Infected grasses as inoculum for Fusarium infestation and mycotoxin accumulation in wheat with and without irrigation

Grasses growing next to agricultural fields influence the Fusarium abundance, the species composition, and the mycotoxin accumulation of wheat plants, especially the field parts directly adjacent to grasses, are highly affected. Grasses are a more attractive and suitable habitat for Fusarium fungi compared to other arable weeds and occur at mostly every semi-natural landscape element (e.g., kettle holes, hedgerows, field-to-field-borders). In our study, we analyzed the ability of a highly Fusarium infected grass stripe (F. graminearum, F. culmorum, F. sporotrichioides) to infect an adjacent wheat field with these species. Results show that the primary inoculated Fusarium species were as well the dominant species isolated from the wheat field. Regarding transects originating from the grass stripe going into the field, the results demonstrate that wheat ears next to the infected grass stripe have a higher Fusarium abundance and furthermore show higher mycotoxin accumulation in the wheat kernels. This effect was highly promoted by irrigation. Therefore, grass stripes next to arable fields must be considered as reservoirs for fungal infections and as a source for a contamination with mycotoxins. Supplementary Information The online version contains supplementary material available at 10.1007/s12550-022-00470-2.


Introduction
Wheat (Triticum aestivum L.) is one of the most economically important crop on which global food security is depending. Unfavorably, numerous plant diseases can reduce yields and thereby threaten the food security. Wheat is highly susceptible to fungal infections and fungal associated diseases can be responsible for 15-20% of crop losses worldwide each year (Figueroa et al. 2018;Savary et al. 2019), whereas viral and bacterial diseases are usually less important (Oerke 2006).

3
Regarding fungal infections like wilts, rots, and blights (Rampersad 2020), mainly Fusarium head blight (FHB), the economically most devastating wheat disease, caused by state of the art up to 19 different Fusarium species, poses farmers to challenges (Parry et al. 1993;Champeil et al. 2004;Landschoot et al. 2011;Müller et al. 2012). Growth of fungal pathogens in wheat as well as the mycotoxin is highly influenced by agricultural, topographic, and also climatic factors (Dill-Macky and Jones 2000;Vogelgsang et al. 2019). A strong multifactorial influence of tillage, previous crop, variety, soil and air moisture, and other site factors on fungal pathogen community and on mycotoxin accumulation in wheat is observed. Different cropping systems and the interaction between different phytopathogenic fungi and bacteria also determine the incidence and impact of fungal plant diseases (Classen et al. 2005;Drakopoulos et al. 2021). Mycotoxins produced by Fusarium fungi during its growth in the wheat ears could have different toxic effects on both animal and human health after uptake of contaminated wheat products (Zain 2011;Chhaya et al. 2021). Deoxynivalenol (DON), zearalenone (ZEN), and nivalenol (NIV) are the most frequently reported mycotoxins and mostly detected on wheat, maize, rice, barley, and rye (Visconti et al. 1992;Bottalico 1998). In Europe, DON and ZEN, mainly produced by F. graminearum and F. culmorum are the most detected toxic secondary metabolisms. F. graminearum is more common in warmer, while F. culmorum is more often to be found in colder regions of Europe (Pasquali et al. 2016).
However, crop residues enduring in the field from last season are regarded as the primary inoculum for an infection with FHB (Landschoot et al. 2011;Leplat et al. 2013). In recent years, also (arable), weeds are of interest as reservoir for fungal spores and as a source infection for crop plants. Weeds are part of every agricultural system and occur alongside the crops in the field season, more since a lot are resistant against herbicides. Seventy percent of gramineous weed species in China were already resistant to herbicides, making them hard to control (Dong et al. 2020). At seminatural habitats (e.g., hedgerows in the field, field margins, kettle holes), weeds can even grow permanently over several field seasons because they were not harvested alongside the crops, giving species of Fusarium a suitable habitat to overwinter or outlast times when their main host is not available. Originating from these weeds, the fungal spores can immigrate into the field by wind, rain and through transport by organisms such as insects (Rossi et al. 2002;Paul et al. 2004;Heitmann et al. 2021;Hoffmann et al. 2021). However, weeds offer broad host range for fungal pathogens, including Fusarium, and can act as both alternative and alternate host for overwintering (Kumar et al. 2021), development and sporulation. Especially gramineous weeds harbor high abundances of Fusarium fungi while are also infested with a diverse species composition (Lofgren et al. 2018;Fulcher et al. 2019a, b B;Dong et al. 2020;Gerling et al. 2022). Wild grasses were also determined to be important in the survival of F. graminearum, F. sporotrichioides, F. equiseti, F. avenaceum, F. poae, F. oxysporum, and F. culmorum (Landschoot et al. 2011). Senescent but also fresh weeds were analyzed in Croatia, isolating different Fusarium species with F. graminearum as the most frequent one (Postic et al. 2012). Also, Mourelos et al. (2014) isolated F. graminearum from a great variation of weed species, and Suproniene et al. (2019) detected that the Fusarium strains isolated from weeds are also pathogenic for wheat. This leads to the assumption that also weeds, besides crop residues and soil, are an important habitat for fungal pathogens to survive and sporulate. The community composition of Fusarium on arable weeds could be depending on many factors including the study side, crops, the weeds, and the climate condition. Regional differences in species profile may exist due to different crop rotation systems and local climatic conditions (especially temperature and precipitation) (Liu and Van Der Fels-Klerx 2021).
Organic farming is becoming increasingly important in European agriculture (Hansen et al. 2001). As a result, the control of weeds with herbicides is severely limited and weeds are becoming more common in crop fields. Also, the establishment of flower stripes or semi-natural landscape structures such as hedges and wide edge grass stripes to preserve biodiversity (pollinators, amphibians, ground beetles, birds) (Benton et al. 2003;Lozada-Gobilard et al. 2021) cause weeds and grasses to be more and more common in agricultural landscapes and thus -besides all positive effects -also a habitat for phytopathogenic mycotoxin producing fungi. The impact of these measures on the spread of plant diseases in adjacent cereal fields is still unknown.
The life cycle of filamentous fungi like Fusarium spp. is highly influenced by moisture conditions in their living environment. Various studies mentioned environmental conditions, including relative humidity, as essential for the development of Fusarium (Backhouse et al. 2004;Osborne and Stein 2007;Landschoot et al. 2011;Kelly et al. 2015). Parikka et al. (2012) expected a shift in the species composition of Fusarium on cereal grain in northern Europe due to changing climate conditions. Liu and Van der Fels-Klerx (2021) mentioned that the increase of precipitation but also increasing temperatures in Europe may promote the spread of Fusarium spp. The changing weather conditions caused by climate change do not only influence the abundance of Fusarium, but also the presence of mycotoxins in wheat crops among others, due to their existence depending on multiple factors including temperature and precipitation. However, the contribution of weeds in the Fusarium head blight disease cycle is not fully understood by now (Keller 1 3 et al. 2014;Miedaner et al. 2017;Fulcher et al. 2019a, b;Martínez et al. 2021).
Therefore, we hypothesize that (1) a highly Fusarium infected grass stripe acts as an infection source of different Fusarium species for the adjacent wheat field, with this effect decreasing with increasing distance from the grass stripe; (2) the mycotoxins produced by the Fusarium are accumulated in high concentrations in this transition zone from the grass stripe into the field and thus differ from the average toxin concentration inside the field; (3) these transmission effects are enhanced by irrigation on one half of the field and lead to higher Fusarium infestation and mycotoxin concentrations of wheat plants in this field side.

