Population dynamics of Rhizoctonia , Oculimacula , and Microdochium species in soil, roots, and stems of English wheat crops

This study aimed to elucidate the population dynamics of Rhizoctonia , Oculimacula , and Microdochium species, causing the stem base disease complex of sharp eyespot, eyespot, and brown foot rot in cereals. Pathogen DNA in soil, roots, and stem frac tions, and disease expression were quantified in 102 English wheat fields in two seasons. Weather data for each site was collected to determine patterns that cor-relate

populations of these species in soil and in planta can change due to selection pressure from environmental (e.g., rainfall, temperature) and agronomic (e.g., fungicides) factors (Turner et al., 2002).
The last studies in the UK, more than 20 years ago (Nicholson & Turner, 2000;Turner et al., 1999), showed that O. acuformis was the main pathogen causing eyespot. However, since then, the introduction and use of the fungicide cyprodinil with increased effectiveness against O. acuformis (Parnell et al., 2008) is expected to have caused changes in the eyespot population in favour of O. yallundae. Turner et al. (1999) also reported M. majus as the major BFR pathogen, and although seasonal variations have been known to occur (Nicholson & Turner, 2000), the pathogen predominance in Microdochium populations remains unclear. The distribution, relative abundance, and pathogenic activity of R. cerealis worldwide is presumed to be increasing due to climate change and changes in agronomic practices such as earlier sowing dates (Hamada et al., 2011); however, the incidence of this pathogen in UK wheat is unknown. Furthermore, limited information exists on the populations of other Rhizoctonia spp. likely to cause disease in English wheat crops. R. solani is a soilborne pathogen species complex of 13 anastomosis groups (AGs) (Carling et al., 2002) causing diseases in a broad range of crops including wheat (Ogoshi, 1996). Symptoms caused by R. solani include pre-and post-emergence damping off, root rot, foliar blight, and stem rots. R. solani AGs 2-1, 4, 5, 8, and 11 are known to cause disease on the roots or stems of wheat (Woodhall et al., 2012a). Isolates of R. solani attack young roots of their hosts (Harris & Moen, 1985), and the most recent study has shown that AG 2-1 is capable of reducing the number of primary roots, root volume, and root surface area on 6-day-old wheat seedlings (Sturrock et al., 2015).
Environmental factors such as meteorological conditions influence the development of fungal pathogens in soil and in planta and consequently alter root and stem base disease development and severity. For example, moisture absorbed by infected straw, on which the pathogen resides, is required for sporulation of ascospores and conidia (Rowe & Powelson, 1973) for eyespot infection to occur. Regular rainfall events and cooler temperatures also favour the development of M. nivale and M. majus in planta (Xu et al., 2008), whilst ambient temperature increases from 16 to 28 °C reduce the disease index for wheat seedlings infected with R. cerealis and halts disease progression (Burpee et al., 1980). Information on the relative abundance and niche of species within the pathogen complex during the physiological development of the host informs us on the risk of yield loss associated with their stem base diseases.
Such information can be used to improve local management decisions, because the effectiveness of control methods differ for individual species within the complex.
The aim of this study was to define the population dynamics of Oculimacula, Microdochium, and Rhizoctonia spp. as the principal pathogens causing SBD in soil and on stems of naturally infected English winter wheat crops during two seasons. The individual objectives were to (a) determine the incidence and severity of diseases on the roots and stems of winter wheat crops; (b) quantify targeted species of Rhizoctonia spp., Oculimacula spp., and Microdochium spp. at different developmental stages of the crop using quantitative real-time PCR (qPCR); and (c) identify meteorological factors that influence disease severity and pathogen DNA accumulation in soil and in planta.

