Time point of virus yellows infection is crucial for yield losses in sugar beet, and co-infection with beet mosaic virus is negligible under field conditions
Abstract
Beet chlorosis virus (BChV), beet mild yellowing virus (BMYV) and beet yellows virus (BYV) transmitted by Myzus persicae cause virus yellows (VY) disease in sugar beet. M. persicae also transmits beet mosaic virus (BtMV), which is often associated with VY. So far, field trials to determine the effect of infection time point on yield have used 100% inoculation density and little is known about the yield effect of BtMV in mixed infections with VY species under field conditions. Therefore, we conducted sugar beet field trials using a new inoculation protocol with densities of 3%–10% in combination with different infection time points; we also tested the effect of BtMV/VY species mixed infections on white sugar yield (WSY). We observed a wide range of WSY losses for BChV (3.6%–26.8%), BMYV (1.7%–22.0%) and BYV (3.7%–37.0%) depending on infection time point, with no further significant losses after BBCH Stages 18/19, 35 and 39, respectively. Both the time of infection and area under disease progress curve showed excellent correlation with WSY losses for all VY species. BtMV had no significant effect on WSY losses either as a single infection or in mixed infections with BChV or BYV compared to control or single infections of these viruses. However, BMYV/BtMV mixed infection showed significantly increased WSY loss (+13.6%) compared to single BMYV infection. Our results can be used to predict yield losses in practical fields and to develop economic control thresholds for decision support systems.
1 INTRODUCTION
Sugar beet (Beta vulgaris subsp. vulgaris) is cultivated on a global area of 4.5 million hectares, and it accounts for 20% of the world's sugar production (Food and Agriculture Organization of the United Nations, 2024). In the European Union (EU), sugar beet is grown on 1.4 million hectares, producing an annual sugar yield of 16.2 million tonnes. The largest sugar beet producers in the EU are Germany, France and Poland (European Association of Sugar Manufacturers, 2023). Since neonicotinoid seed treatment was banned in 2018 in the EU, sugar beet growers have again been confronted with the virus disease complex virus yellows (VY), which is transmitted by aphids and causes yellowing of sugar beet leaves and significant yield losses (Hossain et al., 2021). VY is caused by three virus species: beet chlorosis virus (BChV), beet mild yellowing virus (BMYV), both belonging to the family Solemoviridae, genus Polerovirus and beet yellows virus (BYV, family Closteroviridae, genus Closterovirus). While BMYV and BChV cause a yellowish-orange chlorosis on leaves (for BChV, particularly interveinal chlorosis), which can lead to premature leaf death, BYV causes a yellowish discolouration that can lead to reddish necrosis on older leaves (de Koeijer & van der Werf, 1999; Stevens, Freeman, et al., 2005). Beet mosaic virus (BtMV, family Potyviridae, genus Potyvirus) is often associated with VY, as it is also aphid transmitted, but causes mosaic-like patches on leaves rather than the typical yellowing symptoms (Wintermantel, 2005). The main vector for all mentioned virus species is the green peach aphid Myzus persicae. While BMYV and BChV are transmitted persistently, BYV is transmitted semipersistently and BtMV nonpersistently (Dusi & Peters, 1999; Kozłowska-Makulska et al., 2009; Limburg et al., 1997). Results from the early 2000s showed that BMYV was mainly present in western and northern Europe, BChV was more common in the west and south, and BYV was found exclusively in the south (Stevens, Patron, et al., 2005). Hossain et al. (2021) conducted a Europe-wide monitoring from 2017 to 2019, during and after the authorization of neonicotinoid seed treatment and found that BYV is now also prevalent in northern regions, while BChV and BMYV were mainly found in western and northern Europe. BtMV was found only sporadically, and mixed infections with all viruses were also detected in randomly taken samples.
In the post-neonicotinoid era, the only option to control M. persicae as a vector for VY in sugar beet is broadcast application of insecticides. However, M. persicae is a high-risk organism for insecticide resistance and is already resistant to many active ingredients (Bass et al., 2014). In addition, only a few active ingredients are registered in the EU (e.g., Germany) for the control of M. persicae in sugar beet. Moreover, the EU Farm-to-Fork strategy aims to reduce the use of plant protection products by 50% by 2030 (European Commission, 2020). Breeding sugar beet varieties tolerant or resistant to VY could be an effective solution to solve the problem. However, varieties with high levels of tolerance or resistance to all VY species and high yield potential are not yet available and the breeding process will take several more years (Rollwage et al., 2024). In the short term, therefore, insecticides need to be applied more precisely as part of integrated pest management. Another challenge is that M. persicae has a long flight period in spring, and therefore, a long period of initial infection in sugar beet is possible (Bell et al., 2023; Luquet et al., 2023). In this context, the use of decision support systems (DSS) in combination with forecasting models for M. persicae and VY occurrence can be helpful. One component of DSS is the definition of economic control thresholds (Jeger et al., 2018; Jones, 2004; Nutter et al., 1993). To estimate these thresholds, infestation–loss ratios need to be developed via field trials with different inoculation time points.
