Mycotoxins in the environment: II. Occurrence and origin in Swiss river waters.

Thirty-three different mycotoxins were surveyed over nearly two years in a typical Swiss wastewater treatment plant (WWTP), as well as in Swiss midland rivers. Out of these, 3-acetyl-deoxynivalenol, deoxynivalenol (DON), nivalenol (NIV), and beauvericin (BEA), were detected. DON was quantified in all WWTP effluent grab samples with a maximum concentration of 73.4 ng/L, while the lowest concentration was observed for BEA with 1.3 ng/L. NIV was detected in about 37%, the other three compounds in 9-36% of the weekly or fortnightly integrated flow proportional river water samples. Concentrations were river discharge dependent, with higher numbers in smaller rivers, but mostly in the very low ng/L-range, with a maximum of 24.1, and 19.0 ng/L for NIV and DON, respectively. While NIV and DON prevailed in summer and autumn, BEA occurred mostly during winter. Summer and autumn seasonal load fractions were, however, not correlating with other river basin parameters indicative of the probably most obvious seasonal input source, that is, Fusarium graminearum infected wheat crop areas. Nevertheless, together with WWTP effluents, these two sources largely explained the loads of mycotoxins quantified in river waters. The ecotoxicological relevance of mycotoxins as newly identified aquatic micropollutants has yet to be assessed.

Mycotoxins in the environment: I. Production and emission from an agricultural test field.

Mycotoxins are secondary metabolites that are naturally produced by fungi which infest and contaminate agricultural crops and commodities (e.g., small grain cereals, fruits, vegetables, and organic soil material). Although these compounds have extensively been studied in food and feed, only little is known about their environmental fate. Therefore, we investigated over nearly two years the occurrence of various mycotoxins in a field cropped with winter wheat of the variety Levis, which was artificially inoculated with Fusarium spp., as well as their emission via drainage water. Mycotoxins were regularly quantified in whole wheat plants (0.1-133 mg/kg(dry weight), for deoxynivalenol), and drainage water samples (0.8 ng/L to 1.14 µg/L, for deoxynivalenol). From the mycotoxins quantified in wheat (3-acetyl-deoxynivalenol, deoxynivalenol, fusarenone-X, nivalenol, HT-2 toxin, T-2 toxin, beauvericin, and zearalenone), only the more hydrophilic ones or those prevailing at high concentrations were detected in drainage water. Of the total amounts produced in wheat plants (min: 2.3; max: 292 g/ha/y), 0.5-354 mg/ha/y, i.e. 0.002-0.12%, were emitted via drainage water. Hence, these compounds add to the complex mixture of natural and anthropogenic micropollutants particularly in small rural water bodies, receiving mainly runoff from agricultural areas.

Experimentally determined soil organic matter-water sorption coefficients for different classes of natural toxins and comparison with estimated numbers.

Although natural toxins, such as mycotoxins or phytoestrogens are widely studied and were recently identified as micropollutants in the environment, many of their environmentally relevant physicochemical properties have not yet been determined. Here, the sorption affinity to Pahokee peat, a model sorbent for soil organic matter, was investigated for 29 mycotoxins and two phytoestrogens. Sorption coefficients (K(oc)) were determined with a dynamic HPLC-based column method using a fully aqueous mobile phase with 5 mM CaCl(2) at pH 4.5. Sorption coefficients varied from less than 10(0.7) L/kg(oc) (e.g., all type B trichothecenes) to 10(4.0) L/kg(oc) (positively charged ergot alkaloids). For the neutral compounds the experimental sorption data set was compared with predicted sorption coefficients using various models, based on molecular fragment approaches (EPISuite's KOCWIN or SPARC), poly parameter linear free energy relationship (pp-LFER) in combination with predicted descriptors, and quantum-chemical based software (COSMOtherm)). None of the available models was able to adequately predict absolute K(oc) numbers and relative differences in sorption affinity for the whole set of neutral toxins, largely because mycotoxins exhibit highly complex structures. Hence, at present, for such compounds fast and consistent experimental techniques for determining sorption coefficients, as the one used in this study, are required.

Analysis of selected phytotoxins and mycotoxins in environmental samples.

Natural toxins such as phytotoxins and mycotoxins have been studied in food and feed for decades, but little attention has yet been paid to their occurrence in the environment. Because of increasing awareness of the presence and potential relevance of micropollutants in the environment, phytotoxins and mycotoxins should be considered and investigated as part of the chemical cocktail in natural samples. Here, we compile chemical analytical methods to determine important phytotoxins (i.e. phenolic acids, quinones, benzoxazinones, terpenoids, glycoalkaloids, glucosinolates, isothiocyanates, phytosterols, flavonoids, coumestans, lignans, and chalcones) and mycotoxins (i.e. resorcyclic acid lactones, trichothecenes, fumonisins, and aflatoxins) in environmentally relevant matrices such as surface water, waste water-treatment plant influent and effluent, soil, sediment, manure, and sewage sludge. The main problems encountered in many of the reviewed methods were the frequent unavailability of suitable internal standards (especially isotope-labelled analogues) and often absent or fragmentary method optimization and validation.

