ARTICLE 80 FEED & ADDITIVE MAGAZINE March 2026 which mycotoxins are likely to be present and the vulnerabilities of a specific species. Analyses of broiler and laying hen diets highlighted a high feed risk potential from FUM and aflatoxin (AFLA). ZEN presented the largest risk factor for piglet diets, while FUM and AFLA were more of a challenge for sow and grow-finish pig feeds. In ruminants, ZEN and DON were a high risk for dairy calves, followed by AFLA and FUM. Dairy cows could face different challenges, with high risks from ZEN, T2/HT2, and DON in the total mixed ration, and DON and AFLA in concentrate feed. Aquafeed samples had high risk levels for AFLA, FUM, and ZEN. NEW MODEL FOR FORECASTING MYCOTOXIN PRESENCE, CONTAMINATION IN 2026 Avinash Bhat, MSc, Technical Specialist at MasterLab, noted that the dataset created by years of sample testing has allowed Trouw Nutrition to generate a new predictive model for mycotoxins. Globally, the system is designed to provide a predicted average contamination level and indicate a probable range of contamination. For example, the predicted average for DON was 460ppb, with most of the contamination between 200ppb and 750ppb. Zearalenone results indicate an average of 73ppb, with about 95% of samples expected to contain the toxin at levels ranging from 50ppb to 150ppb. The predictive system can also be used to focus on ingredients within a region. Results at those levels, however, could be influenced by limited historical data. Q1 2026 Prediction for DON & ZEA % Probability Range of Contamination (ppb) % Probability Range of Contamination (ppb) DON ZEA Month Jan 2026 464 402 - 580 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 100 100 99.99 0.01 0.0 395 - 580 402 - 580 460 464 Feb 2026 Mar 2026 Jan 2026 73.0 73.0 73.0 4.0 4.6 3.6 95.97 95.40 96.43 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 51.6 - 94.9 50.0 - 94.9 53.0 - 94.9 Feb 2026 Mar 2026 Predicted Average (ppb) 90% CI (ppb) Month Predicted Average (ppb) 90% CI (ppb) <250 <50 50 to 150 >150 to 250 >250 to 500 >500 250 to 750 >750 to 1500 >1500 to 3500 >3500 Figure 2. Like a weather forecast, the predictive model is most accurate in shorter windows, providing information on what contaminants may be lurking in feed ingredients. Avinash Bhat, MSc
RkJQdWJsaXNoZXIy MTUxNjkxNQ==