Various field studies with different inoculation densities (10% or 100%) of viruliferous wingless M. persicae have been carried out over the last 35 years for BChV, BMYV and BYV in the United Kingdom and Germany (Hossain et al., 2021; Smith & Hallsworth, 1990; Stevens et al., 2004). In field trials in Germany, an inoculation density of 10% led to white sugar yield (WSY) losses of 24% for BChV, 23% for BMYV and 10% for BYV. A mixed infection of BChV and BYV resulted in a WSY loss of 43% (Hossain et al., 2021). Field trials from the United Kingdom with an inoculation density of 100% showed sugar yield losses of 8%–26% for BChV, 5%–29% for BMYV and up to 47% for BYV, depending on the inoculation time point (Smith & Hallsworth, 1990; Stevens et al., 2004). Both Stevens et al. (2004) and Smith and Hallsworth (1990) showed a negative relationship between the infection time point of BChV, BMYV and BYV and the relative loss in sugar yield.
In contrast, there are only a few studies on the yield effect of BtMV under field conditions. A single infection (more than 90% infection rate) with BtMV is estimated to reduce sugar beet yield by less than 10% (Shepherd et al., 1964). Shepherd et al. (1964) also reported that co-infection of BtMV and BYV resulted in slightly higher yield losses compared to single infection of BYV under field conditions. Synergistic effects for co-infection of BtMV with BYV or beet western yellowing virus (BWYV, family Solemoviridae, genus Polerovirus), a close relative of BMYV and BChV, were reported by Wintermantel (2005). In a controlled environment, co-infection of BtMV with BYV or BWYV resulted in stunted growth and significantly reduced plant biomass compared to single infection of each virus. Co-infection also resulted in more severe symptoms for BWYV and a higher virus titre than single infection of each virus. Synergistic effects of co-infection of potyviruses and members of the Closteroviridae or Solemoviridae family have also been described for other crops (Barker, 1989; Karyeija et al., 2000).
As described by Hossain et al. (2021), low inoculation densities are required to detect differences between treatments (like cultivars or insecticides) in field trials, as high inoculation densities do not reflect natural infestation of M. persicae in practical fields. The recently updated European and Mediterranean Plant Protection Organization (EPPO) protocol for trials with aphids as vectors of viruses in sugar beet also recommends a low inoculation density of 3%–10% (European and Mediterranean Plant Protection Organization [EPPO], 2023). So far, the relationship between area under disease progress curve (AUDPC) and yield losses in sugar beet has not been described. In addition, infestation–loss ratios can assist in the selection of breeding material for resistance breeding. In the case of BtMV infection, it is essential to know the overall yield effect of single infection or of mixed infection with other VY species under field conditions in order to focus breeding efforts or to develop infestation–loss ratios. As there are no field studies on the influence of inoculation time points of BChV, BMYV and BYV at low inoculation densities and no data on the effect of BtMV single and mixed infections on disease development and sugar beet yield, we conducted field inoculation studies on these two topics in central Germany in 2023. In the first part of our study, we investigated the effect of low inoculation densities (3%–10%) and inoculation time points of BChV, BMYV and BYV on (i) the development of disease symptoms, (ii) AUDPC and (iii) yield and quality parameters of sugar beet. In the second part of our study, we investigated the effect of BtMV single infection and mixed infection with BChV, BMYV or BYV on (i) the development of disease symptoms, (ii) AUDPC and (iii) yield and quality parameters of sugar beet.
2 MATERIALS AND METHODS
2.1 Field sites and experimental design
The field trials with varied inoculation time points were conducted at two sites in 2023 (Harste and Sieboldshausen), and the BtMV mixed infection trial was conducted at one site (Harste). Both sites are in central Germany in the river Leine valley near Göttingen. The long-term mean annual precipitation for Sieboldshausen is 702 mm and for Harste 624 mm (1991–2020). The long-term average annual temperature for both sites is 9.4°C (Deutscher Wetterdienst, 2024).
