between the amounts of rain, effective temperature sums and contents of mycotoxins does not exist.
It still appears more probable that the 'right' period of rainfall during the growing season exerts greater significance.For example, if heavy rainfalls occurred during heading and harvesting time, the risk of growth of Fusarium fungi and production of toxins would exist.Oldenburg et al. (2000) studied the factors which influence most the infection of grain, especially wheat.The authors found the climate to have the greatest impact, followed by infection pressure/tillage, corn as preceding crop, plant protection, cultivars and plant nutrition.Langseth et al. (2001) reported from Norway that the DON level was especially high in 1988 and 1992, years with spring drought and much precipitation in July in the period during and after anthesis.According to McMullen et al. (1997) during the last two decades fusarium head blight disease occurs especially during rainy flowering seasons of the crop and besides of DON, an infection of the head by Fusarium can lead to severe yield loss and reduced kernel quality.On the other hand, rains occurring at harvest time may increase the risk of appearance of mycotoxins considerably.However, warm-air-drying as primarily employed in Finland is evidently a security-enhancing factor with respect to grain (Jayas and White 2003).
The results showed that the use of various nitrogen fertilization levels during 1997-1999 only slightly affected the mycotoxin concentrations (Fig. 3).The highest mean DON concentrations were observed in the data from 1999 within a range of 101-730 µ g kg -1 (Fig. 3).The DON concentrations in Häme (Jokioinen) were, to some extent, higher than those in South Ostrobothnia (Ylistaro), although the amounts of rainfall were similar.No remarkable differences between the various varieties were found; the highest individual concentrations were determined from the variety Aarre (Fig. 3).Aarre is an early, highyielding and high-quality oat variety and in addition has good lodging resistance.It is earlier than Veli by one day.Aarre is used for feed and food products and export and it is developed by Boreal Plant Breeding Ltd of Finland.It was introduced in the National variety catalogue in 1995.

Organic farming
The trials for conventional as well as organic cultivation were implemented during 1997-1998 at 6 locations: Jokioinen, Mietoinen, Laukaa, Partala, Ylistaro and Ruukki.Only at the Partala Research Station in eastern Finland was conventional cultivation not used as comparative data.
More precise comparison of these cultivation methods could be carried out only with 2 varieties, Ve li and Puhti, although 5 other varieties Aarre, Katri, Kolbu, Leila and Roope were also included in the organic cultivation trials.In contrast to the results obtained in the official variety trials and nitrogen fertilization trials, DON was also the most commonly appearing myco-

Hietaniemi, V. et al. Contents of trichothecenes in oats
toxin in samples from the conventional and organic cultivation trials (Fig. 4).Correspondingly, the results corroborated previous indications that the toxin concentrations during the rainy, cold year of 1998 were clearly lower than in 1997.There were no large differences in DON concentrations in the cultivation methods studied.The mean levels appeared to be similar in organic and conventional cultivation (Fig. 4).
According to Leblanc et al. (2002) a particular group at risk for exposure to DON may be the consumers of organic products as they may regularly eat foods containing higher levels of DON than the general population.On the basis of the work of Schollenberger (2002) the DON content of flour samples originating from conventionally produced wheat was significantly higher than that of samples from organic production.Similar results were obtained by Döll et al. (2000) and Usleber et al. (2000), who found lower DON contents in wheat, rye and flour samples of organic origin than in those from conventional production.In agreement with this study, Marx et al. (1995) reported from southern Germany that the incidence of DON in conventionally and organically produced wheat was 88% and 76%, whereas the mean contents were 420 and 486 µ g kg -1 , respectively.In addition, according to Döll et al. (2000) the organic farming system with the practice of a well-balanced crop rotation, tillage and fertilization showed benefits concerning the contamination with Fusarium toxins, especially DON.

Conclusions
The results of the official variety and nitrogen fertilization trials and comparison of conventional and organic cultivation, suggest that future research and follow-up on trichotecenes must be emphasized and continued.The results of the trials showed that the mycotoxin DON was found most frequently in Finnish oats during 1997-1999.According to the results on an average 55% of the oat samples respective to the official variety trials in 1997-1999 contained DON within the range of 50-896 µ g kg -1 .Correspondingly, the range and frequencies of other toxin findings were as follows: 3-AcDON 50-310 µ g kg -1 (5%), NIV 50-575 µ g kg -1 (16%), T-2 toxin 50-349 µ g kg -1 (1%) and HT-2 toxin 50-507 µ g kg - 1 (4%).A directive currently under preparation in the European Union will specify the tolerance limits for DON (1500 µ g kg -1 ).In comparison to previous studies, the contents of trichothecenes in grains appeared similar or lower to those reported earlier in the Northern or Southern Hemisphere (Karppanen et al. 1985, Tanaka et al. 1988, Hietaniemi and Kumpulainen 1991, Müller and Schwadorf 1993, Rizzo 1993, Pettersson et al. 1995, Groves et al. 1999, Janardhana et al. 1999, Döll et al. 2000, Langseth and Rundberget 2001, Rizzo et al. 2001, Eskola 2002, Schollenberger et al. 2002, Salay and Mercadante 2002).The results also showed that no distinct differences were found in DON contents of various varieties.The differences in DON concentrations between organic and conventional cultivation were small.In addition, the results showed that the use of various nitrogen fertilization levels only slightly affected the trichothecene concentrations.
Nevertheless, the importance of background factors with respect to samples is often forgotten in monitoring the quality of grains, not to mention a deeper familiarity with their impact and, through the same, better control over grain quality.Evidently, the incidence of rain during heading time represents a risk factor (McMullen et al. 1997, Döll et al. 2000, Oldenburg et al. 2000, Langseth et al. 2001), although it has not been indicated here.The present results also showed that more precise research into the effects of cultivation methods in relation to fungi and toxins is necessary (Norred 2000).Good familiarity with, the effects of the preceding crop, rotation, seed purity, various soil preparation methods, direct sowing and pesticides on the formation of mycotoxins is a minimal requirement.A question of its own right also concerns breeding for resistance against Fusarium fungi Vo l. 13 (2004): 54-67. and their associated toxins (de Vries 2000, Hollins et al. 2003).Knowledge of these factors is a prerequisite for good cultivation-related directives that industry and farmers should follow to ensure high-quality production of oats both in Finland and internationally.

Fig. 1 .
Fig. 1.Location of the trial sites in Finland.

Fig. 2 .
Fig. 2. Deoxynivalenol (DON) contents of six oats cultivars across all sites of official variety trials during 1997-1999 in Finland.

Fig. 4 .
Fig. 4. Mean deoxynivalenol (DON) contents in oats during organic and conventional cultivation trials in Finland in 1997 and 1998.

Table 1 .
Precipitation, effective temperature sum, soil pH, soil type and mean deoxynivalenol (DON) content of oats in official variety trials during 1997-1999 at different trial locations in Finland.