| Soil and plant sampling
The present study sampled 102 commercial winter wheat fields on 22 farms in 20 counties across England. Farms were chosen to give a representative sample of winter wheat fields in the north, Midlands, and south of England. On each farm the fields were randomly identified except for the criteria that wheat predominated in the rotation.
This resulted in soil and plant samples being collected from a wide range of soil textures, environmental conditions, and production systems in England. During the 2011/12 and 2012/13 seasons, 52 and 50 fields were sampled, respectively. Farms sampled in 2011/12 were retained for the following season but new fields were identified where wheat came into the rotation.
In each season, soil samples from each field were collected prior to the sowing of the crop (presowing). Soil and plant samples were collected at growth stage (GS) 21-31, GS 37-45, and GS 65-75 (Zadoks et al., 1974). At each sampling period, samples were removed from the same 1 ha area of the field, allowing changes in disease progression and pathogen populations to be observed in the same area. The area sampled was chosen away from the headland, in the mid-field area of the crop. This study used a systematic pattern of collecting soil and plants as this provides a better estimate of changes of populations over random sampling methods (Campbell & Neher, 1994

| Eyespot, sharp eyespot, and BFR
Classification of eyespot, sharp eyespot, and BFR severity was based on the key described by Scott and Hollins (1974). Roots were visually assessed for root rot symptoms and stems were visually assessed for eyespot, BFR, and sharp eyespot symptoms. Stems were assessed at GS 21-31 and GS 37-45 on all tillers, and at GS 65-75 on the main stem of each plant. All tillers were assessed at early growth stages to try to increase the detection of these diseases, which can be difficult to identify early in the season.

| DNA extraction
DNA was extracted from 250 g of soil using the method developed by Woodhall et al. (2012b). Thirty-six soil cores removed from each field were first homogenized and a representative 250 g sample was used for DNA extraction.
Probes and primers are shown in Tables S1, S2 and S3.

| Quantification of Rhizoctonia spp.
The amplification and quantification of Rhizoctonia spp. in soil or plant material were performed using qPCR (TaqMan) assays in 96-well plates using the Applied Biosystems 7500 real-time PCR system. Cycling conditions and reagent volumes are presented in Table S4. Environmental Master Mix 2.0 (Applied Biosystems) for real-time PCR assays was used to target Rhizoctonia spp. in DNA extracts originating from soil samples.

| Analysis of qPCR (TaqMan) assays
The cycle threshold (C t ) value for each reaction was assessed using the Sequence Detection Software's default threshold setting of 0.2 ΔRn (fluorescence) units. Each sample was tested in two replicates and an average C t was taken. Target DNA in soil samples was quantified by including five DNA standards on each PCR run.
The total DNA of standards was first quantified using a NanoDrop spectrophotometer and adjusted to known concentration from the appropriate culture to produce a dilution series of four 10-fold dilutions. Target DNA was then determined by linear regression on C t .
Detection limit for all pathogens in soil assays was 10 −4 pg/g of soil.

| Quantification of Oculimacula spp. and Microdochium spp.
The amplification and quantification of Oculimacula spp. and Microdochium spp. in soil or plant material was performed using qPCR assays in 96-well plates on the CFX96 Touch real-time PCR Detection System (Bio-Rad). Cycling conditions and reagent volumes are presented in Table S4.
DNA from known isolates of each species were included in each assay to make standard curves (10-10 −6 ng/μl) and target DNA was then quantified by linear regression. All qPCR assays contained negative controls of nuclease-free water. Quantification of Rhizoctonia spp., Oculimacula spp., and Microdochium spp. DNA in soil samples was expressed as picograms of DNA per gram of soil (pg/g of soil) and for plant material as picograms per nanogram of total DNA (pg/ ng). The detection limit of all pathogens in plant assays was 10 −4 pg/ ng of total DNA.

