VII Testing different Septoria models in field trials

Transcription

1 Applied Crop Protection 2017 VII Testing different Septoria models in field trials Lise Nistrup Jørgensen, Annemarie Fejer Justesen, Thies Marten Heick, Niels Matzen & Birgitte Boyer Frederiksen As part of a project from Miljøstyrelsen (Danish EPA), Septoria models (a simulation model and a humidity model) have been tested for two seasons. Recommendations given from the models have been inferior or similar to the old Crop Protection Online model. The new models need further adjustments including adjustments for cultivar resistance and can be considered more as a risk assessment than models providing specific spray recommendations. Spore trapping during two seasons have shown a constant low level of ascospore release during the season from 3 sites with peaks during ripening and harvest. This general release during the season adds to the major adaptability of the fungi to new cultivars and fungicides. The disease development has been monitored using both visual assessments and qpcr biomass assessments of leaf samples. The qpcr method measured latent infections in 22% of the samples with 0% attack. In the project new models for control of Septoria are being developed and tested. The decision support system Crop Protection Online (CPO) has for many years been recommending treatments for control of Septoria based on the number of days with precipitation (Hansen et al., 1997). Treatments are recommended if 4 days with rain (> 1 mm) have occurred starting at 32. If the crop has been treated, the crop is seen as protected for 10 days before a new risk period is initiated. A new model based on leaf wetness and periods with high relative humidity is being investigated (Figure 1) as an alternative to the existing model along with an updated version of a more complex growth model, the SIM model (Bligaard et al., 2016). Trials were carried out at 3 sites in 2016 and 2017 in order to test the new models. In each season, the trials were located at Flakkebjerg, near Horsens (LMO) and at Holeby (Lolland). At Flakkebjerg, two cultivars were tested in The specific timings of the treatments are given in Table 1. Disease and yield data from the trials are given in Tables 2-4. In 2016, the complex growth model recommended an early treatment in April, 31-32, at all sites. In Jutland, a second treatment was recommended when Model 1 was used. The humidity model (Model 2) recommended a treatment at all 3 sites following an event with 85% relative humidity for a period of 20 hours. Finally, CPO was recommending 1 treatment in Jutland and at Flakkebjerg, but none at Holeby (Table 2). Different treatments were used in comparison with the different models, using Bell at 3 different timings. In 2017, the trial plan was changed slightly following a general trend towards more fungicide resistance (Heick et al., 2017). The dose of Bell was increased from 0.5 to 0.75 l/ha and Prosaro EC 250 was included at 32 and 55 instead of using a repetition of Bell at all 3 timings. The treatment using the complex growth model recommended 1 treatment at Horsens but 2 treatments at the other two sites. The humidity model (Model 2) recommended 2 treatments at Flakkebjerg and Holeby, but 3 treatments at Horsens. CPO released only 1 treatment at Holeby and 3 treatments at Horsens and Flakkebjerg (Table 1). The models used either Bell or Prosaro EC 250 in line with the product given in the standard treatments. In both seasons, the standard treatments using 3 timings provided the best control. Control levels varied between 65 and 75% control. Yield responses in 2016 were lower than those obtained in On ave- 77

