The Occurrence of Mycotoxins in Raw Materials and Fish Feeds in Europe and the Potential Effects of Deoxynivalenol (DON) on the Health and Growth of Farmed Fish Species-A Review

Abstract

The first part of this study evaluates the occurrence of mycotoxin patterns in feedstuffs and fish feeds. Results were extrapolated from a large data pool derived from wheat (n = 857), corn (n = 725), soybean meal (n = 139) and fish feed (n = 44) samples in European countries and based on sample analyses by liquid chromatography/tandem mass spectrometry (LC-MS/MS) in the period between 2012-2019. Deoxynivalenol (DON) was readily present in corn (in 47% of the samples) > wheat (41%) > soybean meal (11%), and in aquafeeds (48%). Co-occurrence of mycotoxins was frequently observed in feedstuffs and aquafeed samples. For example, in corn, multi-mycotoxin occurrence was investigated by Spearman's correlations and odd ratios, and both showed co-occurrence of DON with its acetylated forms (3-AcDON, 15-AcDON) as well as with zearalenone (ZEN). The second part of this study summarizes the existing knowledge on the effects of DON on farmed fish species and evaluates the risk of DON exposure in fish, based on data from in vivo studies. A meta-analytical approach aimed to estimate to which extent DON affects feed intake and growth performance in fish. Corn was identified as the ingredient with the highest risk of contamination with DON and its acetylated forms, which often cannot be detected by commonly used rapid detection methods in feed mills. Periodical state-of-the-art mycotoxin analyses are essential to detect the full spectrum of mycotoxins in fish feeds aimed to prevent detrimental effects on farmed fish and subsequent economic losses for fish farmers. Because levels below the stated regulatory limits can reduce feed intake and growth performance, our results show that the risk of DON contamination is underestimated in the aquaculture industry.

Keywords: deoxynivalenol (DON); fish; fish feed; growth; maize (corn); mycotoxins; soybean meal; survey; toxic effects; wheat.

Figure 1

Kernel density plot with the probability of estimated log critical concentration 5% (CC5), and boxplots of log CC5 for DON exposure in (a) fish species, n = 146, (b) rainbow trout, n = 56, (c) salmonids, n = 67 and (d) all fish species excluding rainbow trout, n = 90.

Figure 2

Effect of dietary DON concentration on feed intake (a) and growth (b) for all fish species in the dataset (n = 63) and for rainbow trout only (c,d; n = 35). Feed intake and growth values are expressed as percentage (%) of feed intake and growth seen in the control groups of the respective studies. The estimated relationships for all fish species were: (a) feed intake = 100.4 (±2.2) e^{−0.132 (±0.013) × DON}, pseudo-R² = 0.74; (b) growth = 99.0 (±2.6) e^{−0.165 (± 0.016) × DON}, pseudo-R² = 0.85. The estimated relationships for rainbow trout were: (c) feed intake= 101.1 (±2.3) e^{−0.188 (±0.016) × DON}, pseudo-R² = 0.81; (d) growth = 98.9 (±2.6) e^{−0.200 (±0.018) × DON}, pseudo-R² = 0.87. In all prediction equations above, DON concentration in the feed is expressed in mg/kg.

Figure 3

Effects of natural dietary DON (▲) and pure DON (△) on feed intake (a) and growth (b) in rainbow trout (n = 35). Feed intake and growth values were expressed as percentage (%) of the feed intake and growth of the control treatment in the respective studies. The estimated relationships were: (a ▲) feed intake = 100.9 (±1.5) e^−0.221 (±0.016) × DON pseudo-R2 = 0.91; (a △) feed intake = 100.9 (±1.08) e^−0.129 (±0.008) × DON pseudo-R² = 0.96 and (b ▲) growth = 101.1 (±1.5) e^−0.260 (±0.018) × DON pseudo-R² = 0.92; (b △) growth= 100.9 (±1.1) e^−0.157 (± 0.009) × DON pseudo-R² = 0.97. In all prediction equations above, DON is the concentration in the feed expressed in mg/kg.

Figure 4

Relationship between feed intake (% control) and growth (% control) in: (a) all fish species in the dataset (n = 63); (b) rainbow trout (n = 35) fed diets with DON.