Both trials were set up in a complete randomized block design with four replications. The VY-susceptible cultivar ZR 2313 (Bundessortenamt, 2020) was sown in all trials with a row distance of 0.45 m and a sowing density of 120,000 plants/ha. Sugar beet seeds were treated only with fungicides (7 g/unit penthiopyrad, 14 g/unit hymexazol). The plot size was 43.2 m2 (12 rows × 8 m) for the inoculation time-point trials and 64.8 m2 (18 rows × 8 m) for the BtMV mixed infection trial. In the inoculation time-point trial, a noninoculated control and five inoculation time points between BBCH Stages 10 and 39 for BChV, BMYV and BYV were tested. The mixed infection trial included a noninoculated control, BtMV, BChV, BMYV and BYV single infections and a combination of BtMV with each of the other three virus species, inoculated at BBCH 14/16 and 18/19, respectively (Table 1). In the mixed infection trial, each treatment was in each block, while the inoculation time-point trial was divided into three subtrials with a complete randomized block design for each virus to minimize the risk of virus spread and contamination between the different virus species. The subtrials were separated by 24 border rows of cultivar ZR 2313 and additionally treated with insecticide (10 g/unit tefluthrin). The sowing date was 4 May 2023 in Harste and 27 April 2023 in Sieboldshausen. In all trials, no foliar insecticides were applied, and all other crop management measures followed the official recommendations of the regional advisory service.
| Trial | Location | Virus | Treatment | Inoculation method |
|---|---|---|---|---|
| Inoculation time point | Sieboldshausen, Harste | BMYV | Control, no inoculation | None |
| Inoculation BBCH 10/12 | 3% density aphid inoculation | |||
| Inoculation BBCH 14/16 | ||||
| Inoculation BBCH 18/19 | ||||
| Inoculation BBCH 35 | ||||
| Inoculation BBCH 39 | ||||
| BChV | Control, no inoculation | None | ||
| Inoculation BBCH 10/12 | 3% density aphid inoculation | |||
| Inoculation BBCH 14/16 | ||||
| Inoculation BBCH 18/19 | ||||
| Inoculation BBCH 35 | ||||
| Inoculation BBCH 39 | ||||
| BYV | Control, no inoculation | None | ||
| Inoculation BBCH 10/12 | 10% density aphid inoculation | |||
| Inoculation BBCH 14/16 | ||||
| Inoculation BBCH 18/19 | ||||
| Inoculation BBCH 35 | ||||
| Inoculation BBCH 39 | ||||
| Mixed infection | Harste | Control | No inoculation | None |
| BtMV | Single infection | 30% density mechanical inoculation | ||
| BChV | Single infection | 100% density aphid inoculation | ||
| BMYV | ||||
| BYV | ||||
| BChV + BtMV | Mixed infection | 30% density mechanical and 100% density aphid inoculation | ||
| BMYV + BtMV | ||||
| BYV + BtMV |
2.2 Aphid rearing, inoculation densities and methods
For all field trials, M. persicae were obtained from the aphid rearing unit at the Institute of Sugar Beet Research (IfZ), Göttingen, Germany. The aphids were maintained on noninfected sugar beet plants in tents in the greenhouse under controlled conditions (24.0°C, humidity 75%–80%, photoperiod of 16/8 h light/dark). To produce viruliferous wingless M. persicae, adult aphids were placed on virus-infected sugar beet plants to feed for at least 72 h. The four different virus species (BChV, BMYV, BYV and BtMV) were maintained in separate sugar beet plants.
In the inoculation time-point trial, rows 7 to 9 of the plot were inoculated with viruliferous wingless M. persicae. The inoculation density was 3% for BChV and BMYV and 10% for BYV. The inoculation density of 10% for BYV was chosen because, based on the semipersistent transmission mechanism and our experience from previous trails (results not published), it can be assumed that not all aphids carry virus and therefore inoculation at low densities is not reliable. The plants to be inoculated were evenly distributed over the three rows of the plot and 10 viruliferous (BChV, BMYV or BYV) wingless M. persicae were placed on each plant using a brush. Each virus species was inoculated by one person on each date to avoid contamination between virus species. No aphids were released in the untreated controls. The specific inoculation date per site, virus and treatment can be found in Table S1.
