The fungal pathogen Microdochium majus, causing snow mould, seedling blight and foot rot, results in severe yield losses in small grain cereals. There are few options to control this pathogen in organic production. In this study, aqueous extracts or botanical powders prepared from chamomile, meadowsweet, thyme and Chinese galls were tested in vitro against M. majus conidia germination and mycelial growth, respectively. Subsequently, three botanicals were chosen, applied as powders with different seed coating adhesives, and tested for their effect on the incidence of M. majus from naturally infected wheat seed lots and on seedling emergence from soil under controlled environmental conditions. Furthermore, seed treatments with warm water, a bacterial product or one chosen botanical were tested in a growth chamber and in a field experiment over three consecutive years. Of the botanicals tested, Chinese galls showed the highest efficacy in controlling M. majus, reducing conidia germination and mycelial growth by up to 97 and 100%, respectively, and reducing the incidence from infested seeds by up to 59%. In two growth chamber experiments, total seedling emergence increased by up to 30 and 59% compared with the control treatments following an application with Chinese galls. Under field conditions, yield increase through Chinese galls, the bacterial product and the warm water treatment was 19, 10 and 37% compared with the untreated control, respectively. This study demonstrates the potential of Chinese galls to control M. majus in wheat. Options for improved formulations or combinations of heat-based treatments with Chinese galls are discussed.
The fungal pathogens Microdochium majus (teleomorph Monographella nivalis) and M. nivale belong to the most important seed and soilborne pathogens of winter cereals and perennial grasses in temperate and cold climatic areas (Tronsmo et al., 2001), causing (pink) snow mould, seedling blight (Cassini, 1981) and foot rot (Pettitt et al., 1993). The main hosts of these Microdochium species are wheat, rye, triticale, barley and oats, as well as turf and forage grasses (Cristani, 1992; Tronsmo et al., 2001).
Infection by Microdochium species can result in severe yield losses due to strongly reduced emergence (Cristani, 1992; Humphreys et al., 1995). For example, in a field experiment in Ireland with different winter wheat varieties, plant establishment after sowing ranged between 26 and 69%, with a significant negative correlation between establishment and M. nivale incidence of the seeds (Humphreys et al., 1995). In organic agriculture, treatments with synthetic fungicides are prohibited, hence other measures are needed to prevent yield loss. Resistant or less susceptible varieties would be an option, but despite several studies including cold-hardening of seeds (e.g. Ergon et al., 2003) or pretreatment with resistance inducers (e.g. Hofgaard et al., 2010), resistance of winter cereals to Microdochium spp. has not been satisfactory (Gołębiowska & Wędzony, 2009). With respect to the effect of sowing time, contradictory results were obtained. Serenius et al. (2005) observed less infection of winter rye by M. nivale with postponed sowing in autumn 2 weeks after the recommended sowing time, whereas Millar & Colhoun (1969) reported that seedling blight incidence caused by M. nivale was greatest in cold and dry soil conditions.
Another option in organic agriculture is the direct control of seedborne snow mould by treatments with hot humid air (Forsberg, 2004) or warm water (Winter et al., 1998; Waldow et al., 2007), which efficiently reduces the incidence of M. nivale. However, this method has not been implemented in cereal production, because seed batches have to be dried afterwards. Electron seed treatment (Eschrig et al., 2007) has been shown to be very effective against pathogens situated on the seed surface but proved to have only minor effects against Microdochium species, which are often located within the embryo. Application of compost (Boulter et al., 2002) as well as treatments with bacteria (Johansson et al., 2003) or fungi (e.g. Smith & Davidson, 1979) have been tested in vitro or under greenhouse and field conditions. Some of the bacterial isolates applied as seed treatments showed disease-suppressing effects that were not significantly different from treatments with a synthetic fungicide (Johansson et al., 2003). As of yet, Cerall®, a formulation containing a strain of the soil bacterium Pseudomonas chlororaphis (Johnsson et al., 1998) and currently approved for use in Austria, Finland, Germany, Lithuania, Sweden and Switzerland, is the only registered biocontrol product against Microdochium spp. and a number of other seedborne diseases in wheat, rye, triticale and spelt. Still, Waldow et al. (2007) and Krebs et al. (2011) observed only partial effects.
