Disease yield loss relationships

Although the infected leaf area was significantly reduced through a fungicide application in 2013 and 2014 for Kabatiella eyespot, as well as in 2013 for Turcicum leaf blight and common rust, this did not convert into a significant benefit in yield (in terms of biomass or grain production). Wise and Mueller (2011) maintain that the efficacy of a fungicide is dependent on the severity of the disease that is present in the field, with a low disease severity resulting in inconsistent yield responses. Nevertheless, this unpredictable correlation

169 has also been reported in some cases for high disease pressure. Bradley and Ames (2010), for example, found significant differences between treated and non-treated plots in terms of disease severity but without significant differences in yield for grey leaf spot (Cercospora zeae-maydis). The authors concluded that this could have been due to the high disease severity which was also present in the treated plots due to a late fungicide application.

In the trials conducted in Mittich and Ostenfeld, the lack of a positive yield response to a fungicide application may instead be related to the late occurrence of the disease. A noticeable increase in the infected leaf area only occurred when the plant was in the ripening stage and grain and biomass yield had almost formed (BBCH 85). This infers that although a statistically significant reduction in the infected leaf area was recorded at the end of September, this is not necessarily a driving factor or representative for yield losses. Despite a significant impact on the infected leaf area, Jørgensen et al. (2015) also reported no significant yield increases from fungicide treatments due to the minor severity and late arrival of the diseases in maize fields in Denmark. Thus, at late stages of plant development, the outbreak of the disease has to be severe enough to have a significant impact on yield. This is also supported by Urban (2012), who found that the outbreak of Turcicum leaf blight which occurred in the late season (BBCH 85-89) in 2010 led to significant losses in grain yield. For Kabatiella eyespot, the level of infected leaf area recorded at the end of September (≤6%) in our trials in 2013 and 2014 were recorded by Urban (2012) one month earlier in 2011, with a noticeable increase in the infected leaf area in September resulting in significant yield losses.

Another explanation for the results of our trials could be closely related to the resistance of the respective hybrid used. Hybrids with moderate resistance against Turcicum leaf blight, as was the case for the variety NK Silotop in 2014, can counteract the impact of the disease to a greater extent. Although higher levels of disease severity were registered in the untreated control compared to the treated plots, yield was not significantly affected. This explanation is less plausible for Inzing in 2013 as the sown variety, Zidane, is classified as susceptible. In this case, however, the low levels of disease are unlikely to have been sufficient to have a significant negative effect on yield, even with a susceptible hybrid. This reasoning also applies to the tested varieties in Ostenfeld in 2013 and 2014, Ronaldinio and Kalvin, which are considered moderately susceptible to Kabatiella eyespot (levels of resistance in commercial hybrids are not published). Significant yield losses were not observed here either. Despite the inconsistency of these results, likely owing to the aforementioned low disease severity, the importance of hybrids and variation in their susceptibility to leaf diseases should not be underestimated. Several authors report significant differences in performance among hybrids with respect to resistance levels for Turcicum leaf blight (Guiomar 2011; Khot et al. 2006). Furthermore, information about the susceptibility of the regional varieties sown is frequently provided by the regional body responsible for plant

170 protection and private companies (Hiltbrunner et al. 2015). For Kabatiella eyespot, significant differences in disease severity have also been reported among different maize hybrids (Prończuk 2004; Formento et al. 2014).

An additional reason for the inconsistencies between infected leaf area and yield response in our trials is related to the high variation in the data among repetitions of the same treatment.

Although repetitions were analysed separately in the linear regression, the high variation in the data did not allow a significant negative relationship between yield and infected leaf area to be established in three of the four field trials. The trial in which a significant negative correlation was obtained (Inzing 2014) presented relatively harmonious results for repetitions of the same treatment, with lower standard deviations than the other trials in general. Thus, despite the fact that disease pressure was lower in 2014, a negative correlation between yield and infected leaf area could be determined due to the consistency of the data. At the same time, the high data variation could be due to external factors. In a field trial, these can include the soil, water supply and fertilisation, factors which can never be completely uniform (Schuster and Geidel 1978). In addition, for the trials in Ostenfeld, the distribution of Kabatiella eyespot in the field was inconsistent. This was especially evident in 2014, where certain areas of the trial were more infected than others. This was caused by external factors unrelated to the effect of the fungicide.

For example, higher levels of humidity may have been concentrated in certain zones of the trial due to its north-south orientation. Due to the height of the maize plant, prolonged radiation (from higher exposure to sunlight) in the first blocks of repetitions (southern end of the trial) would have caused a lower level of humidity. As a result, the plants dried out at a quicker rate than in other areas of the trial. In these blocks, the severity of the disease was noticeably lower. Thus, the orientation of field trials could be an important factor.

Finally, a potential, so-called “greening effect” could also have had an influence on the correlation. In addition to the control of pathogens through inactivation, fungicides from the strobilurine and triazole groups have secondary effects, stimulating physiological activity in the plants to which they are applied. The “greening effect” refers to an increase in photosynthetic activity through the higher production of chlorophyll, retarding the senescence of the plant and producing a higher yield (Gerhard & Habermeyer 1998; Bryson et al. 2000;

Venancio et al. 2003; Häuser-Hahn et al. 2004). While an inhibition of the ethylene biosynthesis is reported for triazoles, which leads to a delayed senescence (Siefer &

Grossmann 1996), strobilurins improve nitrogen metabolism, maintaining the green leaf area for longer (Häuser-Hahn et al. 2004).

A potential example of the above phenomenon is the trial in Ostenfeld in 2013. Here, two treatments applied at the vegetative stage (BBCH 55), namely fluopyram + prothioconazole

171 and carbendazim + fluxilazole, provided a better control against Kabatiella eyespot than propiconazole + azoxystrobin applied at flowering (BBCH 63) in terms of infected leaf area, yet propiconazole + azoxystrobin provided a yield increase which was 6 dt/ha higher, on average. A delay in the ripening could be observed in Inzing in 2014, where plots sprayed with propiconazole + azoxystrobin applied at flowering (BBCH 65) were notably greener than the other plots (Figure 79). Nonetheless, the infected leaf area recorded for all plots (non-treated and (non-treated) was similar. This result could not be methodically confirmed because quantitative data on the green leaf area (green leaf area rating) were not registered.

Fig. 79. Mittich 2013. Example of greening effect observed between non-treated control and treated with propiconazole + azoxystrobin applied at flowering (BBCH 63).