Material and methods
In 2020, a field experiment was conducted in a winter wheat field on the research area at the Leibniz Centre of Agricultural Landscape Research (ZALF) Müncheberg to evaluate the influence of highly infected grasses and irrigation on the spread and the diversity of Fusarium fungi.

Study site
The examined winter wheat field (preceding crop: winter wheat), located in Müncheberg, Germany (52.5176 N;14.1217 E), was 3339 m 2 in total (53 m wide × 63 m length), divided into two field sides. The soil type was Sl4D soil (sandy clay loam pallid soil). The wheat variety "Tobak" was cultivated, which is susceptible for Fusarium head blight (susceptibility of the wheat cultivar to an infection with Fusarium spp. expressed as a rating of 7 on a scale from 1 to 9; Anonymous, 2022). Detailed information on the agricultural management practices and applications are shown in Table S1. All the residues of the previous crop were removed from the field; the soil was ploughed and then cleared again. Emerging weeds in the field were suppressed or minimized at all times with herbicides or by hand. No fungicides against Fusarium infections were applied. Additionally, herbicides decimate the weeds emerging in winter. In spring, the field was checked again, and the few remaining weeds were eliminated. The management of the trial aimed very consistently at minimizing all sources of infection except the inoculated grass stripe.

Inoculation of the grass stripe
The grass stripe was infected with Fusarium fungi through soil inoculation. Fusarium isolates used as inoculum in this study originated from wheat plants and maintain in the culture collection of fungi of the working group "Fungal Interactions" at the ZALF in Müncheberg. To perform the inoculation, wheat kernels were infected with three different Fusarium species: F. graminearum, F. culmorum, and F. sporotrichioides. Wheat kernels were moistened with sterile water and autoclaved three times, before infected with Fusarium fungi by mycelium-covered nutrient media (potato dextrose agar PDA, Merck, Heidelberg, Germany) and incubated by 24 °C for 14 days in the dark. In total, 3600-g wheat kernels were used, 1200 g per Fusarium species. Inoculum for every Fusarium species was produced separately. Afterwards, the infected kernels from each species were uniformly spread over the soil (7.5-g inoculum from each species/m 2 ) in November 24, 2020, and lightly pressed into the ground. F. graminearum was chosen for being the main causal agent of FHB in wheat and other small-grain cereals in Europe. F. culmorum is a soil-borne pathogen and also highly associated with FHB. F. sporotrichioides is less pathogenic to wheat, though was also selected due to its high abundance on grasses at small water bodies on wheat fields in Brandenburg/Germany (Gerling et al. 2022).