Botanicals based on preparations from medicinal and aromatic plants have high potential for the control of various fungal pathogens. Frequent examples include garlic (Allium sativum), caraway (Carum carvi), meadowsweet (Filipendula ulmaria), lavender (Lavandula stoechas), chamomile (Matricaria chamomilla), rosemary (Rosmarinus officinalis), sage (Salvia fruticosa), thyme (Thymus vulgaris) and yucca (Yucca schidigera) against fungal pathogens from the genera Aspergillus, Candida, Cladosporium, Fusarium, Leptosphaeria, Penicillium and Verticillium (e.g. Giordani et al., 2004; Kumar et al., 2008; Tolouee et al., 2010; Wulff et al., 2012). Another plant-based preparation that has been investigated is the material from ground Chinese galls (Galla chinensis; syn. G. rhois, Chinese sumac, Wu Bei Zi; Tian et al., 2009). These galls are produced by aphids feeding on leaves of Chinese sumacs or nutgall trees (Rhus spp.; Ahn et al., 2005).
For the application against fungal diseases with medicinal and aromatic plants, extracts or essential oils have been commonly used (e.g. Tinivella et al., 2009). In contrast to spray treatments onto the foliage, the formulation and delivery of botanicals onto seeds represent a greater challenge in terms of adhesion, persistence and lasting efficacy. For seed treatments with biological control agents, powder applications were reported to be successful (e.g. Sabaratnam & Traquair, 2002). So far, only one plant-based powder preparation has been registered and commercialized; Tillecur®, a product containing yellow mustard flour, is highly effective against common bunt (Tilletia caries; syn. T. tritici) in wheat (Koch et al., 2006). A wettable powder formulation or seed coating (Taylor & Harman, 1990) could be promising for other plant-based preparations because all possible active ingredients are present in the product. In order to achieve sufficient adherence, persistence and possibly a slow-release effect (McQuilken et al., 1998), the choice of an appropriate adhesive is crucial. Numerous compounds have been tested, such as starch (e.g. Adekunle et al., 2001), talcum (e.g. Sabaratnam & Traquair, 2002) and various commercial products or synthetic binders (e.g. Legro, 2004). As of yet, no sufficiently effective and biocompatible product is available for control of M. nivale or M. majus.
The first aim of this study was to investigate the effect of four different botanicals on in vitro conidial germination and mycelium growth of a selected strain of M. majus. Secondly, the three botanicals with the best in vitro efficacy were chosen to test their effect on the growth of M. majus from infested wheat seeds when applied with different formulation methods. The third aim was to investigate the efficacy of the botanical with the best in vitro and in planta results in a climate chamber experiment on emergence of M. majus-infested wheat seeds, and under field conditions towards its potential to increase plant emergence and yield.
Materials and methods
Fungal material and inoculum production of starter cultures
For the experiments on conidial germination and mycelium growth, a single conidium strain of M. majus (Mm0327) was selected. The strain was isolated in 2003 in an experimental field in Zurich-Reckenholz, Switzerland, from grains of the winter wheat cultivar Runal and deposited as CBS 121295 at the public culture collection of the Fungal Biodiversity Centre, Centraalbureau voor Schimmelcultures, The Netherlands. At the research station Agroscope Reckenholz-Tänikon, a stock culture was maintained at 5°C in screw cap slants (15 cm, diameter 1·5 cm), filled to three quarters of its volume with autoclaved shrub soil (‘Staudenerde’ (41% white peat, 36% bark humus, 20% expanded clay, 3% clay, pH 5·5–6·3), Obiter), amended with 1% ground rolled oats and moistened with 3 mL sterile deionized water. For molecular identification of the M. majus strain (in vitro experiments) and for seed lots infected with M. majus (in planta experiments), a forward primer for M. nivale (EFNiv/F) and M. majus (EFMaj/F) and a reverse primer common for both species (EFMic/R; Glynn et al., 2005) were used and confirmed the species status for both the strain and the pathogen within the seeds. Starter cultures of fungal inoculum were produced by placing individual aliquots from stock cultures in 9 cm diameter Petri plates containing potato dextrose agar (39 g L−1, CM0139; Oxoid Ltd), amended with streptomycin sulphate (0·1 g L−1; Fluka, Sigma-Aldrich; PDA +) after autoclaving for 20 min at 121°C. Plates were incubated for 6 to 7 days at 19 ± 1°C with a photoperiod of 12 h dark/12 h near-UV light.
Dried and chopped flowers of Matricaria chamomilla (origin: Egypt), flowers of Filipendula ulmaria (origin: Bulgaria and Poland) and whole plants of Thymus vulgaris (origin: Poland and Peru), purchased from Hänseler AG and from Berg-Apotheke, were finely ground with a centrifugal mill (mesh size 0·08 mm; Retsch ZM 200, Schieritz & Hauenstein AG). Meal of Galla chinensis galls (origin: Sichuan, China; purchased from Berg-Apotheke) was reduced to the same mesh size.
