March 9, 2022
Trouw Nutrition's global mycotoxin analysis: What 2021 raw materials mean for 2022's feeds

Trouw Nutrition, Nutreco's livestock feed business, is sharing results from a global mycotoxin analysis of complete feeds and raw materials sampled in 2021.
The almost 32,000 samples collected and analysed globally, showed mycotoxin contamination ranging from 18% to 63%, with the highest being aflatoxin (AFLA) contamination and the lowest for T-2/HT2 toxin (T2HT2).
Based on the mycotoxin concentrations (ppb), mycotoxin risk was the highest for pet food and aqua feed followed by ruminants, pigs and poultry feeds. These results show a need for proper sampling and analysis of feeds and raw materials to assess contamination threats and introduce targeted mycotoxin mitigation strategies that respond to emerging concerns.
The need for mycotoxin analysis
Understanding the mycotoxins present in feed ingredients and complete feeds is necessary to provide safe feed for efficient production as toxin contamination can impair animal health and performance. Knowledge of mycotoxin presence also plays a role in safe food production because some mycotoxins can be transmitted into animal proteins including milk, eggs and meat. This transfer has led to regulatory limits in some areas.
When ingredients and feeds are checked for mycotoxin contamination, priority is given to a selection of "Big 6" mycotoxins such as deoxynivalenol (DON), T-2HT2, zearalenone (ZEA), fumonisins (FUM), AFLA and ochratoxin (OTA). All six, and many other mycotoxins belong to one of four toxin groups – fusarium toxins, aspergillus toxins, penicillium toxins and ergot toxins.
Mycotoxin monitoring can be done at several different points in time – after crop harvested in the field, while ingredients are at the feed mill and after a complete feed is mixed. Trouw Nutrition mainly uses the Mycomaster lateral flow technology for the analysis of "Big 6" mycotoxins as well as using LC-MS/MS for silage and TMR analysis.
Mycotoxin interpretation tools such as Mycotoxin Monitor, which provides a regional risk analysis for specific species, and Mycotoxin Adviser, which provides customer- and species-specific feedback, are used by Nutreco colleagues and customers across the world.
What is the difference between % mycotoxin contamination and concentrations?
A few key terms are important for the better understanding of the mycotoxin data interpretation: % contamination refers to the number of commodities, out of 100, that are tested and found to have mycotoxin concentrations at and above the level of detection (LOD). Mycotoxin concentration, on the other hand, means the amount of a specific mycotoxin present in 1kg of feed or raw material and the amount should be at and above limit of quantification (LOQ). It is usually expressed in ppb (µg per kg feed) or ppm (mg per kg feed).
% contamination with mycotoxins indicates the prevalence of mycotoxins in various commodities and regions whereas concentrations help in understanding the potential toxicity of feeding mycotoxin-contaminated commodities to various species of animals in a specific farm.
A feed with the same level of a specific mycotoxin can illicit different toxic responses in different species of animals.
Dataset overview
Close to 32,000 analyses performed in this dataset represented 43 countries grouped into five regions (Figure 1).
The highest % of samples were from Europe followed by North America, LATAM, MEA and Asia regions. Feeds contributed maximum to the dataset (39%) followed by grains (35%), grain by-products, protein meals, silages and other protein products.
Maize, wheat and barley were the major grains tested while soybean meal, sunflower meal and cotton seed meal were the chief protein meals. Many by-products were part of the testing including wheat bran, maize DDGS, maize gluten meal and maize flour.
As in our previous survey, broiler feed, layer feed, pig feeds and ruminant feeds made the bulk of feed samples tested but there were also considerable number of silages, pet food and aqua feeds featured in the 2021 survey.

Analysis of entire dataset
As can be seen in table 1, 89% of the samples tested positive for at least one mycotoxin when LOD filter was not applied to the dataset (table 1).
When the analysis are performed using rapid tests, the instrument will display mycotoxin concentrations, even those below LOD. For example, if LOD is 3ppb for AFLA, the instrument will also display results for 1ppb and 2ppb.
We should not consider samples with 1ppb and 2ppb of AFLA as positive samples. When respective LOD were applied to each mycotoxin, only 51% of samples were found to be contaminated and these are the correct figures. The two most detected mycotoxins across all the samples were AFLA and ZEA, which were present in 63% and 53% of samples, respectively.
It is a routine practice in the industry to use average or mean to describe concentrations of mycotoxins in the commodities. Although this is easy to understand, the chances of data getting skewed is very high. Even if one or two sample out of 1,000 samples tested contain high levels of a mycotoxin, the average of 1,000 samples can be skewed to the upper side.
For example, across the entire data set in this survey, the minimum concentration of AFLA was 3ppb and the maximum found was 2,010ppb (table 1). Given the outliers in the data, the average concentration of AFLA was determined to be about 12ppb.
However, the median, or central number in the data set, was 3ppb. Calculating the mycotoxin risk based on the average concentration may therefore overestimate the toxicity than what is actually occurring.
The Box and Jitter Plot (Figure 2) provides a better picture of mycotoxin concentrations as it incorporates mean, median, quartiles and outliers into the analysis. It is clear from the Figure 2 that the variability within the dataset was the highest for AFLA followed by ZEA, T2HT2 and OTA. Variability was the lowest for FUMO and DON as the mean (diamond shape) was closer to the median (a line within the box).
Table 1. contamination at global level on entire datasheet


Figure 2. The Box and Jitter plot can better explain mycotoxin data as it not only provides mean (diamond shape) but also median and different quartiles.
Regionally, there was a variation in what mycotoxins were present in higher concentrations (Figure 3). For example, Asia reported the highest mycotoxin concentrations for AFLA, FUMO and DON, while North America had higher concentrations of OTA and T2HT2. LATAM reported higher concentrations of ZEA as compared to others.