2 rage, the triple treatments gave 4.7 and 12.1, respectively, in the two seasons. The models gave yield responses equal to the standard treatments, although it was less for Model 1 in In 2017, all models gave average yield responses between 8 and 10. In one of the trials the models gave very negative responses, which was strange as this trial had a severe attack of Septoria. The trial with the resistant cultivar Sheriff at Flakkebjerg did only respond half as much as the susceptible cultivar Hereford. Also for this cultivar there were no or very limited net yield responses. All 3 models gave recommendations varying between 1 and 3 treatments in Model 1 was adjusted between 2016 and 2017 to respond more to spore production during elongation and season. Model 2, using 20 hours with 85% RH, did in a few cases show that treatments were not triggered because only 19 hours were measured, which did not fulfil the requirements. The CPO decision support system, using days with precipitation as the indicator of treatment timing, generally gave competitive recommendations. Summary of two seasons The conclusion from this year s trials is that the humidity model and CPO recommended very similar input at two sites (Horsens and Flakkebjerg) but in Lolland, where the season was very dry, CPO did not recommend a treatment, which was correct based on yield data from this season. Model 1 probably missed the best timing at Flakkebjerg and in Lolland. But at Horsens the two timings gave a good gross yield, but a less good net yield (Table 3). In 2017, all models recommended several treatments. The models gave relatively good effects and positive yield responses (Table 5). The results also made it clear that the input should be differentiated depending on the susceptibility of the cultivars. Model 1 needs to be developed further, and an interface to run the model does not exist at present. The prototype for the humidity model has been developed and can be used when organising the data needed to run the humidity model (available through LandbrugsInfo). Table 1. Detailed dates for application in models, Flakkebjerg LMO Holeby SIM model 10 May 10 May + 7 June 4 May Humidity model 26 May 26 May 27 May CPO 24 May 26 May None 2017 Flakkebjerg LMO Holeby SIM model 11 May + 23 May 26 May 9 May + 24 May Humidity model 11 May + 8 June 6 May, 26 May + 15 June 18 May + 1 June CPO 3 May + 23 May + 8 June 6 May+ 26 May + 15 June 23 May Table 2. Detailed yield data from the 3 validation trials carried out in Treatments, l/ha Leaf 2 % Septoria Yield and increase Leaf 2 Leaf 1 1. Untreated Bell 0.5 Bell 0.5 Bell Bell Bell 0.5 Bell Bell 0.5 Bell 0.5 Bell SIM model Humidity model CPO No. of trials LSD TGW g 78

3 Table 3. Detailed yield data from the 3 validation trials carried out in Treatments, l/ha A B 55 C % Septoria Yield and increase Leaf 2 Leaf 1 75 Leaf 2 1. Untreated Prosaro 0.5 Bell Bell Bell 0.75 Prosaro Prosaro 0.5 Bell 0.75 Prosaro Prosaro 0.5 Bell 0.75 Prosaro SIM model Humidity model CPO No. of trials LSD TGW g Table 4. Detailed yield data from the 3 validation trials carried out in Yield and increase Cost Net yield Yield and increase Cost Net yield Yied and increase Cost Net yield 1. Untreated Bell 0.5 Bell 0.5 Bell Bell Bell 0.5 Bell Bell 0.5 Bell 0.5 Bell SIM model Humidity model CPO LSD NS Table 5. Treatment timings and trial plan used in the field trials Treatments Yield and yield increase, Net yield, Hereford Sheriff Average Hereford 1. Untreated Sheriff Average 3 trials 2. AB B BC ABC ABC + insect SIM model Humidity model CPO NS

4 Figure 1. Measurements from a climate station are included on a prototype platform with the aim of developing a new Septoria risk model. The project is financed by the Danish Environmental Protection Agency. The platform helps to optimise the timing of spraying against Septoria and visualise when spraying is needed or when the crop can be expected to be protected. Models aim at predicting early and severe attacks of septoria tritici blotch in order to know when to recommend spraying and avoid significant attacks developing. 80