In vitro experiment – conidial germination
For this experiment, aqueous extracts were used instead of plant-based powders, because preliminary trials showed that germinating conidia are hardly visible on agar containing powder particles. For each of the four botanicals, 10 g powder were suspended in 100 mL autoclaved deionized water and stirred for 3 h at ambient temperature. The aqueous suspensions were subsequently filtered using fluted filters (diameter 15 cm, 520 A ½, Schleicher & Schuell). Three concentrations of the extracts were tested: 0·1, 0·5 and 1%. Microscope slides (76 × 26 mm) were placed in 9 cm diameter Petri plates onto moistened (2 mL sterile deionized water) filter papers (diameter 8·5 cm, Nr. 591, Schleicher & Schuell) and three water agar plugs (1 cm diameter) were placed on each slide. Each treatment consisted of two Petri plates, resulting in a total of six agar plugs. Conidial suspensions were obtained by washing each incubated plate of the starter cultures with 7 mL sterile deionized water and adjusting the resulting suspension to a concentration of 3·3 × 104 conidia mL−1. For each botanical and each concentration, 15 μL extract were pipetted onto each agar plug. Sterile, deionized water served as the control treatment. To compare the efficacy of the botanicals with a synthetic fungicide, a treatment with Pronto® Plus (0·035%; 25·5% spiroxamine, 13·6% tebuconazole) was included. Extracts, water and fungicide solutions were allowed to evaporate for 20 min. Subsequently, 15 μL conidial suspensions were pipetted onto the agar plugs. Petri plate lids were closed and plugs were incubated for 24 h at 10°C and 70% relative humidity (RH) in the dark. Conidia were killed and stained with one drop of a Pronto® Plus (0·19%) and cotton blue (0·5%) mixture. The germination rate was assessed with the aid of a light microscope (×400 magnification) by determining the ratio of germinated conidia from a total of 30 conidia within three different visual fields. A conidium was assigned as germinated when the germination tube was longer than the width of the conidium.
In vitro experiment – mycelium growth
Autoclaved PDA medium in Schott flasks was placed in a water bath (60°C) and while stirring constantly, amended with streptomycin sulphate and the desired quantities of the four botanical powders before pouring into 9 cm diameter Petri plates. Concentrations of powders were 0·1, 0·5 and 1% (0·1 g, 0·5 g, 1 g powder 100 mL−1 medium, respectively). PDA + without powders served as the control treatment. Using a cork borer, mycelial plugs (diameter 0·5 cm) were cut from starter cultures and for each PDA plate, one plug was placed in the centre with the mycelial side facing the agar. For each treatment, five Petri plates were used. Plates were incubated in the dark at 20 ± 1°C and 50% RH for 6 days. Subsequently, radial growth was determined by measuring the diameter of the fungal colony at two positions (smallest and largest diameter) and calculating the average of both values.
Incubation chamber experiment – incidence of M. majus from infected seeds on agar
Winter wheat seeds (cultivar Siala) infected with M. majus (infection rate 25–30%) were treated with the botanicals F. ulmaria, T. vulgaris and G. chinensis and two different adhesives, DiscoAg-Red (L203; DR) and Organic Binder (A6.6041; OB; Incotec Holding BV). The adhesive OB has been specially developed for this research and its formulation is based on natural components including a polymer which is certified organic through the inspection and certification body ECOCERT of the EU and the National Organic Program of the USA. For each botanical, 2 g were used for 100 g seeds and for each adhesive, two different application methods were evaluated. The first method consisted of dispersing the powder with water and the adhesive in a liquid seed treater (Hege 11, inductor 190 V, volume 20 to 3000 g seeds, Hege Maschinenbau), resulting in a slurry that was applied in a single step onto the seeds. For the second method, seeds were coated using a ‘sandwich’ technique (McQuilken et al., 1998), in which seeds were first coated with the adhesive (Hege 11), followed by a powder application with a rotating machine (Turbula®, type 2A, 3 × 380 V, Willy A. Bachofen AG) and a second layer of the adhesive (Hege 11). Several preliminary trials were conducted to determine the most suitable concentration and amount of adhesive in terms of viscosity and adherence of the botanical (data not shown). For the slurry method, 100 g seeds were treated with 4·7 mL of DR or OB (both 40%) for F. ulmaria and T. vulgaris and with 4·5 mL of DR or OB (40%) for G. chinensis. For the coating method (‘sandwich’ technique), 3·5 mL (2 × 1·75 mL) of DR or OB (50%) were used for all three botanicals. Treatments with the adhesives but without the botanicals, as well as grains without any treatment, were included as controls. Following treatment, seeds were air dried at 30°C for approximately 60 min. The treated and untreated seeds were placed on PDA + agar in 9 cm diameter Petri plates. For each treatment, 10 plates with 10 seeds in each plate were used. Plates were incubated for 6 days at 19 ± 1°C with a photoperiod of 12 h dark/12 h near-UV light. Subsequently, the number of M. majus colonies growing from seeds was determined and expressed as incidence percentage.