Microclimate data
Soil humidity (m 3 /m 3 ), air humidity (%), leaf wetness (%), and air temperature (°C) were monitored from March 2021 through the duration of the experiment using 6 microclimatic measuring stations (Hobo H21 Micro Station, Onset Computer Corporation, MA, USA) with soil moisture smart sensor (S-SMD-M005), leaf wetness smart sensor (S-LWA-M003), and temperature/RH smart sensor (S-THB-M002) all of them provided by Onset Computer Corporation (Bourne, MA, USA). Except for the soil humidity sensor (measurement in 10 cm depth), all sensors were positioned at a height of 30 cm above ground. The sensors provided measurements every hour.

Sampling design
Samples of the grasses from the inoculated grass stripe were taken in May and June 2021 at 6 different sampling points. Fifteen grass blades were cut 2 cm above the ground. The abundance of Fusarium and furthermore the species composition were analyzed by culture-dependent methods to monitor the inoculation success and the development of the inoculated species. Grass samples in June 2021 were taken while wheat was flowering and was meanwhile most susceptible to fungal infections (Góral et al. 2019).
In July 2021, 2 weeks after full flowering at BBCH stage of medium to late milk development (BBCH 75-77) (Zadoks et al. 1974), wheat ears were sampled a long six transects starting next to the grass stripe up to 64 m into the wheat field. Sampling points were set at 1 m, 2 m, 4 m, 8 m, 16 m, 33 m, and 64 m. The transects were 5 m apart from each other. Fifteen different wheat ears were randomly picked in a 0.5 m area around the sampling point and cut 2 cm below the ear. Wheat ears were sampled for further investigations regarding the abundance of filamentous fungi and the diversity of Fusarium species. All samples were collected in paper bags and immediately transported to the laboratory for further investigations. Collected plants were analyzed by culture-dependent and culture-independent methods for the presence of fungi, mainly filamentous fungi of the genera Fusarium.

Culture-dependent method
Potato dextrose agar (PDA; Merck, Heidelberg, Germany) supplemented with chloramphenicol to suppress bacterial growth, and synthetic nutrient agar (SNA) (Nierenberg 1976) were used as described detailed by Leslie and Summerell (2006) to determine the colony forming units (cfu) per gram of fresh matter of Fusarium (FUS_cfu/gFM) and the species analysis of Fusarium. Regarding the grasses, ten pieces (about 1 cm length) of each grass sample were cut out and weighted (Kern 572-35; Kern&Sohn GmbH Balingen-Frommern, Germany) to calculate the colony forming units to 1 g of plant fresh matter for further statistical analyses. Afterwards, grass pieces were plated onto two PDA containing petri dishes (diameter 9 cm); five pieces (mixed from leaves, stems, and inflorescences) were placed on each plate.
Wheat samples were prepared as described above by using 10 kernels per sampling point from 10 different wheat ears (one kernel per ear), randomly chosen from the bottom, the middle and the tip of the ear.
Plated samples, both grass samples and wheat samples, were incubated for 2 days at 24 °C in darkness to support the fungal growth and further 2 days under UV light (12 h UV light/12 h daylight) at room temperature to support sporulation and coloration of the mycelium. After 4 days, the colonies of Fusarium (FUS_cfu/gFM) were counted on each petri dish.
For the morphological identification of different Fusarium species, again PDA and SNA petri dishes were used. Colonies from the initial plate were transferred onto one plate of both PDA and SNA. New petri dishes were treated as described above, but with of a longer UV light treatment depending on the growth rate of the different Fusarium species. UV light treatment up to 14 day for slow growing species is possible. The cultures grown on PDA supports the formation of the typical coloring of the fungal mycelium, while SNA supports each Fusarium species in developing its species-specific macro-and microconidia. Fusarium isolates were identified to species level using light microscopy (Jenaval, Carl Zeiss, Jena, Germany), and the identification was mainly based on morphological characteristics described in details by Leslie and Summerell (2006).

Culture-independent method (qPCR approach)
For further analyses by qPCR, remaining grass and wheat ear samples were dried (60 °C for at least 48 h) and ground using a vibrating cup mill RS200 (Retsch, Haan, Germany): 1300 rpm for 1.5 min for grasses and 1000 rpm for 45 s for wheat ears. Afterwards, the grounded material (250 mg for grasses and 50 mg for wheat ears) was mixed and used for genomic DNA extraction according to the customized standard protocols of the following DNA extraction kits: Nucle-oSpin® Soil Kit (MACHEREY-NAGEL GmbH & Co. KG, Düren, Germany) for grasses and DNeasy Plant Mini kit (QIAGEN GmbH, Hilden, Germany) for wheat ears.
The lyses for the grass samples were adapted to plant material, and samples were centrifuged by 13,000 rpm instead of 11,000 rpm. The quantification of DNA gene copy numbers of Fusarium by a qPCR approach with genusspecific primers was described in detail by (Müller et al. 2018). The QuantStudio™ 12 K Flex Real-Time PCR System (Thermo Fisher Scientific Inc., Waltham, MA, USA) and the software "QuantStudio™ 12 K Flex Software v1.x" (Thermo Fisher Scientific Inc., Waltham, MA, USA) was used for the qPCR assay. Quantification of Fusarium gcn/ gDM was performed using the HOT FIREPol® Probe GC qPCR Mix (Solis Biodyne, Tartu, Estonia). The standard curves were generated by using DNA of F. graminearum strain name "Fg486". Reactions were carried out under the following thermal conditions: 95 °C for 10 min (hold stage) and 45 cycles of 95 °C for 15 s and 67 °C for 1.5 s (PCR stage). The fungal strains used for the preparation of the standard curves were stored in a culture collection of fungi of the working group "Fungal Interactions" at the ZALF in Müncheberg. All measurements were done in duplicate and qPCR assays contained negative controls. The genome copy numbers were expressed in FUS_gcn/g DM for Fusarium fungi.