Based on our experience from field trials in 2021 and 2022 (results not published), where we achieved very low infection rates in sugar beet with BtMV-infected viruliferous aphids (100% inoculation density), we decided to mechanically inoculate BtMV in 2023. Therefore, in the BtMV mixed infection trial, the virus was mechanically inoculated on 30% of the plants in rows 5 to 7 and BChV, BMYV or BYV were superinoculated 10 days later with 10 viruliferous aphids on 100% of the plants in rows 5 to 7. The same inoculation density was chosen for the plots with single infection with BChV, BMYV or BYV. The inoculation density of 100% was chosen to infect all plants simultaneously. For the mechanical inoculation, BtMV containing plant sap was prepared by grinding leaves of BtMV-infected sugar beet plants from the greenhouse 1:10 (wt/vol) in 10 mM phosphate buffer on 5 June 2023. On the same day, two Celite 535 pretreated leaves per plant (BBCH Stage 14/16) were mechanically inoculated using a cotton swab dipped in plant sap in rows 5 to 7 of the plots. Subsequently, at BBCH Stage 18/19, the viruliferous aphids for BChV, BMYV and BYV were also released in rows 5 to 7. The single BtMV inoculation was made in the same way, but the plants received additional non-viruliferous wingless M. persicae. Each single or mixed infection treatment was inoculated by one person to prevent contamination.
2.3 Disease ratings and area under disease progress curve calculation
2.4 Harvest and quality analysis
The sugar beets from the three-row core plots (10.8 m2) were harvested on 18 October in Harste and on 23 October 2023 in Sieboldshausen using a one-row beet harvester. The harvested and defoliated roots were stored in sacks overnight and processed the following day at IfZ. They were automatically washed, weighed and processed into beet brei. Subsamples of the brei were quick-frozen (−60°C), stored at −20°C and analysed for sugar, potassium, sodium and α-amino nitrogen according to the standard methods of the sugar industry (International Commission for Uniform Methods of Sugar Analysis, 2007). White sugar yield (WSY) in t/ha was calculated according to Märländer et al. (2003).
2.5 Statistical analysis
Statistical data analysis was performed using the software R (v. 4.2.3; R Foundation for Statistical Computing). A linear model was used to analyse differences between the inoculation time points per virus species (inoculation time-point trial) and the effect of the different single and mixed infections (BtMV mixed infection trial) on percentage symptomatic plants, AUDPC, root yield, sugar content and WSY. For the inoculation time-point trial, each virus species was analysed separately. For single sites, the fixed effect of treatment and block was tested. In the inoculation time-point trial, the fixed effects of site, inoculation time point, block and the interaction of site and inoculation time point were additionally tested. The residuals of the linear models were checked graphically for normal distribution and homoscedasticity (Kozak & Piepho, 2017), and additionally for normal distribution using the Shapiro–Wilk test. The package ‘agricolae’ was used for calculating the analysis of variance (ANOVA). If the inoculation time point or the single and mixed infections were significant in the ANOVA (p < 0.05), their mean values were compared with a post hoc t test using the ‘emmeans’ package. A linear regression model was used to calculate the relationship between WSY and AUDPC or percentage of infected plants at harvest for both trials. For the inoculation time-point trial, the relationship between WSY loss and AUDPC or inoculation time point (days after sowing) was additionally analysed. To determine the WSY losses, the mean WSY per inoculation time point was calculated for each virus species and site and normalized to the mean of the noninoculated control per virus species and site. All figures were created with the package ‘ggplot2’ in R.
3 RESULTS
3.1 Inoculation time-point trial
3.1.1 Disease epidemiology
The first yellowing symptoms, in a range of 3%–13% symptomatic plants per plot, were detected in mid-June at both sites in the plots of the first two inoculation time points (BBCH Stages 10/12 and 14/16) of BYV (Figure 1), while in the BChV and BMYV subtrials no to very few symptomatic plants (0.2%–1.5%) were visible for these two time points. These treatments reached about 75%–100% symptomatic plants at both sites by mid-August. With later inoculation, virus spread was much slower. In October, the difference between the second and third inoculation time point for BChV and BMYV was 51% and 62%, respectively, in Harste and 63% and 73%, respectively, in Sieboldshausen. For BYV, the percentage of symptomatic plants increased from 5% to 10% symptomatic plants in June to almost 100% in July for the first two inoculation time points. Differences in the percentage of symptomatic plants between the second and third inoculation time point were 4% in Harste and 6% in Sieboldshausen and thus much lower than for BChV and BMYV. At both sites, no major differences were observed in October between the last three inoculation time points (BBCH Stages 18/19, 35 and 39) for BMYV. In contrast, the percentage of symptomatic plants decreased from the third to the fifth inoculation time point for BChV and BYV. Infection of the entire plot (100% symptomatic plants) was achieved by all virus species for the first two inoculation time points. The percentage of symptomatic plants in the noninoculated controls, due to contamination or natural influx of viruliferous M. persicae, was less than 15% in October at both sites and in all virus species, with values for BYV being highest.