Growth chamber experiment – seedling emergence from soil
Experiment 1 – three botanicals, one application method
Winter wheat seeds (cultivar Fiorina) infected with M. majus (infection rate 30%) were treated at Incotec with F. ulmaria, T. vulgaris and G. chinensis and the two adhesives DR and OB. For this, an encrusting and pelleting process (EPM03) with a rotary batch coater (Satec concept ML 2000, diameter 30 cm, SATEC seed coating) was used. For 100 g seed, 2 g of the respective botanical and 3·5 mL of DR or OB (50%) in complete mixes (slurry) were used. Treated seeds were dried for 5 min in unheated air. Treatments with the adhesives but without botanicals and untreated seeds served as controls. Seeds were sown in plastic trays (dimensions: 30 × 47 × 6 cm) containing moistened shrub soil (Obiter) at a depth of 2 cm. For each treatment, three trays with 100 seeds each (10 rows with 10 seeds) were sown. After sowing, the soil was watered and trays were placed in a greenhouse for 24 h at 20 ± 1°C to stimulate germination. Trays were then wrapped in plastic bags, transferred to a growth chamber and incubated for 21 days at 5°C in the dark without watering. Subsequently, trays were unwrapped and further incubated for 14 days at 10°C in the light (fluorescence and red light, 350 μmol m−2 s−1) and watered as needed. At the end of the second incubation period, the number of emerged seedlings was counted and the ratio of healthy looking and abnormal seedlings (twisted, truncated, without coleoptiles) was determined.
Experiment 2 – one botanical, different application methods
As all in vitro experiments and the seedling emergence in soil from Experiment 1 demonstrated a superior effect of G. chinensis compared with the other botanicals, the following experiments were conducted solely with G. chinensis. As above, 2 g of G. chinensis powder for 100 g seed (winter wheat cv. Siala, infection rate 35%) were used. With respect to the adhesives and for 100 g seed, 4·5 mL of DR or OB (40%) were applied for the slurry method, whereas 3·5 mL (2 × 1·75 mL) of DR or OB (50%) as described for the incubation chamber experiment were used for the coating method. In order to compare the effects of these small-scale seed treatments conducted at Agroscope ART with those from large-scale seed treatments, the coating method by Incotec as described for Experiment 1 was also included. Furthermore, to compare the effect of botanicals with a physical and a biological method, a treatment with warm water (45°C, 2 h; Winter et al., 1998) and a treatment with the bacterial product Cerall® (active ingredient Pseudomonas chlororaphis, 1 mL for 100 g seeds; Stähler Suisse SA) were tested as well. Treatments with the adhesives but without G. chinensis and untreated seeds served as controls. The number of trays and seeds per tray for each treatment was the same as in Experiment 1. Seeds were sown, incubated and seedling emergence was rated as described above.
Field experiment – plant emergence and yield
Over three consecutive years, the effect of seed treatments with G. chinensis powder was evaluated in the field. Treatments were the same as in the growth chamber experiments, except that for the adhesive, only OB was used. For the field experiments sown in 2008 and 2009, a single winter wheat seed lot of cv. Siala infected with M. majus (35% infection rate) was chosen. For the experiment sown in 2010, another seed lot of Siala (25–30% infection rate) was used.
The experiments were carried out on the experimental farm of the Research Station Agroscope ART in Zurich-Reckenholz. The soil type in 2008 and 2009 was a loamy anthrosol with 2·8 and 2·2% organic matter, respectively. In 2010, the soil type was a loamy cambisol with 2·4% organic matter. Plot size was 1·2 × 8·7 m and wheat was drilled at 150 kg ha−1. Each treatment consisted of four plots. The sowing dates for the 3 years were 10 November 2008, 21 October 2009 and 3 November 2010. Husbandry operations were standard for the farm except that no fungicides were applied. Seedling emergence was determined at growth stages (Zadoks et al., 1974) DC 11 to DC 12 after snow-melt. For each plot, four rows were randomly selected while excluding the border rows, and the number of emerged seedlings within 1 m of each row was counted. Assessment of seedling emergence for the different years took place on 18 March 2009, 24 March 2010 and 9 March 2011. Plots were combine-harvested on 30 July 2009, 1 August 2010 and 25 July 2011. Wheat grains were passed through a grain-cleaning machine (aspiration cleaner Kongskilde KF12, Kongskilde Industries) to remove harvest by-products and grain yield (t ha−1) was determined at approximately 14% moisture content. Weather data were obtained from a MeteoSwiss operated weather station (SwissMetNet) located at Zurich-Reckenholz c. 0·5 to 1 km from the experimental sites. Data included daily mean air temperature (at 2 m height) and sum of precipitation (1·5 m) and were taken according to WMO guidelines (Anonymous, 2008).