| Meteorological data
Meteorological records for mean daily maximum (Temp max ), minimum (Temp min ) temperature (°C), total rainfall (mm), and mean daily relative humidity (%) were obtained from the ECMWF (European Centre for Medium-Range Weather Forecasts). The ECMWF uses a numerical model system combined with observational data for reanalysis of the past weather. Therefore, the meteorological data presented are estimates based on the ECMWF model output for the time period of this study. The area data covers a grid with a spatial resolution of 0.25° latitude/longitude (grid cell size c.27.5 km). The latitude/longitude of each field was obtained using global positioning system (GPS) coordinates of the area sampled within each field using Google Earth (Google Earth, 2011). Field coordinates were matched to the nearest grid cell in the ECMWF model to obtain records for that particular field. Records of the meteorological factors for each field were compiled from the date the first samples were collected (presowing) until the date the last samples were collected (GS 65-75) in each survey. Table S5 presents the means for each meteorological variable. Data from the meteorological records were split into three intervals based on the period of time between sampling of each field. These intervals were linked to the seasons during the period of study. Therefore, the interval between presowing to GS 21-31 is the autumn/winter period, the interval between GS 21-31 and GS 37-45 is spring, and between GS 37-45 and GS 65-75 is summer.  Table S6. PC1 and PC2 captured most of the variation and were used to visualize the data.

| Incidence and severity of root rot and stem base diseases
Disease incidence as a percentage of fields sampled and as a percentage of plants sampled per field and disease index (DI) for BFR, eyespot, sharp eyespot, and root rot are shown in Table 1. Root rot was present in >90% of crops at GS 21-31, declining as plants matured in both seasons. BFR incidence ranged between 92% and100% of crops and was the dominant SBD apart from GS 37-45 in 2011/12, when eyespot predominated. Eyespot was more prevalent in 2011/12, occurring in 100% and 98% of crops at GS 37-45 and GS 65-75, respectively. Sharp eyespot also occurred more frequently in 2011/12 than in 2012/13, and was identified in 87% and 90% of crops at GS 37-45 and GS 65-75, respectively.
Disease incidence as a percentage of plants sampled per field identified root rot on around 27% of plants at GS 21-31 (Table 1).
Overall, BFR occurred most commonly, found on 24%-46% of plants across all growth stages. Eyespot was highly prevalent in 2011/12 TA B L E 1 Mean disease incidence (I, %) and disease index (DI) for root rot, eyespot, brown foot rot, and sharp eyespot on English winter wheat fields and plants collected at growth stage (GS) 21-31, GS 37-45, and GS 65-75

| Pathogen quantification in stem bases and upper stems
Oculimacula spp. and Microdochium spp. were the most frequently detected species, present in >90% of samples at each growth stage (

| Correlation coefficients (r) between root rot, SBDs, and meteorological variables
Meteorological data were split into three intervals based on the period of time between sampling of each field. These intervals were linked to the seasons during the period of study (Table 6).
Meteorological data showed that throughout 2011/12 the mean Temp max and Temp min were higher than in 2012/13 (Table S5). Mean total rainfall in 2012/13 over the autumn/winter period doubled that in 2011/12 (Table 6). In contrast, in 2011/12 rainfall in the spring and   Root rot was weakly but positively correlated with each meteorological variable in the spring ( Table 6). The strongest correlation was with relative humidity (RH) (r = .50). Eyespot DI at GS 21-31 was negatively associated with rainfall and RH over the autumn/winter period. The stronger positive correlations (r > .55) were with each meteorological variable in the spring and lower temperature and rainfall in the summer. BFR was only correlated with variables in the autumn/winter period, with positive correlations with rainfall and humidity and negative associations with Temp max and Temp min (Table 6). Sharp eyespot at GS 21-31 had a weak but positive correlation with rainfall over the autumn/winter period, whilst in the summer period sharp eyespot had strong correlations with rainfall (r = .60), relative humidity (r = .47), and Temp max (r = −.50).

| PCA for soil, roots, stem bases and stems
Associations at GS 21-31 between species and diseases with meteorological variables in the autumn/winter period are shown as PCA biplots in Figure 1 for soil (a), roots (b), and stem bases (c). PC1 and PC2 explained 59% of the variation in the soil biplot, and vectors for   with any of the pathogens. However, again the short vector for root rot suggested this factor had a minor effect on the variation in this data set. Oculimacula spp. and R. cerealis were again grouped together, suggesting an association between these species in roots.
The biplot for stem bases (Figure 3c; PC1 and PC2 = 46%) showed eyespot closely aligned to R. cerealis rather than to Oculimacula spp., which grouped with Microdochium spp. and Temp min . Sharp eyespot was grouped with total rainfall and relative humidity but was sharp eyespot and eyespot were clustered along the x axis along with rainfall, but only weakly associated with their causal agents.