Analyses of deoxynivalenol and zearalenone
For the analyses of DON and ZEN, harvested and threshed samples were dried at 60 °C for at least 48 h and ground in an ultra-centrifugal mill with vibratory feeder (ZM 200, Retsch Haan Germany). The Fusarium toxins DON and ZEN were extracted as described in detail previously by Müller et al. (2018) and analyzed on a HPLC system consisted of an Ultimate 3000 unit (Thermo Fisher Scientific GmbH, Dreieich, Germany) combined with a diode array detector (DAD3000) and a fluorescence detector (FLD 3400RS). DON was separated on a Synergie Polar RP 100A column (2.5-μm particle size, 130 × 3 mm i.d., Phenomenex LTD, Aschaffenburg, Germany). Methanol:water (25:75, v/v) was used as eluent with a flow rate of 0.65 ml/min at 30 °C. The analyte was detected by measuring the UV-absorbance at 220 nm and 280 nm simultaneously.
ZEN was analyzed after separation on a Lichrospher 100 RP18 column (5-μm particle size, 150 × 4.6 mm i.d., VDS OptiLab Berlin, Germany) with methanol:3 mM phosphoric acid (65:35, v/v) as eluent with a flow rate of 0.65 ml/min at 25 °C. ZEN concentration was determined by fluorescence detection (extinction: 274 nm, emission: 456 nm). Mycotoxin standard substances were obtained from Romer Labs Diagnostic GmbH (Tulln, Austria). Each analysis was performed in duplicate. All toxin concentrations were calculated on the dry matter (DM) of the substrate (ng/g DM), and the toxin detection limits in the grains were 30 ng DON and 2 ng ZEN per gram of substrate DM.

Statistics
All statistical tests were performed using SPSS (IBM SPSS Statistics V 22.0). For the visualization of the gcn/gDM of fungal abundances as boxplots, a logarithmic transformation LOG (x + 1) was applied to the data of the culture-independent method (qPCR approach). Via a Kolmogorov-Smirnov test, the abundance data of Fusarium (cfu/gFM and gcn/gDM) was tested for normal distribution. The differences in fungal quantities between the different Fusarium species (gcn/gDM), and the differences between the mycotoxin accumulations (ng/g) were compared by Kruskal-Wallis test and Mann-Whitney U test. In all figures, a/b indicate significant differences between parameters investigated (p values < 0.05).

Microclimate
Except for air temperature, we were able to detected differences regarding the microclimatic factors on the irrigated and the non-irrigated field side. Especially in June, but also in July, we found a higher soil moisture, a higher air humidity and a higher leaf wetness on the field side with irrigation ( Fig. S1).

Inoculated grass stripe with F. graminearum, F. sporotrichioides, and F. culmorum in May and June
The Fusarium abundance (FUS_cfu/gFM) determined in the two sampling months decreased from May to June (Fig. 1): the mean total abundance of Fusarium on the grass stripe in May was 598 FUS_cfu/gFM, compared to 434 FUS_cfu/ gFM in June. The abundance of Fusarium was 1.4-fold lower in the second month. Regarding the irrigation, the measured abundances of the total Fusarium abundance in May were almost even. In June, the abundance of Fusarium on the irrigated field side was 1.3-fold higher compared to the field side with no irrigation. In both months, F. sporotrichioides was the predominant species on the grass samples investigated. The abundance of F. sporotrichioides (SPORO_cfu/ gDM) was higher in May and decreased until June, while there was no difference associated to the irrigation detected: the abundance of F. sporotrichioides was almost even on both sides. Also, the abundance of F. culmorum decreased from May to June. In May, the abundance F. culmorum on the irrigated field side was even to the non-irrigated field side, and in June, the abundance on the irrigated side was 3.7-fold higher.
In contrast to the other Fusarium species analyzed, the abundance of F. graminearum (GRAM_cfu/gFM) increased from May until June. We determined 16 × more GRAM_cfu/ gFM in June compared to May. In May, F. graminearum was only isolated in low abundances on the irrigated field side. Also in June, more F. graminearum (1.4-fold) was detected on the field side with irrigation, but further on the nonirrigated side as well. In May, F. culmorum was the second most isolated Fusarium species of the grass samples, while in June, it was F. graminearum. The abundance of F. sporotrichioides in May was 39-fold higher than the abundance of F. graminearum, while in June, the difference was only 26%.
We also isolated F. equiseti from the grass stripe, even though this species, compared to the others named above, was not inoculated. We found low abundances of F. equiseti (11 EQUI_cfu/gFM) both in May as well as in June, independently from the irrigation.