In the ANOVA (Table S2), no significant interaction was found between site and inoculation time point for the AUDPC (%-days) for each virus species. Therefore, AUDPC for BChV, BMYV and BYV was calculated as the mean of both sites (Figure 2). Highest AUDPC values across all virus species were achieved for the different inoculation time points of BYV. For all virus species, the noninoculated control displayed the significantly lowest AUDPC. The AUDPC differed significantly between the first and the second inoculation time point for BChV and BMYV, but not for BYV. The greatest differences in AUDPC were observed between the second and the third inoculation time point for BChV, BMYV and BYV. While the AUDPC for all viruses was significantly highest for inoculation at BBCH Stage 10/12, AUDPC decreased significantly for later inoculations up to BBCH Stage 39 for BYV, BBCH Stage 35 for BChV and BBCH Stage 18/19 for BMYV.
3.1.2 Yield and quality parameters
In the ANOVA (Table S2), no significant interactions were found between site and inoculation time point for the yield and quality parameters for each virus species. The greatest reduction in root yield (t/ha) across all inoculation time points compared to the noninoculated control was achieved at both sites in the BYV subtrial (2.2–30.5 t/ha) compared to BChV (4.4–25.5 t/ha) and BMYV (0.9–18.6 t/ha; Table 2). The mean sugar content (%) of both sites was significantly lower for the first two inoculation time points of BChV and BYV compared to the corresponding noninoculated control. For BMYV, this was only the case for the first inoculation time point. The significantly greatest reduction in WSY (%) compared to the noninoculated control of each virus was achieved for the first two inoculation time points for all virus species (BChV 26.8% and 16.6%, BMYV 22.0% and 17.2%, BYV 37.0% and 30.4%) (Figure 3). For BChV, the mean WSY (t/ha) of both sites for the last inoculation time point was not significantly lower than in the noninoculated control, while for BMYV and BYV the last two inoculation time points had no significant effect on WSY (Table 2).
| Site | Inoculation time pointa | BChV | BMYV | BYV | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Root yield (t/ha) | Sugar content (%) | White sugar yield (t/ha) | Root yield (t/ha) | Sugar content (%) | White sugar yield (t/ha) | Root yield (t/ha) | Sugar content (%) | White sugar yield (t/ha) | ||
| Sieboldshausen | Control | 95.1 ± 2.6 CD | 17.8 ± 0.2 B | 15.3 ± 0.3 D | 99.1 ± 1.7 D | 17.1 ± 0.4 A | 15.3 ± 0.2 D | 92.3 ± 2.3 D | 17.7 ± 0.2 C | 14.8 ± 0.3 D |
| BBCH 10/12 | 78.1 ± 3.4 A | 16.9 ± 0.2 A | 11.9 ± 0.6 A | 80.5 ± 4.1 A | 16.9 ± 0.2 A | 12.2 ± 0.5 A | 61.7 ± 5.8 A | 16.4 ± 0.2 A | 9.0 ± 1.0 A | |