Mycotoxins: diffuse and point source contributions of natural contaminants of emerging concern to streams.

To determine the prevalence of mycotoxins in streams, 116 water samples from 32 streams and three wastewater treatment plant effluents were collected in 2010 providing the broadest investigation on the spatial and temporal occurrence of mycotoxins in streams conducted in the United States to date. Out of the 33 target mycotoxins measured, nine were detected at least once during this study. The detections of mycotoxins were nearly ubiquitous during this study even though the basin size spanned four orders of magnitude. At least one mycotoxin was detected in 94% of the 116 samples collected. Deoxynivalenol was the most frequently detected mycotoxin (77%), followed by nivalenol (59%), beauvericin (43%), zearalenone (26%), ß-zearalenol (20%), 3-acetyl-deoxynivalenol (16%), α-zearalenol (10%), diacetoxyscirpenol (5%), and verrucarin A (1%). In addition, one or more of the three known estrogenic compounds (i.e. zearalenone, α-zearalenol, and ß-zearalenol) were detected in 43% of the samples, with maximum concentrations substantially higher than observed in previous research. While concentrations were generally low (i.e. < 50 ng/L) during this study, concentrations exceeding 1,000 ng/L were measured during spring snowmelt conditions in agricultural settings and in wastewater treatment plant effluent. Results of this study suggest that both diffuse (e.g. release from infected plants and manure applications from exposed livestock) and point (e.g. wastewater treatment plants and food processing plants) sources are important environmental pathways for mycotoxin transport to streams. The ecotoxicological impacts from the long-term, low-level exposures to mycotoxins alone or in combination with complex chemical mixtures are unknown.

Development, validation and application of a multi-mycotoxin method for the analysis of whole wheat plants.

Mycotoxins are known to affect the health of humans and husbandry animals. In contrast to wheat grains used for food and feed, whole wheat plants are rarely analysed for mycotoxins, although contaminated straw could additionally expose animals to these toxic compounds. Since the entire wheat plant may also act as source of mycotoxins emitted into the environment, an analytical method was developed, optimised and validated for the analysis of 28 different mycotoxins in above-ground material from whole wheat plants. The method comprises solid-liquid extraction and a clean-up step using a Varian Bond Elut Mycotoxin(®) cartridge, followed by liquid chromatography with electrospray ionisation and triple quadrupole mass spectrometry. Total method recoveries for 26 out of 28 compounds were between 69 and 122% and showed limits of detection from 1 to 26 ng/gdry weight (dw). The overall repeatability for all validated compounds was on average 7%, and their mean ion suppression 65%. Those rather high matrix effects made it necessary to use matrix-matched calibrations to quantify mycotoxins within whole wheat plants. The applicability of this method is illustrated with data from a winter wheat test field to examine the risks of environmental contamination by toxins following artificial inoculation separately with four different Fusarium species. The selected data originate from samples of a part of the field which was inoculated with Fusarium crookwellense. In the wheat samples, various trichothecenes (3-acetyl-deoxynivalenol, deoxynivalenol, diacetoxyscirpenol, fusarenone-X, nivalenol, HT-2 toxin, and T-2 toxin) as well as beauvericin and zearalenone were identified with concentrations ranging from 32 ng/gdw to 12 × 10(3) ng/gdw.

Multi-residue screening method to quantify mycotoxins in aqueous environmental samples.

Mycotoxins are naturally occurring secondary metabolites of fungi colonizing agricultural products on the field or during storage. In earlier work, we have shown that two common mycotoxins, i.e., zearalenone and deoxynivalenol, can be present at significant levels in the aquatic environment. This raises the question about the relevance of a wider range of mycotoxins in natural waters. In this investigation, we present the first validated method for analysis of some additional 30 mycotoxins in drainage, river, and wastewater treatment plant effluent water. The method includes solid-phase extraction over Oasis HLB cartridges, followed by liquid chromatography with electrospray ionization triple quadrupole mass spectrometry. Absolute method recoveries for 13 of the 33 mycotoxins were higher than 70% in wastewater treatment plant effluent (at 25 ng/L), and 27 compounds had method detection limits (MDLs) below 10 ng/L. The applicability of this method is illustrated with selected data from our ongoing monitoring campaigns. Specifically and for the first time, beauvericin and nivalenol were quantified in drainage and river water samples with mean concentrations of 6.7 and 4.3 ng/L and 6.1 and 5.9 ng/L, respectively. These compounds thus add to the complex mixture of natural and anthropogenic micropollutants in natural waters, where their ecotoxicological risk remains to be evaluated.