Gene copy numbers of Fusarium spp. on wheat ears on the irrigated and the non-irrigated field side and along the transect
Regarding Fusarium, the abundance of the FUS_gcn/gDM on the irrigated field side was significantly higher compared to the non-irrigated side ( Fig. 2A).
However, concerning the transect, high abundances up to 26,000,000 FUS_gcn/gDM were found at the first sampling point at 1 m next to the infected grass stripe (Fig. 2B). In 4 m distance, there is still a FUS_gcn/gDM of 6,000,000 to 1 3 be found. These high Fusarium abundances only occurred on the irrigated field side. On the non-irrigated field side, the abundance of Fusarium differs from 1,250,000 to 18,300 gcn/gDM. The highest abundances on the non-irrigated field side were measured at 8 m and 64 m, the lowest at 33 m, but the differences in the abundance were not statistically significant.
Concerning Fusarium, there is a gradient from the grass stripe along the transect to be found on the irrigated field side. Significant differences were detected starting at 16 m away from the infected grass stripe, compared to the sampling point directly next to the stripe. From 1 to 4 m the FUS_gcn/ gDM decreased, while at 8 m the abundance increased again up to the level of 2 m. In contrast, no significance along the transect could be identified on the non-irrigated field side.

Diversity of different Fusarium species determined on wheat kernels in July 2021 on the irrigated and the non-irrigated field side and along the transect
Concerning the different Fusarium species isolated from the wheat field in July, F. graminearum was the dominant species on the irrigated field side ( Table 1). The latter accounted for 77% of the whole Fusarium fungi isolated from the wheat kernels. This abundance is significantly different to the abundances of F. sporotrichioides (7%) and F. culmorum (1%), among the other detected species. Besides the three inoculated Fusarium species, we identified four additional Fusarium species on the field with irrigation: F. equiseti, F. cerealis, F. avenaceum, and F. poae. F. equiseti accounted for 5% of the total Fusarium abundance and was already isolated from the grass stripe in May and June. F. cerealis and F. avenaceum were detected for the first time during the field study and were both found in abundances of 3%. F. poae, also initially found in July on wheat ears, accounted for 1% of the total Fusarium abundance. In total, seven different Fusarium species in different abundances were determined on the irrigated field side from wheat kernels in July, during the ripening of ears.
On the non-irrigated field side, the species composition of Fusarium was more balanced. F. culmorum, the least isolated species on the irrigated field side, was the dominant species on the non-irrigated field side and accounted for 36% of the total Fusarium abundance. F. graminearum (20%) and F. sporotrichioides (17%), both inoculated species together with F. culmorum, were also frequently detected on the non-irrigated field side. Again, F. equiseti, which was not inoculated at the beginning of the field experiment, was found in the same high amount as F. sporotrichioides. This makes F. equiseti, together with F. sporotrichioides, the third most isolated Fusarium species on wheat ears on the field side without irrigation. Furthermore, just like on the irrigated side, F. cerealis (8%), even in higher abundances, and F. avenaceum (2%) were isolated. We detected three additional Fusarium species on the non-irrigated field side. F. poae was only isolated from the irrigated field side.
On the irrigated field side, F. graminearum, predominant species on this field side, was detected in similar high abundances from 1 m up to 8 m along the transect (Fig. 3A). Between 8 and 33 m away from the inoculated grass stripe, there was a decrease in the abundance of F. graminearum to be found, just as between 16 and 33 m. But only between ears at 8 m and 33 m, the difference was significant. No isolates of F. graminearum were determined from 64 m away from the grass stripe, also none of F. sporotrichioides or F. culmorum.
At 1 m, directly next to the grass stripe, F. sporotrichioides was the second most detected species, but with abundances 3.4-fold lower than the abundance of F. graminearum on wheat ears. At 2 m, 8 m, and 11 m, we also isolated F. sporotrichioides, but only in low abundances (0.61-0.71 SPORO_cfu/gFM). F. culmorum was only determined at two sampling points: 1 m and 33 m. Compared to F. graminearum, the abundance of F. culmorum at 1 m was 21.4-fold lower, 6.3-fold lower compared to F. sporotrichioides. We also isolated four additionally Fusarium species, which were not inoculated at the grass stripe in November 2020. These species were detected at all sampling points from 1 m up to 8 m and again at 64 m. At 1 m, the abundance of not inoculated Fusarium species was threefold higher compared to the abundance of F. culmorum. Regarding the irrigated field side, none of the primary inoculated Fusarium species were detected at the last sampling point, 64 m away from the infected grass stripe.
On the non-irrigated field side, the abundances of all three inoculated Fusarium species are decreasing from 1 m up to 4 m along the transects (Fig. 3B). F. sporotrichioides was the predominant detected species isolated from wheat ears at 1 m, next to the artificial infected grass stripe, followed by F. culmorum. The abundance of F. graminearum was 2.4-fold lower compared to F. sporotrichioides. At 2 m and 4 m, the abundances of the three inoculated species were even. F. sporotrichioides was isolated up to 8 m into the field without irrigation. F. culmorum was determined from 1 to 4 m in decreasing abundances. Concerning additionally detected Fusarium species (Fig. 3A/B), we isolated those at all sampling points except 4 m and 33 m. At 2 m, the other Fusarium species were dominant, and also at 8 m, 6 m, and 64 m, additionally occurring Fusarium species were found in higher abundances than some of the inoculated species: at 8 m, the abundance of F. sporotrichioides was twofold lower compared to the other species, at 16 m, we determined the same for F. culmorum. Additionally occurring species were the second most detected Fusarium species at 64 m, but found in 3.4-fold lower abundances than F. culmorum. All species were isolated in low abundances: no statistically significant differences between the cfu/gDM of the different species were detected.