5 Collecting spores of Septoria During the season spores were trapped in five Burkard spore traps placed at four different sites. One near Holeby Lolland, 1 near Gedsergaard (2016 only) Falster, 1 at the trial site at Horsens and 2 at Flakkebjerg 1 outside and 1 in the crop (Table 6). The traps collect airborne particles by impaction onto a sticky tape, which is fixed onto a rotating drum. Every week the tapes were changed and cut into pieces, which each represented one day. DNA from all particles on the tape sections were extracted according to the method described by Duvivier et al. (2013). For each day a QPCR test specific for Zymoseptoria tritici was run to measure the quantity of Septoria spores collected on the tape. This provides a picture of the spore concentration released during the season, which again might have an impact on the disease epidemic. Two types of spores are produced ascospores and pycnidia spores. Ascospores are windspread, while pycnidia spores are mainly splashborne. The QPCR used in this study was developed and tested by Duvivier et al. (2013). The QPCR method enables the quantification of Septoria DNA in a sample, but it cannot distinguish pycnidia spores from ascospores. However, the spore traps are placed so that the orifice of the spore trap is 1 m above ground to avoid trapping of pycnidia spores spread in water droplets. The amount of spores per day was calculated by preparing a standard curve based on a dilution series of a pycnidia spore suspension with a known concentration. A pycnidia spore contains four to eight nuclei per spores whereas an ascospore contains 2 nuclei; thus the measured numbers of spores per day may be higher if converted to ascospores. The detection threshold was estimated to approximately 20 pycnida spores (~70 ascospores) per daily tape section. Pycnidia spores. Ascospore. Table 6. Sites and periods at which spores were collected Start of collection End of collection Start of collection End of collection Flakkebjerg 1, near wheat crop 21 April 10 November 6 April 28 November Flakkebjerg 2, placed in a wheat crop 21 April 10 November 6 April 28 November LMO placed close to a wheat field 22 April 12 July 6 April 8 October Gedsergaard Gedser. Used for collection of beet pathogens but used similarly for Septoria Holeby - Lolland. Used for collection of beet pathogens but used similarly for Septoria 29 June 30 September April 21 October 6 April 3 December The analysis and graphs indicate that minor release of spores takes place during most of the season. In certain intervals major peaks of release have taken place. This is particularly seen in late July and August at all locations. At Flakkebjerg higher numbers of spores are seen in the air samples from the spore trap placed in the field compared with the one placed outside the field, indicating that pycnidia spores may also have been trapped when the spore trap is located inside the crop. However, both spore traps generally follow the same pattern of spore release. From April to harvest small peaks of spore release were 81

6 occasionally seen, which may be caused by the release of ascospores from pseudothecia developing on the wheat plants. After harvest and throughout the autumn several and generally higher peaks of spore release were detected and are most likely due to the release of ascospores from plant debris. Data from spore collection in 2016 were shown in Applied Crop Protection 2016 (Jørgensen et al., 2017), and data from 2017 are shown below (Figures 2-4). Increases in spore release have been linked to periods with wet conditions few days prior to releases (Duvivier et al., 2013). Similar negative correlations have been found to sun radiation and high temperatures. In agreement with other investigations small releases of spores take place throughout the season. Peaks of spores occurring mainly in the autumn are known to be of major importance for the carryover effect from one season to the next (Duvivier et al., 2013). Photos of a Burkard 7-day volumetric spore trap located in the field. The trap is linked to a vacuum pump, and airborne particles are impacted onto a sticky tape, which rotates at a speed that is equivalent to 1 week. 82

7 50% Septoria assessed at 75 is considered to clearly reduce yield. Figure 2. Spores collected by a spore trap at Flakkebjerg during the growing season The spore trap Flakkebjerg 1 was placed outside a wheat field and Flakkebjerg 2 in a wheat field. Samples were collected from 4 April-23 November

8 Figure 3. Spores collected by a spore trap at Horsens during the growing season The spore trap was placed in a wheat field. Samples were not collected June 2017 and 11 July-7 August Figure 4. Spores collected by a spore trap at Holeby during the growing season The spore trap was placed 20 m from a wheat field at Sofiehøj, Holeby, and was moved to a sugar beet field about 500 m away. The distance to wheat fields was approximately 50 m in three directions. 84