Experimental design and analysis
All laboratory, incubation and growth chamber experiments were set up in a completely randomized design and were performed twice. The field experiment was set up in a randomized complete block design and was conducted three times. For all experiments, results from the experimental runs were pooled. In case of a failed normality test and in order to approach normal distribution, percentage data (conidial germination rate, incidence of M. majus from grains, seedling emergence in growth chambers) were arcsine transformed whereas data from radial mycelial growth, number of emerged plants in the field and yield were ln transformed before analysis of variance (anova). Apart from one-way anovas analysing the effect of one treatment factor only, two-way anovas were also conducted for experiments where other factors than the botanicals were expected to be important (e.g. application procedure, different adhesives, years). When the overall effect of the tested factor was significant in anova, an all-pairwise multiple comparison procedure according to Holm–Sidak (α = 0·05; Holm, 1979) was conducted in order to evaluate differences between treatment means. For plotting of graphs, untransformed data were used. All statistical analyses were conducted using sigmastat (Systat Software).
In vitro experiment – conidial germination
The mean germination rate of M. majus conidia from the control treatment was 96%. The treatment with the synthetic fungicide Pronto® Plus completely inhibited germination (data not shown). When data from all three concentrations were combined, the reduction of germination through the four botanicals ranged between 1% (T. vulgaris – mean of 95% germinated conidia) and 60% (G. chinensis – mean of 39% germinated conidia). The reduction was highly significant (P <0·001) for G. chinensis at 0·5 and 1·0% as well as for F. ulmaria at 0·5% (Fig. 1). None of the treatments with M. chamomilla or T. vulgaris significantly reduced the germination rate. The greatest reduction of 97% (2·5% germination) was obtained with G. chinensis at 1·0%.
In vitro experiment – mycelium growth
For the control treatment, the mean colony diameter of M. majus after 6 days incubation was 7·1 cm. When data from all three concentrations of incorporated powders were combined, reduction of colony growth was greatest with G. chinensis (mean diameter 0·9 cm). Clearly less effective were incorporations by F. ulmaria (4·1 cm), T. vulgaris (5·3 cm) and M. chamomilla (6·6 cm; Fig. 2). Highly significant (P <0·001) reductions of mycelial growth were obtained with G. chinensis at all three concentrations, for F. ulmaria at 0·5 and 1·0%, for T. vulgaris at 0·5 and 1·0% and for M. chamomilla at 1·0%. Incorporation of G. chinensis at 0·5 and 1·0% completely inhibited mycelial growth of M. majus.
Incubation chamber experiment – incidence of M. majus from infected seeds on agar
The mean incidence of M. majus based on the number of colonies from untreated seeds was 24% compared with 23 and 24% from seeds treated with the adhesives OB and DR, respectively, with no significant differences. For the treatments containing adhesives and botanicals, no significant difference was found between the coating and the slurry application technique (data not shown). Overall, the reduction of the fungal incidence was significantly better (P <0·006) when botanicals were applied with the adhesive OB, compared with the adhesive DR (Fig. 3). The best effect was obtained with G. chinensis, followed by F. ulmaria. The efficacy in reducing fungal incidence was significant (P <0·001) for both G. chinensis treatments and for the F. ulmaria treatment applied with the adhesive OB. The alteration of M. majus incidence through the botanicals compared with the control treatment ranged from an increase of 12% (DR + T. vulgaris: 27% incidence) to a decrease of 59% (OB + G. chinensis: 10% incidence; Fig. 3).
Growth chamber experiment – seedling emergence from soil
In Experiment 1 with three botanicals and the two adhesives applied by Incotec, mean emergence of total and healthy seedlings from the control treatment was 49 and 42%, respectively. The adhesives DR and OB without botanicals showed no significant effect on total seedling emergence (Fig. 4). Treatment with botanicals resulted in a mean increase of total emerged seedlings between 16 and 59%. Highly significant (P <0·001) effects were observed for both treatments with G. chinensis (emergence with DR: 72%, with OB: 77%), both T. vulgaris treatments (with DR: 61%, with OB: 59%) and for the F. ulmaria treatment with OB (60%) (Fig. 4). The effect of the botanicals on the ratio of healthy seedlings was similar. However, apart from the superior G. chinensis treatments (with DR: 66%, with OB: 71%), only the T. vulgaris treatment with DR (55%) significantly (P <0·001) increased the ratio of healthy looking seedlings when compared with the control treatments (data not shown).