Analyses of DON and ZEN concentration along the transects on the irrigated and the non-irrigated field side
The highest levels of DON in wheat kernels were detected on the irrigated field side. Up to 30,000 ng/g DON were analyzed via HPLC at the first sampling point directly next to the infected grass stripe (Fig. 3C). Along the transect into the wheat field, the DON concentration decreased until 110 ng/g at 64 m. There were significant differences in the mycotoxin concentration in wheat kernels on the irrigated field side at 16 m, 33 m, and 64 m, compared to the sampling point next to the infected grass stripe (1 m). The highest decrease in DON concentration was measured between 8 and 16 m: 72.5% less DON at 16 m. At the last two sampling points (33 m and 64 m), only a low concentration of DON of 718 ng/g and 110 ng/g were detected via HPLC, especially compared to the concentration at sampling point 1 m. Concerning the non-irrigated field side, the DON concentration along the transect is lower compared to the irrigated side. A trending decrease of the DON concentrations from 790 to 124 ng/g was measured, but none of the differences between the sampling points along the transect were statistically significant.
Regarding ZEN, a mycotoxin contamination up to 14 ng/g at 1 m, next to the infected grass stripe, was measured. At the last sampling point at 64 m, no ZEN accumulation in a recognizable amount was detected (Fig. 3D). From 1 to 4 m the ZEN concentration decreased, while it is even at 4 m and 8 m. After 8 m, it decreased again until the end of the transect. The measured differences were significant from 1 to 16 m and also 33 m. The ZEN accumulation on the irrigated field side never exceeded the toxin regulations set by the EU, and no ZEN contamination was found on the non-irrigated field side.  Fig. 3 A Mean values of the cfu/gFM of F. graminearum (GRAM_ cfu/gFM), F. sporotrichioides (SPORO_cfu/gFM), F. culmorum (CULM_cfu/gFM) and other Fusarium species (OTHER_cfu/gFM) along a transect from 1 to 64 m into a wheat field on the irrigated field side. a/b indicate significant differences (*p < 0.05), B Mean values of the cfu/gFM of F. graminearum (GRAM_cfu/gFM), F. sporotrichioides (SPORO_cfu/gFM), F. culmorum (CULM_cfu/gFM) and other Fusarium species (OTHER_cfu/gFM) along a transect from 1 to 64 m into a wheat field on the non-irrigated field side. C Mean values of the DON concentration in ng/g on wheat ears from the irri-gated and the non-irrigated field side along a transect from 1 to 64 m into a wheat field. The mean values are plotted with their standard deviation of the mean. a/b indicate significant differences (*p < 0.05), D Mean values of the ZEN concentration in ng/g on wheat ears from the irrigated field side along a transect from 1 to 64 m into a wheat field. The mean values are plotted with their standard deviation of the mean. Asterisk brackets indicate significant differences (*p < 0.05) ZEN was under the toxin detection limit of 2 ng/g at sampling point 64 m on the irrigated field side and on all sampling points on the nonirrigated field side Regarding the yield, no statistically significant differences were detected: the yield was almost even at the irrigated and the non-irrigated field side.