9 Quantification of Septoria (Zymoseptoria tritici) DNA in wheat leaves 2017 Leaves were collected at regular intervals throughout the season and divided into groups of leaves with visible attack and leaves without visible attack. At the sampling time, growth stage and the level of Septoria were assessed on each of the leaf layers. DNA was extracted from the leaf samples, and the level of Zymoseptoria tritici DNA was measured by the QPCR method described by Bearchell et al. (2005). A clear gradient across the canopy indicates a higher level of attack on the lower leaves than on the upper leaves. Data from 2017 are shown in Figures 5-7. Similar data from 2016 are shown in Applied Crop Protection 2016 (Jørgensen et al., 2017). The DNA analysis gave 18 cases out of 90 of pre-symptomatic readings in 2016 and 28 cases out of 111 in 2017 where no visible attacks were seen on the leaves, indicating that the DNA method can detect latent attack. Generally, a good link between disease severity and DNA measurement was seen as shown in Figure 8 from specific cultivars and localities. In a few cases for the late growth stages only moderate DNA content was seen despite assessments showing severe attack. Part of this poor correlation might be due to the leaves being very dry and senescent for this very late assessment. Summary of two seasons testing of leaf samples In 10 crops during two seasons leaf samples were picked at approximately 10-day intervals from three different cultivars located at 3 sites (Flakkebjerg, Horsens and Holeby). The leaves were sectionalised in different leaf layers, % visible attack assessed on each leaf layer and following this DNA was extracted and measured. The samples showed clearly how the disease makes progress in the crop increasing % attack step by step as new leaves develop. In approximately 22% of the samples categorised with 0% attack, minor levels of DNA were detected. Generally, Hereford was seen as the cultivar with most severe attack, and here the highest level of DNA was measured. The correlations between measured DNA and % attack of Septoria were only moderate (R2 = ). Analysis of air samples collected April-November in two seasons and at 3-4 sites showed that Z. tritici spores are present throughout the period with small peaks of released spores in the spring (April-June) and larger peaks in the autumn/winter. Major peaks are seen in late July/August. 85

10 Figure 5. Link between DNA and % attack of Septoria in Hereford and Sheriff. Data from Flakkebjerg with moderate levels of diseases. Low = lower leaves with attack; top = top leaves without attack. The numbers 1, 2, 3, 4, 5 and 6 refer to the leaf layer, with 1 indicating the top leaf layer at each growth stage and 6 was the lowest leaf layer assessed. 86

11 Figure 6. Link between DNA and % attack of Septoria in Hereford and Sheriff. Data from Horsens (LMO field) with moderate to high levels of diseases. Low = lower leaves with attack; top = top leaves without attack. The numbers 1, 2, 3, 4, 5 and 6 refer to the leaf layer, with 1 indicating the top leaf layer at each growth stage and 6 was the lowest leaf layer assessed. 87

12 Figure 7. Link between DNA and % attack of Septoria in Torp. Data from Holeby with low to moderate levels of diseases. Low = lower leaves with attack; top = top leaves without attack. The numbers 1, 2, 3, 4, 5 and 6 refer to the leaf layer, with 1 indicating the top leaf layer at each growth stage and 6 was the lowest leaf layer assessed. Figure 8. Link between measured DNA and % attack of Septoria in the two cultivars Sheriff and Hereford. References Bearchell, S, B. A. Fraaije, M. W. Shaw and B. D. L. Fitt (2005). Wheat archive links long-term fungal pathogen population dynamics to air pollution. PNAS 102: Bligaard, J., L. N. Jørgensen, J. E. Ørum, G. C. Nielsen, J. G. Hansen and J. Axelsen (2016). Septoriamodel med vejrdata versus en Septoria Timer i afgrøden. Plantekongres 2016, pp Duvivier, M, G. Dedeurwaerder, M. De Proft, J.-M. Moreau and A. Legrève (2013). Real-time PCR quantification and spatio-temporal distribution of airborne inoculum of Mycosphaerella graminicola in Belgium. European Journal of Plant Pathology 137: Hansen, J. G., B. J. M. Secher, L. N. Jørgensen and B. Welling (1994). Threshold for control of Septoria spp. in winter wheat based on precipitation and growth stage. Plant Pathology 43: Heick, T. M., A. F. Justesen and L. N. Jørgensen (2017). Anti-resistance strategies for fungicides against wheat pathogen Zymoseptoria tritici with focus on DMI fungicides. Crop Protection 99: Jørgensen, L. N., A. F. Justesen, T. Heick, N. Matzen and B. B. Frederiksen (2017). Testing different Septoria models (MS project). In: L. N. Jørgensen, B. J. Nielsen, P. K. Jensen, S. K. Mathiassen, S. Sørensen & T. Heick (eds.) Applied Crop Protection DCA report No Aarhus University. pp