Table 1 shows the interactions on wheat grain at 15 and 25 °C and both 0.995 and 0.955 a_w. F. graminearum and F. poae were dominant over F. culmorum at most a_w levels and the two temperatures. Where the other Fusaria were not dominant over F. culmorum, mutual inhibition on contact type interactions occurred. Other phyllosphere fungi were mostly dominated by F. culmorum in all the conditions tested. The sum of the I_D scores shows that F. culmorum was overall more competitive than the other species, and thus dominant under the conditions examined on wheat grain.

Figure 1 shows the effect of such interactions on the relative growth of F. culmorum on wheat grain. Growth rates were unaffected by interactions with C. herbarum, F. graminearum, F. poae, and M. nivale and M. nivale var majus at 0.995 and 25 °C. At 0.995 a_w and this temperature, all the fungi inhibited growth of F. Culmorum, except for C. herbarum and P. verrucosum. However, at 15 °C, growth of F. culmorum was stimulated by interactions in dual culture with A. tenuissima, P. verrucosum, C. herbarum, and the Microdochium species.

Figure 2 shows the effect of interactions with other species in dual culture on the production of DON at 15 and 25 °C at 0.955 a_w. It is shown that in the presence of M. Nivale, there was stimulation of the production of DON by F. culmorum. In the presence of some species, DON production was inhibited. Table 2 summarizes, based on the statistical analyses, whether DON was stimulated or inhibited or whether there was no effect due to interactions on wheat grain.

The effect of a_w and temperature interactions on relative competitiveness of A. carbonarius against different fungi are shown in Table 3. The relative total I_D values under the different conditions are also shown. Overall, three different interaction types were found in the study. At 30 °C, A. carbonarius generally dominated all the other fungi, scoring 4-0. The exceptions were the other Aspergillus section Nigri species and E. nigrum. At 20 °C, A. carbonarius was only able to dominate the pink yeast. Against all other fungi A. carbonarius was mutually antagonistic using contact scoring (2-2), whilst with E. nigrum it was mutually antagonistic at a distance, scoring 3-3.

The growth rates varied with temperature and a_w level of treatment. When comparing the growth rate of A. carbonarius in the absence of competitors against growth after interactions in dual culture, there was a slight inhibition of the area of colonization. Table 4 shows the relative stimulation or inhibition of growth by competitors when compared to A. carbonarius alone. This shows that depending on the interacting species, there was an effect on growth of the mycotoxigenic OTA producing species.

Figure 3, Figure 4 and Figure 5 show the effect of interactions between A. carbonarius and other fungi on OTA production under different environmental conditions. OTA production was greatest at 20 °C and 0.987 a_w with stimulation occurring when grown against most other species (Figure 3). OTA production decreased when the a_w was decreased (drier conditions) and with increasing temperature. The lowest OTA production was at 30 °C and 0.93 a_w. In general, there was a stimulation of OTA production at 0.987 a_w. However, OTA production was reduced at 0.93 and 0.95 a_w when compared with that produced by A. carbonarius alone. Interaction with B. cinerea consistently resulted in increased OTA production when compared to the control. Competition with the pink yeast (Sporobolomyces species) and Aspergillus section Nigri (uniseriate) also resulted in reduced OTA produced by A. carbonarius than when grown alone.

It is interesting that F. culmorum was often not able to achieve dominance over F. graminearum, regardless of the environmental conditions, and they were mutually antagonistic to each other at 0.955 a_w. This supports published work reported on FEB, which has shown dominance of F. graminearum inoculum and infection in the UK and indeed in Europe [13,14]. The interactions also had a variable effect on colonization by F. culmorum. In some cases, growth was significantly increased (F. culmorum versus A. tenuissima) at 25 °C and both 0.995 and 0.955 a_w. In other cases, growth was significantly reduced (F. culmorum versus F. graminearum, 25 °C and 0.995 a_w). However, changes in growth rate did not seem to correlate with changes in interaction type. For instance, F. culmorum was dominant in contact with A. tenuissima in all conditions assayed at 25 °C with growth was significantly stimulated by this interaction, whereas at 15 °C there was no difference from the control.

Earlier interaction studies by Marin et al. [8] examined Fusarium species on maize and found that there was no correlation between populations of F. verticillioides and four other interacting species and interaction scores. Studies by Lee and Magan [7] found that A. ochraceus growth rates were reduced by the presence of other fungi on maize and no stimulation observed, regardless of temperature and a_w levels tested. This suggests that growth rate alone does not determine competitiveness of a specific species per se. Recent studies have suggested that carbon nutritional patterns may also play a significant role in the level of niche occupation or niche exclusion, and that this is further influenced by the changing environmental factors [4,9].

