Mycotoxin occurrence in maize produced in Northern Italy over the years 2009–2011: focus on the role of crop related factors
Published 2015-07-10
Keywords
- aflatoxins,
- fumonisins,
- logistic regression,
- modelling
How to Cite
Abstract
The occurrence of mycotoxins associated with Fusarium spp. and Aspergillus flavus in Northern Italy, and the role of cropping systems, were investigated on 140 field samples collected over the years 2009–2011. Samples were analysed for fumonisins B1 and B2 (FBs), aflatoxins B1, B2, G1 and G2 (AFs), deoxynivalenol (DON) and zearalenone (ZEN) using validated analytical methods. Information on: maize hybrid, preceding crop, tillage applied, mineral nutrition, pest and disease control, severity of European Corn Borer (ECB) attack, sowing and harvesting dates, kernel moisture at harvesting and longitude of the sampled province, were also collected. During this period there were distinct differences in FBs and AFs concentrations between years and geographic origins, and very low contamination with DON and ZEN was always found. The incidence of AFs exceeded 75% across all samples, and was almost 100% for FBs. The meteorological trends were quite different in the 3 years surveyed. 2009 was the coldest in June and the warmest in August, 2010 the most humid, and in 2011 cold weather occurred during flowering and dry conditions during ripening. The run of a logistic equation with the backward stepwise approach selected three parameters, (seeding week, ECB severity and longitude of sampling province) to predict AFB1 contamination and four parameters (year, sowing week, ECB severity and longitude of sampling province) to predict FB contamination. The internal validation gave good results, with 76% correct predictions. The probability of harvesting maize with more than 5 µg kg-1 of AFB1 varied between 86 and 5%, and the probability of harvesting maize with more than 4,000 µg kg-1 of FBs varied between 81 and 2%, respectively, for conducive and non-conducive environments. Therefore, considerable variability was found even if a limited area and only 3 years were considered.