| BBCH 14/16 | 83.1 ± 6.5 B | 17.2 ± 0.3 A | 12.9 ± 1.1 B | 87.1 ± 1.8 B | 17.0 ± 0.3 A | 13.4 ± 0.5 B | 69.4 ± 7.6 B | 16.8 ± 0.2 B | 10.4 ± 1.3 B | |
| BBCH 18/19 | 90.6 ± 6.1 C | 17.6 ± 0.1 B | 14.4 ± 1.0 C | 92.1 ± 2.8 BC | 17.4 ± 0.1 A | 14.4 ± 0.5 C | 82.3 ± 1.5 C | 17.3 ± 0.2 C | 12.9 ± 0.3 C | |
| BBCH 35 | 90.7 ± 4.0 C | 17.8 ± 0.1 B | 14.6 ± 0.6 CD | 93.7 ± 3.6 CD | 17.5 ± 0.2 A | 14.8 ± 0.6 CD | 87.6 ± 3.2 CD | 17.5 ± 0.2 C | 13.9 ± 0.5 CD | |
| BBCH 39 | 96.1 ± 2.6 D | 17.7 ± 0.4 B | 15.4 ± 0.0 D | 96.9 ± 6.1 CD | 17.3 ± 0.7 A | 15.1 ± 0.3 D | 90.0 ± 10.8 D | 17.5 ± 0.3 C | 14.2 ± 1.9 D | |
| Harste | Control | 95.1 ± 4.7 C | 16.3 ± 0.3 B | 13.9 ± 0.4 C | 90.6 ± 5.6 C | 16.0 ± 0.4 B | 13.0 ± 1.2 C | 87.8 ± 3.1 B | 16.4 ± 0.4 B | 12.9 ± 0.8 B |
| BBCH 10/12 | 69.6 ± 2.7 A | 15.4 ± 0.3 A | 9.5 ± 0.3 A | 72.0 ± 3.5 A | 15.3 ± 0.4 A | 9.8 ± 0.3 A | 60.7 ± 3.3 A | 15.7 ± 0.2 A | 8.4 ± 0.6 A | |
| BBCH 14/16 | 82.7 ± 2.3 B | 15.6 ± 0.4 A | 11.4 ± 0.5 B | 72.4 ± 3.1 A | 15.5 ± 0.3 A | 10.0 ± 0.6 A | 63.5 ± 5.1 A | 15.7 ± 0.2 A | 8.8 ± 0.8 A | |
| BBCH 18/19 | 85.4 ± 5.9 B | 16.0 ± 0.4 B | 12.2 ± 0.9 BC | 83.2 ± 5.9 B | 16.0 ± 0.3 B | 11.9 ± 1.0 B | 81.9 ± 3.8 B | 16.4 ± 0.2 B | 12.0 ± 0.6 B | |
| BBCH 35 | 87.6 ± 6.6 BC | 16.2 ± 0.1 B | 12.7 ± 1.0 BC | 89.7 ± 1.6 C | 15.9 ± 0.2 B | 12.7 ± 0.3 BC | 85.1 ± 6.4 B | 16.4 ± 0.1 B | 12.5 ± 0.9 B | |
| BBCH 39 | 87.8 ± 10.6 BC | 16.2 ± 0.2 B | 12.7 ± 1.7 BC | 88.7 ± 4.3 BC | 16.0 ± 0.3 B | 12.7 ± 0.5 BC | 85.0 ± 8.3 B | 16.4 ± 0.2 B | 12.5 ± 1.3 B | |
| Mean of sites | Control | 95.1 ± 3.5 D | 17.1 ± 0.8 C | 14.6 ± 0.8 E | 94.9 ± 5.9 C | 16.6 ± 0.7 BC | 14.1 ± 1.4 D | 90.0 ± 3.5 C | 17.1 ± 0.7 B | 13.8 ± 1.1 D |
| BBCH 10/12 | 73.8 ± 5.3 A | 16.1 ± 0.8 A | 10.7 ± 1.4 A | 76.3 ± 5.8 A | 16.1 ± 0.9 A | 11.0 ± 1.4 A | 61.2 ± 4.4 A | 16.0 ± 0.4 A | 8.7 ± 0.8 A | |
| BBCH 14/16 | 82.9 ± 4.5 B | 16.4 ± 0.9 A | 12.2 ± 1.1 B | 79.8 ± 8.2 A | 16.3 ± 0.8 AB | 11.7 ± 1.8 B | 66.4 ± 6.8 A | 16.2 ± 0.6 A | 9.6 ± 1.3 B | |
| BBCH 18/19 | 88.0 ± 6.2 C | 16.8 ± 0.9 B | 13.3 ± 1.5 C | 87.6 ± 6.4 B | 16.7 ± 0.8 C | 13.2 ± 1.5 C | 82.1 ± 2.7 B | 16.8 ± 0.5 B | 12.4 ± 0.7 C | |
| BBCH 35 | 89.1 ± 5.3 C | 17.0 ± 0.9 BC | 13.6 ± 1.3 CD | 91.7 ± 3.4 C | 16.7 ± 0.9 C | 13.8 ± 1.2 D | 86.3 ± 4.8 BC | 17.0 ± 0.6 B | 13.2 ± 1.0 CD | |
| BBCH 39 | 92.0 ± 8.4 CD | 17.0 ± 0.9 BC | 14.1 ± 1.8 DE | 92.8 ± 6.6 C | 16.7 ± 0.9 C | 13.9 ± 1.4 D | 87.5 ± 9.3 C | 16.9 ± 0.6 B | 13.3 ± 1.8 D | |
- Note: Values are means ± SD. Means with the same letter are not significantly different within virus species, parameter and site (t test, α =0.05).
- a BBCH growth stage at which plants were inoculated.