Practically no studies have been carried out to evaluate the impact of interactions on DON production by F. culmorum. Changes in DON concentrations occurring while interactions remain the same may indicate a change in combative strategy by the fungi in response to environmental stress. For example, F. graminearum was able to dominate F. culmorum on wheat grain at 0.995 and 0.955 a_w. However, at the higher a_w level, DON was significantly reduced, whereas at 0.955 a_w DON levels were increased. As they both produce DON it is possible that under water stress they both increase production of DON resulting in the increased concentrations observed. This could indicate a physiological change in response to ecological pressures, even though the interaction outcomes were macroscopically the same. However, generally it has been suggested that when some fungi are in close proximity to each other, mycotoxin production is stimulated; this is thought to be because the fungi are trying to pre-emptively exclude other competitors [3,4].

For example, Ramakrishna et al. [12] found that T-2 toxin production by Fusarium sporotrichioides was generally inhibited when grown in dual culture with Aspergillus flavus, P. verrucosum and Hyphopichia anomala on barley grain. However, in the present study some stimulation of toxin production was observed under some interacting environmental conditions.

Very few studies have examined the effect the competitiveness of A. carbonarius against other mycobiota of grapes. The effects of a_w and temperature suggest that at 30 °C A. carbonarius is very competitive and able to outcompete and dominate all the non-Aspergillus species, except E. nigrum. A. carbonarius has an optimum growth temperature of around 30 °C [15] and thus would certainly be able to colonize grape-based matrices quickly over a range of a_w conditions. The competitiveness of A. carbonarius in these conditions was also supported by the I_D results and the relative influence on growth rates. At lower temperature conditions, which are sub-optimal for growth (20-25 °C), A. carbonarius was less competitive, except against the pink yeast strain. Under most conditions there was mutual antagonism on contact. Both E. nigrum and Aspergillus section nigri (uniseriate) showed signs of being antagonistic at a distance suggesting the influence of secondary metabolites. These two species have a wide optimum growth range, and in the case of E. nigrum, a higher optimal a_w as well [5]. However, it can compete effectively as it is also a biocontrol agent of some preharvest pathogens and produces a wide range of pigments and secondary metabolites [16].

With regard to OTA production, we had nine different interacting species and nine different environmental conditions. This resulted in outcomes from interactions being both stimulatory and inhibitory to OTA production by A. carbonarius being observed. At 30 °C and with freely available water (0.98 a_w), OTA production was decreased in the presence of all competitive fungi except for A. pullulans, A. alternata and the pink yeast. These were the species which stimulated growth of A. carbonarius. At 20 °C and 0.98 a_w there was a stimulation of OTA production by A. carbonarius when grown in the presence of all species except the pink yeast and the Aspergillus section Nigri (uniseriate) species. As a_w was reduced to 0.95 a_w at this temperature OTA production was suppressed by interaction with all species except Cladosporium species. Previous ecological studies suggest that 20-25 °C and 0.95 a_w are optimum conditions for OTA production [15]. This is very different from those for optimum growth which are 30 °C and 0.98 a_w. Thus, under a_w stress, more OTA may be produced as a defence reaction against competitors to maintain colonization/occupation of the niche. Of course at 0.95 a_w most of the competing fungi (phyllosphere species) are under some stress with reduced growth. Thus the effect of competition in relation to growth and OTA production may be quite complex. The ability of B. cinerea to inhibit OTA production is interesting as it is a saprophyte which normally infects senescing leaves and plant material. It colonizes damaged grapes under high humidity conditions and is able to produce high amounts of hydrolytic enzymes. It may be that this enables B. cinerea to compete effectively with A. carbonarius under some conditions and prevent the biosynthesis of OTA. This again suggests that the interactions between fungi are complex and the influence of changing environmental conditions will influence not only the outcome but the role of mycotoxins in competition.

Table 5 lists the fungi used in interaction studies related to (a) grain (b) grape-based ecosystems.

Wheat-based studies: For wheat mycobiota interaction studies and effect on DON production, experiments were initially conducted on a 2% milled wheat agar medium modified with glycerol to the required a_w levels (0.995, 0.98, 0.955) at 15 and 25 °C. Subsequent studies and those reported in this paper were carried out on layers of wheat grain modified to the required a_w levels by reference to a moisture adsorption curve at 0.995 and 0.955. After equilibration of irradiated (12 kGys) grain, the single layers were placed in sterile 9 cm Petri plates and placed in polyethylene containers containing glycerol/water solutions to maintain the equilibrium relative humidity at the target level. The layers of grain were inoculated with the test fungi using agar plugs or spore suspensions 4 cm apart. The growth of each species was measured over a period of 10-14 days and the type of interactions determined. The experiments were carried out with three replicates per treatment and performed twice.