Discussion
In the present study, an inoculated grass stripe was analyzed as reservoir for an infection with Fusarium spp. for an adjacent wheat field. The grass stripe was highly infected with three different species of the genera Fusarium: F. graminearum, F. culmorum, and F. sporotrichioides. Furthermore, the distribution of Fusarium along a transect into to field, the Fusarium species composition on the wheat ears and the mycotoxin contamination of the wheat kernels were determined. We could clearly show that infection has spread from the grass stripe up to 33 m into the wheat field, as evidenced by the increased abundance of Fusarium fungi as well as mycotoxin accumulation in the wheat ears near the grass stripe. This effect was remarkably enhanced by altered microclimatic conditions (humidity and temperature) that prevailed on one side of the field as a result of the irrigation.
Gerling et al. (2022) measured the Fusarium abundance of different weed species at the edges of semi-natural landscape elements in arable lands, finding out that grasses are a more attractive habitat for different Fusarium species than herbaceous plants. Both, the fungal abundance and the species diversity was highest on grass samples. In the field experiment shown here, we were able to create a strong infection of grass stripe due to soil inoculation up to 600 FUS_cfu/gDM. This seems to confirm the findings of the just cited study: grasses are suitable hosts for different species of Fusarium fungi. It is assumed that different Fusarium species and their community composition on both grasses and wheat are strongly determined by abiotic factors such as soil and air humidity and temperature as well as by their own competitive strength, aggressiveness and spore dispersal. It was general consent that F. graminearum was limited to regions with warm temperature (28 °C) and with moderate to high rainfall during anthesis (Parry et al. 1993;Backhouse and Burgess 2002;Xu and Nicholson 2009) and therefore was the most common causal agent of FHB in the USA, in Canada, Australia, and parts of central Europe (Goswami and Kistler 2004;Osborne and Stein 2007). In cooler, maritime regions like the UK and northern Europe, F. culmorum was prevalent (Parry et al. 1993;Jennings et al. 2004;Gilbert et al. 2013). Due to rising temperatures, F. graminearum incidences increased also in regions where Fusarium culmorum was the main causal agent of FHB (Waalwijk et al. 2003;Xu et al. 2005). Different studies confirm the increase of F. graminearum and the decrease of F. culmorum over the last two decades in the UK (Jennings et al. 2004), in Netherlands (Waalwijk et al. 2003), in Denmark (Nielsen et al. 2011), in Poland (Stępień andChełkowski 2010) and in Germany (Miedaner et al. 2008), which may reflect the climatic effects on these species. The results of the present study regarding the abundance of F. graminearum and F. culmorum in the grass stripe showed a similar pattern but in the small-scale dimensions of our field experiment. Both species were inoculated in the grass stripe. In May, the month with lower temperature, higher abundances of F. culmorum were isolated from the grasses, while no isolates of F. graminearum were found. In June, the temperature raised and also the F. graminearum incidence increased, while the incidence of F. culmorum decreased. F. sporotrichioides also used as an inoculum, was the predominant isolated species in both months on the grass stripe, followed by F. culmorum in May and F. graminearum in June. This leads to the assumption that Fusarium species which are better adapted to rising temperatures may become more prevalent in the future and that the high infestation of F. graminearum indicates a better adaption to changing environmental conditions than F. culmorum (Panwar et al. 2016). However, this assumption seems to be confirmed for grasses as non-crop host plants colonizing as alternate and alternative hosts. F. sporotrichioides is a causal agent of FHB, but is not as often isolated as F. graminearum or F. culmorum (Salas et al. 1999;Lenc et al. 2015), but on grasses, it seems to be competitive, even against F. graminearum. F. sporotrichioides is mainly to be found in temperate regions, but has a temperature range from approximately − 2 to 35 °C (Thrane 2014). The infection of inflorescences of host plants through this species occurs from 10 to 40 °C (Nazari et al. 2014). Gerling et al. (2022 analyzed the abundance of different Fusarium species on arable weeds, including grasses, at the edges of kettle holes in two consecutive autumn/winter seasons with the result that F. sporotrichioides was the dominant isolated species in both seasons. In the present study, F. sporotrichioides was detected on grasses and wheat ears, but in lower abundances than F. graminearum (irrigated field side) and F. culmorum (non-irrigated field side). The competitiveness of F. sporotrichioides against other Fusarium species seems to be better on grass-hosts, compared to wheat plants.
Fusarium fungi colonized the wheat plants in high frequencies and abundances and produced remarkable mycotoxin amounts in the field, especially at the sampling points near the infected grass stripe. This result clearly shows that the infected grass stripe influenced the fungal abundance of Fusarium in the wheat field and the mycotoxin accumulation in the wheat kernels at harvest time. The Fusarium species composition on the wheat ears was strongly influenced by the grass stripe growing next to the field and the Fusarium spores transported by wind and plant-to-plant contact from there. The previously inoculated Fusarium species account for 85% (irrigated field side) and 73% (non-irrigated field side) of the total Fusarium abundance, but in different 1 3 abundances for the species. This result clearly shows the immigration of the inoculated Fusarium species from the grass stripe into the field, which makes a Fusarium infected grass stripe next to a wheat field a reservoir for fungal infections. This pattern was determined on both field sides but seems to be strongly promoted by the irrigation. The species composition of Fusarium depends on several factors, especially on the precipitation and the temperature at flowering stage (Bottalico and Perrone 2002;Covarelli et al. 2015). Therefore, due to climate change, shifts in the composition of Fusarium species on cereal grains in Northern Europe are expected (Parikka et al. 2012). Regarding further frequently detected Fusarium species in our study, F. equiseti is mainly known as a saprophytic species, which can also be a causal agent for FHB, but not as aggressive as F. graminearum or F. culmorum (Langseth et al. 1999). Desai et al. (2020) suggested that high intense rainfall events could promote the saprophytic living Fusarium species. In the present study, this species was more frequent detected on wheat kernels of the non-irrigated field side. F. equiseti was also isolated from the irrigated field side, but in lower abundances. Maybe the saprophytic F. equiseti was not able to compete with the more aggressive F. graminearum species, which was the dominant species on the irrigated field side.