3.1.3 Correlation between disease epidemiology and yield parameters
The regression analysis showed a highly significant (R2 = 0.76–0.88; p ≤ 0.001) negative relationship between WSY and AUDPC at plot level per virus species and site (Figure 4). For the negative correlation of WSY (t/ha) and symptomatic plants at harvest (%), the coefficients of determination (0.40–0.85) were all significant (p ≤ 0.001) (data not shown). Furthermore, the regression analysis showed a strong positive relationship between relative WSY loss and AUDPC, and a strong negative correlation between relative WSY loss and inoculation time point (days after sowing) (Figure 5). The coefficients of determination (0.82–0.95) were highly significant (p ≤ 0.001) in all cases.
3.2 BtMV mixed infection trial
3.2.1 Disease epidemiology
The disease assessment (% symptomatic plants per plot) in the BtMV mixed infection trial was divided into two ratings, one for symptoms of BChV, BMYV and BYV (VY symptoms) and one for BtMV symptoms. In the first rating at the beginning of July, the highest proportion of plants with VY symptoms (>50%) was recorded for BYV single and BYV + BtMV mixed infection (Table 3). In contrast, symptoms were detected on fewer than 5% of the plants for the other treatments at this date. In August, September and October, the percentage of symptomatic plants reached highest values for the BChV and BYV single and mixed infections with BtMV and lowest values for the BMYV treatments. In the noninoculated control and in the single infection with BtMV, a few plants showed VY symptoms, due to contamination or natural influx of viruliferous M. persicae.
| Symptoms | Treatment | Date of assessment | |||
|---|---|---|---|---|---|
| July | August | September | October | ||
| VY | Control | 0.0 ± 0.0 | 4.3 ± 2.8 | 2.0 ± 2.3 | 3.3 ± 3.3 |
| BtMV | 0.0 ± 0.0 | 3.5 ± 2.9 | 1.3 ± 1.5 | 4.8 ± 4.3 | |
| BChV | 4.3 ± 1.7 | 93.8 ± 4.8 | 96.5 ± 4.4 | 100.0 ± 0.0 | |
| BChV + BtMV | 4.0 ± 1.8 | 98.8 ± 2.5 | 96.3 ± 2.5 | 100.0 ± 0.0 | |
| BMYV | 1.0 ± 1.4 | 72.5 ± 17.1 | 81.3 ± 17.5 | 78.8 ± 32.8 | |
| BMYV + BtMV | 2.0 ± 1.4 | 91.3 ± 4.8 | 83.8 ± 32.5 | 98.3 ± 2.9 | |
| BYV | 64.3 ± 32.6 | 100.0 ± 0.0 | 100.0 ± 0.0 | 100.0 ± 0.0 | |
| BYV + BtMV | 50.0 ± 36.5 | 99.5 ± 1.0 | 100.0 ± 0.0 | 100.0 ± 0.0 | |
| BtMV | Control | 0.0 ± 0.0 | 0.0 ± 0.0 | 0.0 ± 0.0 | 0.0 ± 0.0 |
| BtMV | 88.8 ± 13.2 | 100.0 ± 0.0 | 100.0 ± 0.0 | 100.0 ± 0.0 | |
| BChV | 0.0 ± 0.0 | 0.0 ± 0.0 | 0.0 ± 0.0 | 0.0 ± 0.0 | |
| BChV + BtMV | 100.0 ± 0.0 | 100.0 ± 0.0 | 100.0 ± 0.0 | 100.0 ± 0.0 | |
| BMYV | 0.0 ± 0.0 | 0.0 ± 0.0 | 25.0 ± 50.0 | 25.0 ± 50.0 | |
| BMYV + BtMV | 97.5 ± 5.0 | 100.0 ± 0.0 | 100.0 ± 0.0 | 100.0 ± 0.0 | |
| BYV | 0.0 ± 0.0 | 0.0 ± 0.0 | 0.0 ± 0.0 | 0.0 ± 0.0 | |
| BYV + BtMV | 87.5 ± 9.6 | 75.0 ± 50.0 | 100.0 ± 0.0 | 100.0 ± 0.0 | |
- Note: Values are mean ± SD.
At the beginning of July, the percentage of plants with BtMV symptoms was highest in the BChV + BtMV and BMYV + BtMV mixed infections. In October, all BtMV-inoculated treatments achieved a proportion of 100% plants with BtMV symptoms; in addition, BtMV symptoms were observed on 25% of the plants in the BMYV single infection (limited to the fourth repetition), due to contamination or natural influx.