Grape-based studies: Studies were carried out on a synthetic grape-based medium (SGM) representative of mid-veraison [15]. This consisted of D(+) glucose 70g, D(-) fructose 30g, L(-) tartaric acid 7g, L(-) malic acid 10 g, (NH₄)₂HPO₄ 0.67 g KH₂PO₄ 0.67 g, MgSO₄ 7 H₂O 1.5 g NaCl 0.15 g CaCl₂ 0.15 g CuCl₂ 0.0015 g, FeSO₄ 7 H₂O 0.021 g, ZnSO₄ 7 H₂O 0.0075 g, (+) Catechin hydrate 0.05 g, agar 25 g in a litre of medium. This was adjusted with 2 M KOH to pH 4.0-4.2. The studies were carried out at 0.98 and 0.95 a_w,at 25 and 30 °C. Seven day old colonies were used to prepare spore suspensions of (1 × 10⁶) of A. carbonarius, B. cinerea, Phoma sp, pink yeast, white yeast, Aspergillus section Nigri biseriate and uniseriate. Due to problems in harvesting spores of E. nigrum and A. alternate, mycelium from the growing edge of colonies was included in the spore suspension. SGM plates were then inoculated using a 1 μL calibrated loop, with A. carbonarius and one of the other species inoculated approximately 4 cm apart. Controls were inoculated centrally using the 1 μL loop. Each treatment was replicated three times and plates of the same a_w were sealed in plastic bags. The plates were incubated for 15 days and analyzed for OTA production. All experiments were carried out twice with three replicates on each occasion.

Colony diameter was measured by taking two measurements at right angles to each other during the incubation period to calculate a growth rate by linear regression. The colonies were checked regularly for interactions by macroscopic and microscopic analysis. Each interaction was given a score based on mutual intermingling (1-1), mutual antagonism on contact (2-2), mutual antagonism at a distance (3-3), dominance of one species on contact (4-0) and dominance at a distance (5-0). In the case of the dominant interactions, the higher score was always awarded to the more competitive fungus [5]. For example, if A. carbonarius was dominant over B. cinerea upon contact this would result in a 4 and 0, respectively, being awarded to the two fungal species. The scores for each species were totalled to give an overall Index of Dominancy (I_D) value as a measure of competitiveness.

Deoxynivalenol quantification: The method was adapted from Cooney et al. [17]. Grain samples were first dried overnight at 50 °C, milled and a 10 g sub-sample was placed in 40 mL acetonitrile/water (14:1). This was shaken for 2 hr on a rotary shaker before cleanup using an in house cartridge. This cartridge consisted of a 2 mL syringe (Fisher Ltd) packed with a disc of filter paper (Whatman No. 1), a 5 mL lugger of glass wool and 500 mg of alumina/activated carbon (20:1). The sample was allowed to gravity feed through the cartridge. Residues on the cartridge were washed out with acetonitrile/methanol/water (80:5:15; 500 uL). The combined eluate was evaporated to dryness and re-suspended in methanol/water (5:95; 500 uL).

Quantification of DON was carried out using HPLC using a Luna column (100 mm × 4.6 mm i.d., Phenomonex). Separation was achieved using a isocratic mobile phase of methanol/water (12:88) at 1.5 mL/min. Eluates were detected using a UV detector set at 220 nm and an attenuation of 0.01 AUFS. The retention times for DON was 7.5 min and the limit of quantification for DON was 120 ng/mL.

Ochratoxin quantification: Up to 6 agar plugs (4.5 mm diameter) were removed across the mycotoxigenic strain including the interaction zone. These samples were placed into 28 mL Universal bottles and 5 mL methanol was added. The samples were shaken for 1 hr. The extracts were filtered through fluted filter paper (Whatman No 1) containing 0.25 g of Celite^® 545 to improve filtration. This was filtered through a 0.22 um filter (Millex® HV 13mm, Millipore) directly into amber HPLC vials (Jaytee Biosciences LTD, UK) and stored at 4 °C until HPLC analysis was performed.

The quantification method used was adapted from Bragulat et al. [18]. The HPLC system consisted of a Millipore Waters 600E system controller, a Millipore 712 WISP autosampler and a Millipore Waters 470 scanning fluorescence detector (Millipore Corporation Massachusetts USA)(excitation 330 nm, emission 460 nm). The samples were separated using a C18 Luna Spherisorb ODS2 (150 × 4.6 mm, 5 μm) (Pheonomenex), with a guard column of the same material used to extend the column life and reduce drift. Run time for samples was 10 min. The flow rate of the mobile phase (57% acetonitrile, 41% water and 2% acetic acid) was 1 ml/min. The recovery rate for OTA was 88% from grape juice based media with a limit of detection of 0.01µg/g medium. Analysis of the results was carried out on a computer running Kroma system 2000 operating system (Bio-tek Instruments, Milan, Italy).