A comparison with other studies that have investigated the extent to which the effects of edge structures or transition zones affect the neighboring agricultural field is difficult because there are only a few of them.
A study of Raatz et al. (2019) shows that yield decreases next to semi-natural landscape elements up to a distance of 11.3 m from kettle holes and up to a distance of 17.8 m from hedgerows and forest borders. The authors named shade, soil compaction through the use of machinery and also a reduced use of chemical preparations close to water bodies as the main reasons. In a second study, Raatz et al. (2021) analyzed different pests and pathogens regarding their ability to decrease yield from semi-natural landscape elements into wheat fields, but did not find any evidence of yield reduction due to an infection with pathogenic fungi, although the infection rate of wheat leaves next to kettle holes was higher compared to the mid field. This is in agreement with our results: The present study shows no significant differences in yield between the irrigated and non-irrigated field sides, although the former was much more infested with filamentous fungi of the genera Fusarium. Raatz et al. (2021), however, only analyzed the yield quantity, no qualitative parameters. But an infection of wheat with, e.g., FHB can cause both, quantitative yield losses and qualitative changes of nutrients in wheat grains including an accumulation of toxic substances (Martínez et al. 2020): Quality losses in wheat are primarily caused by the contamination of the harvested crop with mycotoxins. If the contamination exceeds the limits set by the EU, the crop may no longer could be sold, fed, or processed and must be destroyed. The limits for DON and ZEN in unprocessed cereals are set by the European Commission regulation (EU) at 1250 ng/g for DON and 100 ng/g for ZEN .
In our experiment, we investigated both, the yield, and the mycotoxin contamination of grains at the harvest. The results show a significantly increased toxin accumulation on the field side which was irrigated. This was determined for both the analyzed mycotoxins, but mainly for DON. The limits for DON were exceeded with a maximum value of 29.000 ng/g on the irrigated field side at the sampling point directly next to the grass stripe. We determined high DON contaminations of the grains up to 16 m into the field. Only at a distance of 33 m from the grass stripe did the DON concentration decline to values below the EU limit (718 ng/g) and at the end of the transect (64 m away from the grass stripe), we analyzed only a low concentration of DON (110 ng/g). On the dryer field side, the limits for DON were not exceeded, same for the limits of ZEN on both field sides, independent from the irrigation. But also, for DON concentration of the non-irrigated field side and for ZEN, there was a decrease to be detected: the highest levels of mycotoxin contamination were analyzed at the sampling points right next to the infected grass stripe.
On both sides of the field, the inoculated Fusarium species were detected, but in different frequencies and abundances regarding the irrigation. A higher Fusarium load on wheat ears was found at the sampling points near to the grass stripe compared to the middle and end of the transect (at the opposite side of the field). Therefore, it can be assumed that the Fusarium species inoculated in the grass strip have spread into the wheat field, even if the final confirmation that they are the same strains is lacking. The results show clearly that grasses at the edges of fields can act as reservoirs for Fusarium and spread the infection on the wheat plants growing next to them, especially when the grasses are highly infected. Gerling et al. (2022) already show that concerning arable weeds, grasses are the most Fusarium-infected weed species in Brandenburg. The influence of the grass stripe on the infection of the wheat plants is enhanced by temperature and moisture conditions, especially at the flowering stage (Leplat et al. 2013;Martínez et al. 2020). This effect was most notable for F. graminearum.
Grasses seem to be an attractive alternative host for different Fusarium species, especially when the main host is absent. Therefore, grasses as non-crop host plants and reservoirs for fungal pathogens will become more important in the next decades and need to be considered in the FHB management. Due to increasing temperatures, F. graminearum may become the dominant species in further parts of (northern-) Europe. Knowledge about a possible shift in the FHB species profile is essential, first of all since F. graminearum is considered the most important trigger of FHB and as the most aggressive DON producer. According to our study, F. graminearum, under beneficial microclimatic conditions, has the potential to spread from a source of infection up to 33 m into a wheat field. As the global food demand is expected to double by 2050 (Godfray et al. 2010), it is more important than ever to ensure the food security of wheat due to an improved FHB management. Knowledge about changes in the intensity of FHB epidemics caused by changing climatic conditions is key for an effective management of FHB in the future. By trying to increase the biodiversity in arable lands, the number of semi-natural landscape elements (e.g., hedgerows, small forests, flower stripes) will increase and along with this also the amount of transition zones between these two habitats.
Our results show that grasses at the edges of wheat fields increase the risk of an infection with FHB for the adjacent wheat field, mainly for the field parts directly next to the grasses. Because of this, we recommend to take grasses into account as a source of infection for FHB, especially after rain events while flowering or when the field is irrigated. Future work is needed to analyze the influence of grasses on the abundance and distribution of more different Fusarium species into wheat fields, especially those belonging to the FHB complex.