| D ISCUSS I ON
This is the first study to quantify the dynamics of a range of fungal species and diseases in soil, root, and stem fractions of English wheat crops and as such provides novel information on the incidence of economically important wheat pathogens, and the risk of their associated diseases. The novel finding of this study is that the predominant Rhizoctonia spp. in soil of English wheat crops was R.
solani AG 2-1, occurring on average in 63% of fields. This is in contrast to a previous soil survey by Goll et al. (2014) using a soil baiting method, which isolated AG 2-1 in just 13% of soil samples (n = 60) from arable fields in the UK. Here, we used species-specific qPCR assays to acquire more accurate quantification of targeted pathogens than soil baiting, although the soil baiting method provides useful information on the viability of pathogens within the soil profiles.
This difference in methodology could explain the higher percentage of fields containing AG 2-1 in our study. The widespread distribution of AG2-1 globally has been demonstrated by previous studies in wheat-growing regions of the USA (Schroeder et al., 2011) and in potato crops in south-eastern Australia (Sparrow et al., 2015). Less is known of the aggressiveness of UK AG 2-1 isolates to wheat in England, but the decline of the pathogen in soils over the seasons suggests the exhaustion or the lack of suitable substrate to sustain its continued accumulation. We found low DNA concentrations of AG2-1 in wheat roots/stems, and together with the lack of significant associations with any assessed disease, this suggests that winter wheat is unlikely to be a major host for this pathogen. Indeed, previous studies by Sturrock et al. (2015) have shown that AG2-1 is less pathogenic to wheat than to oil-seed rape (OSR), thus it is more likely that wheat serves as an alternative host for survival of AG2-1 until the more susceptible OSR crop is planted in the rotation. AG2-1 is highly pathogenic to OSR (Babiker et al., 2013) with much more significant implications for yield loss. least detected species in terms of incidence or biomass. However, the incidence of M. nivale from presowing over the autumn/winter period increased by >100% in both seasons. This suggests that Microdochium spp. may have been introduced into fields on infected seed, which is generally accepted as the most important source of inoculum for these pathogens (Parry et al., 1995). PCA showed that all pathogens except M. majus were generally positively grouped together in the soil, suggesting coexistence. The population densities in recovered roots showed that the most commonly occurring species was O. yallundae (90%) rather than O. acuformis (88%). These were followed by R. cerealis (56%) and R. solani AG 2-1 (52%). The majority of wheat crops in England exhibited root rot symptoms of low to moderate severity, with a peak at GS 21-31, followed by a decline as plants matured.  (Turner et al., 2002). Our studies indicate that the most recent shift is in favour of O. yallundae, as this species was found in higher DNA concentration in stems and roots throughout the growing seasons.
This was most notable at GS 65-75 in 2012/13 when DNA concentrations of O. yallundae were 24-fold higher than of O. acuformis.
The seasonal incidence and severity of eyespot and sharp eyespot across sites were higher in 2011/12 compared to 2012/13 and both correlated at GS 37-45 with higher temperature and rainfall in the spring. As Oculimacula spp. grouped separate from R. cerealis at this stage, it is likely that other factors not included in our analysis may have affected their interactions. Rainfall (McCartney & Fitt, 1998) and temperature (Bock et al., 2009)  Our results also show that visual disease assessments of stems do not agree strongly with individual pathogen DNA in planta. For example, at GS 37-45 BFR and sharp eyespot were grouped with Microdochium spp. whilst eyespot was associated with R. cerealis.
Mixed infections in the SBD complex are known to be difficult to distinguish at early crop growth stages (Turner et al., 1999(Turner et al., , 2001).
However, in our studies eyespot and sharp eyespot were more often misdiagnosed with their causal organisms than BFR. It is likely that this is because in time eyespot is an intermediary disease that occurs after BFR and prior to sharp eyespot. Thus, eyespot symptoms cor-

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.