Figure 3. Mycotoxin concentrations (ppb) per region
Raw material analysis
When looking specifically at raw materials harvested in 2021 (with the potential of these being used in feeds in 2022), primary mycotoxins sought in analyses were present in most feed ingredients. The exceptions are that no samples of soybean meal were tested positive for T-2 and no corn silage was found to have OTA (Figure 4).

Figure 4. Mycotoxin concentrations (ppb) by raw materials
Complete feed analysis
When analysing complete feeds, there tends to be a range of mycotoxin types present and % contamination tends to be higher because of the blend of ingredients (data not shown). In terms of mycotoxin concentrations, AFLA and OTA were the highest in broiler feeds while pet food contained the maximum concentrations of T2HT2 (Figure 5). FUMO and DON concentrations were quite high across the feeds tested while ZEA concentrations were quite low in aqua feeds.

Figure 5. Mycotoxin concentrations (ppb) by feed type
Species risks
Not all species face the same risk from the presence of the same mycotoxin in feed. For example, swine can be highly sensitive to DON, while poultry are not.
However, poultry are more sensitive to T2HT2 than either dairy cows or pigs. In addition to the type and concentration of mycotoxins making a difference, the duration of the exposure, the order in which animals experience mycotoxin negative effects and the presence of other mycotoxins can influence responses to mycotoxin contamination.
Based on the practical mycotoxin guidance values (concentrations), species risk can be determined by the various feeds analysed in 2021 (Table 2).
To reduce the variability in the data, this year's median values (instead of mean values) are considered for risk assessment. Overall, data showed low toxicity for poultry species, followed by low to medium toxicity for pigs and low to high toxicity for ruminant and aqua species. Mycotoxins in dog foods showed maximum risk as dogs are quite sensitive to a number of mycotoxins.
Table 2. 'Colour Card' for 2021 global mycotoxin species risk (median)

Looking into 2022, the assessment of samples collected in late 2021 through early 2022 has indicated high risk for dairy and piglets followed by sows and poultry. This assessment is based on the concentrations of AFLA, DON, FUM and ZEA in these feeds.
Mycotoxin management
Although binding has long been part of the discussion with addressing mycotoxin presence in feeds, the recognition of the multiple types of mycotoxins as problematic suggests that using several modes of action could be important to mitigation efforts.
Binding alone may not address all mycotoxin challenges. However, because different mycotoxins can generate similar problems for the animal - like damaging the gut wall or suppressing immunity – supporting organs damaged by mycotoxin presence is another mitigation practice. Using feed additives that provide a range of responses including adsorbing mycotoxins, supporting gut wall integrity and the immunity system, and delivering antioxidants helps limit the mycotoxins available that are harmful to an animal and can potentially mitigate some of the damages caused.
An in vitro trial done in Germany looked at the adsorption of ergot toxins using Trouw Nutrition's Toxo products. At a pH of 6.5, 90.4% of the ergot extract was absorbed and at a pH of 3.0, 93.6% of the ergot extract was absorbed.
A similar experiment explored the binding of emerging mycotoxins using the Toxo-XL product. In the trial, 1,000ppb of sterigmatocystin, 1,000ppb of enniatin B and 1,000ppb of roquefortine C were exposed to 0.2% of Toxo-XL at a pH of 3.0 or 6.5.
In more neutral conditions, 74.9%, >85.0% and 83.1% of sterigmatocystin, enniatin B and roquefortine C were adsorbed by the product, respectively, demonstrating the potential for a high adsorption of emerging mycotoxins in the GI tract.
Conclusion
The results and analyses demonstrate the importance of proper sampling as well as the use of right statistical parameter (median instead of means for concentrations) to minimise data variation between the samples and applying the LOD so that contamination percentages are not overestimated, said Swamy Haladi, Global Programme Manager for mycotoxin risk management with Trouw Nutrition.
He added: "Remember, we apply the LOD and LOQ precisely which is very critical for correct data interpretation."
Overall, Asia and MEA had the most mycotoxin prevalence followed by LATAM, North America and Europe. Unexpected results of higher prevalence of AFLA in Europe, and DON in Asia, MEA and LATAM may be related to international trading of raw materials.
"In terms of concentration, which is for me the most important one, it's DON, FUM, T-2, ZEA, AF and OTA – in that order," said Haladi. Those results are based on analysis of samples using median concentrations rather than averages.
In 2022, these data suggest a potential threat for piglets, dairy cows, sows and poultry in that order.
However, more information is needed regarding aquaculture feeds and pet food and more research is needed into emerging mycotoxins.
- Trouw Nutrition