In Experiment 2 using G. chinensis applied with different techniques as well as a warm water and a Cerall® treatment, the mean emergence of total and healthy seedlings from the control treatment was 53 and 42%, respectively. None of the treatments with the adhesive OB without G. chinensis showed a significant effect on total seedling emergence (Fig. 5). There was a highly significant (P <0·001) increase of total and healthy emerged seedlings following a treatment including G. chinensis with different application techniques, with mean emergence between 65 and 69% and between 54 and 58%, respectively. The application techniques used by Agroscope ART led to slightly higher total seedling emergence compared with the technique by Incotec, but the differences were not significant (Fig. 5). The emergence following a treatment with Cerall® was similar (total emergence: 66%; healthy: 55%) to that with G. chinensis treatments. The effect of the warm water treatment was substantially greater than all other treatments, resulting in a mean emergence of 96% total seedlings (Fig. 5) and 93% healthy seedlings (data not shown).
Field experiment – plant emergence and yield
There were highly significant interactions (P <0·001) between the effect of the year and the treatment on plant emergence and yield. Of the 3 years, the winter after sowing in 2008 was the coldest with the longest snow cover, whereas the winter after sowing in 2010 was the warmest with the lowest amount of precipitation. The lowest and highest temperature immediately after sowing was recorded for 2008 and 2010, respectively (Fig. S1). When data were combined over all treatments, the average number of plants within 1 m of each row showed a large range from 10 (2009) up to 34 plants (2011) and the yield ranged from 3·8 (2009) up to 6·3 t ha−1 (2011). Nevertheless, even when data were combined over all 3 years, differences were observed between the seed treatments. The average number of emerged plants from the control treatment was 15 within 1 m. The number of plants in treatments containing the adhesive OB without G. chinensis ranged between 18 and 23, whereas in the treatments containing OB together with G. chinensis, the number ranged between 26 and 27 (Fig. 6a). The Cerall® treatment resulted on average in only 21 plants and warm water was the best treatment, resulting on average in 39 plants within 1 m (Fig. 6a). When data were combined over all years, the increased emergence was only significant (P <0·001) for the warm water treatment. The two-way anova with the factors year and seed treatment separated, also demonstrated highly significant effects from all treatments containing G. chinensis, from Cerall® and also from the three treatments containing OB only (Fig. 6a). The yield from seeds of the untreated control was 4·7 t ha−1. Seeds treated with G. chinensis resulted in a yield between 5·5 and 5·7 t ha−1, whereas yield from seeds that received only the adhesives ranged between 4·9 and 5·0 t ha−1. The average yield from the Cerall® and the warm water treatment was 5·2 and 6·5 t ha−1, respectively (Fig. 6b). As observed for the emergence, combined over all years, the increase in yield was only significant (P =0·048) for the warm water treatment. However, the two-way anova also showed highly significant effects on yield from all treatments containing G. chinensis on yield (P <0·001; Fig. 6b).
In this study, the effects of various botanicals on the snow mould-causing pathogen M. majus were tested in vitro and in planta. The choice of botanicals was based on literature data and on preliminary experiments. Aerial parts of the medicinal herb Filipendula ulmaria were reported to have both antibacterial and antifungal effects based on phenolics, in particular flavonoids, and on salicylaldehyde (e.g. Rauha et al., 2000). The antifungal effects were confirmed with preliminary experiments in the laboratory using acetone or ethanol extracts of F. ulmaria showing it to be one of the best performing botanicals with respect to reducing conidial germination and mycelium growth (data not shown). Matricaria chamomilla has been reported several times for its antimicrobial, anti-inflammatory and antiseptic properties (cited in Newall et al., 1996) while Tolouee et al. (2010) also observed antifungal effects through reduced growth and conidia production of Aspergillus niger. The antimicrobial effects of M. chamomilla are suggested to be based on various essential oils, with the sesquiterpene alcohol α-bisabolol as one of the main compounds (Tolouee et al., 2010). The antimicrobial properties of the essential oils of Thymus vulgaris have also been demonstrated before (e.g. Manou et al., 1998). In fact, Thymus essential oils were ranked as some of the most potent inhibitory substances against bacteria (Deans & Ritchie, 1987) and fungi (Giordani et al., 2004), which were mainly attributed to two phenolic compounds, thymol and carvacrol (Rota et al., 2008). Galla chinensis galls, used as a traditional Chinese herb, contain several tannin-derived components, such as gallic acid and methyl gallate (Ahn et al., 2005). The range of activities reported by G. chinensis is particularly broad, including inhibition of melanogenesis in melanoma cells (Chen et al., 2009), antidiarrhoeal (Chen et al., 2006), antioxidant, antibacterial (Tian et al., 2009) as well as antifungal (Ahn et al., 2005) effects. The investigation by Ahn et al. (2005) using extracts of G. chinensis demonstrated reduced conidial germination and appressorium formation of the rice blast-causing pathogen Magnaporthe oryzae, as well as reduced disease areas from various fungal plant pathogens in a whole plant bioassay. As these tannin-derived components of G. chinensis display a rather low pH, it could be speculated that the superior effect might also be due to the acidity. The pH of OB was 3·9 and the pH for the suspended botanicals ranged from 3·2 (G. chinensis) to 5·9 (T. vulgaris). However, fungal pathogens thrive well in a large range from highly acid to neutral conditions (e.g. Matthies et al., 1997).
In vitro experiments possess the advantage that a great number of substances can be tested within a limited amount of time. Active substances within a potentially antifungal botanical may bear different modes of action (e.g. effect on cell respiration, biosynthesis of sterols) and may have different target sites within the life cycle of a fungus, including spore germination, penetration, mycelial growth and/or sporulation (Köller, 1992). In the in vitro experiments of this study, G. chinensis was substantially more effective compared with the other botanicals, and higher concentrations of G. chinensis greatly or completely inhibited M. majus conidial germination and mycelial growth. Remarkably, the efficacy of G. chinensis at a concentration of 1% was almost as high as that from the synthetic fungicide Pronto® Plus, resulting in 97 or 100% reduction of conidial germination, respectively. Bearing the different target sites in mind, it was not surprising that the pattern of results on conidial germination was not always equivalent to that on mycelium growth. In fact, treatments with T. vulgaris did not reduce germination of M. majus conidia whereas high concentrations of T. vulgaris powder incorporated into agar significantly reduced mycelial growth. In general, the effects of botanicals on conidial germination were clearly smaller than those observed on mycelium growth, except for G. chinensis. This finding may be explained by the fact that the germination experiment was the only test system using aqueous suspensions and it is possible that the active substances from M. chamomilla, F. ulmaria and T. vulgaris were rendered available in smaller amounts compared with a test system using entire plant powders. For the seedborne pathogen M. majus, it might be more important to control mycelium growth than conidial germination because it is assumed that mycelium between the pericarp and aleurone layer as well as in the endosperm and the embryo infects the seedling after germination (Rennie & Cockerell, 2006). The second best botanical after G. chinensis in reducing mycelium growth of M. majus was F. ulmaria, confirming the preliminary experiments, followed by T. vulgaris. Consequently, M. chamomilla, with the smallest effect on mycelium growth, was excluded for the subsequent in planta experiments.
The findings from the in vitro experiments were confirmed in the following in planta test systems. All G. chinensis treatments had a substantially higher efficacy compared to those from F. ulmaria or T. vulgaris. For all three botanicals, there was a trend showing that the use of the adhesive OB was slightly better in reducing M. majus colony incidence or disease of seedlings compared with the adhesive DR, which is advantageous because OB is certified for use in organic agriculture. The reason for this finding might be due to the fact that the polymer system in the adhesive DR is less hydrophilic and in OB more hydrophilic. Thus, the DR formulation may have a delayed release of active substances that could affect the efficacy.
In contrast to the results of this study, Ahn et al. (2005) observed antifungal effects only from methanol extracts of G. chinensis applied on leaves and not from water or other extracts. This could be due to the different fungal species examined in the present study, the use of the entire powder instead of extracts and the fact that seeds were treated instead of leaves. Other factors might also have played a role. With respect to essential oils, for example, the origin of the plant, the part of the plant used, the stage of plant development, climatic and growth conditions as well as storage conditions might have an effect on the composition of antibacterial or antifungal substances (cited in Kalemba & Kunicka, 2003).
The technique of warm or hot water, air or vapour treatments, also called thermotherapy, represents a widely used method of disease control that has been proven to be efficient against various pathogenic microorganisms. The application of heat to various plant parts at specific temperature/time regimes kills the conserved pathogen and is only slightly injurious to the host (Grondeau et al., 1994). The low moisture content of seeds results in greater resistance to higher temperatures compared with the pathogen, which is rendered sensitive by wetting through the treatment (Forsberg, 2004). The genus Microdochium seems to have a low heat tolerance (Forsberg, 2004), which might explain the superior effect of the warm water treatment in both the growth chamber and the field experiments of the current study. In general, it is important to determine the optimal combination of time and temperature of exposure for the highest efficiency with the least damage to the host. This study was able to apply a protocol that was already established and validated for wheat and various seed-borne pathogens (Winter et al., 1998).
The effect of the bacterial product Cerall® is assumed to be based on various modes of action, including competition for nutrients and space, stimulated growth of seeds, induced resistance and antibiosis through secretion of biochemical products onto the seed in the soil (e.g. Johnsson et al., 1998). In the experiments from this study, only partial effects were observed and these were substantially smaller than those of the G. chinensis treatments.
In the field experiments, a strong year effect with respect to plant emergence and yield was observed: in 2011, plant emergence was more than three times higher and yield was more than one and a half times greater than in 2009. This might be explained by the winter wheat seed lots and the weather conditions. The seed lot used for the experiment in 2009 had an average M. majus infection rate of 35%, whereas the one for the experiment in 2011 had a rate of only 25 to 30%. The temperature after sowing in 2008 was lower than in the following years, leading to slower plant emergence, and the winter of 2008 and 2009 was in general colder with longer snow cover than in the other years, which favoured the development of M. majus.
In the experiments under controlled environment and under field conditions, different seed treatment methods and equipment were tested in order to determine which application would result in the highest efficacy in controlling M. majus. Surprisingly, no difference was found between the large scale industrial encrusting and pelleting method (EPM03) with a rotary batch coater and two small scale methods either with a single step technique with a liquid seed treater resulting in a slurry or with the ‘sandwich’ technique, applying the botanical between two layers of the adhesive. The ‘sandwich’ technique was expected to perform better than the slurry technique because it was assumed that the botanical would be better protected from being washed off or from adverse environmental effects. This result, though, could be advantageous because a single step method with small-scale equipment would allow cereal growers to treat their seed lots at lower costs.
In the field experiment of 2010, a two-way anova (year × treatment) revealed that treatment of M. majus infected seed with the adhesive OB but without a botanical had an effect with respect to increased emergence. Coating formulations with adhesives can facilitate an increased moisture uptake after sowing and thus could have resulted in an accelerated germination process.
In contrast to vegetable crops, cereal crops are of lower commercial value, hence an economical treatment against fungal diseases is needed, which should not exceed US$ 20 ha−1 (Taylor & Harman, 1990), which corresponds to approximately € 15 ha−1. A seed treatment with G. chinensis at a rate of 2 g for 100 g seed and a seeding rate of 150 kg ha−1 comes to approximately € 60 ha−1 if based on a direct import from China (A. Lenherr, Berg-Apotheke, Zurich, Switzerland, personal communication). Furthermore, the costs for the adhesive OB of approximately € 30 ha−1 have to be added. Hence, with the present dose, G. chinensis might not be used despite excellent biological performance. Thus, lower doses should be investigated in future experiments.
The current study has demonstrated that the medicinal herb G. chinensis applied to seeds with an adhesive has the potential to control M. majus through improved plant emergence, resulting in increased yield in infested seed lots. Nevertheless, there is still room for optimization of the application technique. For example, a finer milling grade below 0·08 mm might not only reduce costs due to smaller doses but could also prevent physical loss of the botanical after application and during sowing. In all growth chamber and field experiments, the warm water treatments were clearly superior compared with the bacterial product and the botanicals. As redrying of seeds is too costly for the cereal industry, experiments have been started with the ThermoSeed™ technique applying moist hot air (Forsberg, 2004), which has been used for control of a number of seedborne pathogens, mainly in vegetable seed. In addition, dry hot air has also been shown to be effective against fungal pathogens (Gilbert et al., 2005). In fact, preliminary experiments at Agroscope ART investigating the effect of dry air between 60 and 80°C at different durations on M. majus incidence and wheat seed germination were promising. Currently, seed treatments combining G. chinensis with moist hot air or dry hot air are tested under controlled environment and field conditions in order to control M. majus. By using synergies of botanicals combined with physical treatments, this approach could be promising for efficiently and economically controlling a number of noxious seedborne diseases in organic agriculture.
The authors greatly appreciate the excellent technical assistance by Andreas Hecker, Eveline Jenny and Fabienne Schwab, as well as the valuable support from the field group of Agroscope ART in maintaining the field sites. They would also like to thank the Swiss Federal Office of Meteorology and Climatology for providing meteorological data and Tomke Musa for extracting and interpreting the meteorological data from 3 years of field experiments.