The AUDPC values for VY in the BYV single and the BtMV + BYV mixed infections were significantly highest, followed by the BChV + BtMV mixed infection and BChV single infection (Figure 6). The differences between the single and mixed infections were not significant for either BChV or BYV. In contrast, the differences between the BMYV single infection and the BMYV + BtMV mixed infection were significant. AUDPC values between 8881.3 and 9450.0 %-days were calculated for the BtMV AUDPC in the inoculated plots, and no significant differences were found between the BtMV single and the respective mixed infections with BChV, BMYV and BYV. The AUDPC for the contamination in the BMYV single infection was significantly lowest.
3.2.2 Yield and quality parameters
The root yield (t/ha) over all experiments ranged between 95.7 t/ha, for the noninoculated control, and 55.1 t/ha, for the BYV mixed infection with BtMV (Table 4). The variation in sugar content (%) was between 16.4% for the BtMV single infection and 14.8% for the BChV + BtMV infection. WSY (t/ha) was highest in the noninoculated control and after BtMV single infection, while the differences were not significantly different.
| Treatment | Root yield (t/ha) | Sugar content (%) | White sugar yield (t/ha) |
|---|---|---|---|
| Control | 95.7 ± 6.1 E | 16.3 ± 0.1 D | 14.0 ± 0.8 D |
| BtMV | 90.2 ± 8.5 E | 16.4 ± 0.2 D | 13.2 ± 1.3 D |
| BChV | 68.8 ± 4.4 CD | 15.1 ± 0.1 ABC | 9.2 ± 0.6 B |
| BChV + BtMV | 65.4 ± 4.3 BC | 14.8 ± 0.4 A | 8.5 ± 0.8 AB |
| BMYV | 77.4 ± 6.6 D | 15.6 ± 0.6 C | 10.7 ± 1.0 C |
| BMYV + BtMV | 67.3 ± 3.5 BC | 14.9 ± 0.6 ABC | 8.8 ± 0.8 B |
| BYV | 59.7 ± 6.8 AB | 15.4 ± 0.4 BC | 8.2 ± 1.1 AB |
| BYV + BtMV | 55.1 ± 1.9 A | 15.1 ± 0.3 ABC | 7.3 ± 0.2 A |
- Note: Values are means ± SD. Means with same letter are not significantly different within parameter (t test, α = 0.05).
Loss of WSY (%) compared to the noninoculated control was highest after BYV + BtMV mixed infection (48%) and BYV single infection (42%), followed by BChV + BtMV mixed infection (39%) and BChV single infection (34%; Figure 7). The differences in WSY between the single and mixed infections were not significant within BChV and BYV, but were significant between BMYV single infection and BMYV + BtMV mixed infection (Table 4). The regression analysis for WSY (t/ha) and VY AUDPC (%-days) at plot level for all variants showed a strong negative relationship (Figure S1). The coefficient of determination was 0.85 and highly significant (p ≤ 0.001). For the negative correlation of WSY (t/ha) and symptomatic plants at harvest (%) with VY symptoms, the coefficient of determination was 0.79 and significant (p ≤ 0.001) (data not shown).
4 DISCUSSION
In the BChV and BMYV subtrials, we observed a slower spread of symptoms for all inoculation time points compared to the BYV subtrial and greater differences between the treatments in October. This effect is primarily due to the fact that BYV is a semipersistent virus that requires a shorter transmission time by aphids than persistent poleroviruses (Kozłowska-Makulska et al., 2009; Limburg et al., 1997). In the literature, inoculation densities of 100% led to fast symptom development and high incidences at harvest, regardless of inoculation time points of BChV and BMYV (Smith & Hallsworth, 1990; Stevens et al., 2004). This could not be shown in our investigations of different inoculation time points. For BYV, our symptom ratings in June and July showed no clear differences between early inoculation time points, as found by Smith and Hallsworth (1990); however, for the later ratings, the differences between inoculation time points were higher in our study. This shows that low inoculation densities, as used in our study, are more likely to reflect natural infestation and so allow better differentiation between expression of symptoms. It should be noted that the symptoms of VY in the field can easily be confused with other causes, especially in the early stage of infection (4–6 weeks after inoculation). In particular, the co-occurrence of capsid bugs (family Miridae), which cause yellowed leaf tips, makes rating difficult (Draycott, 2008). Therefore, it is important that ratings are carried out by experienced persons and that the main leaf vein is checked for pricks to exclude this as a cause of yellowing. In both trials, yellowed leaf tips due to capsid bugs were very rare and their damage usually has no effect on yield (Draycott, 2008).
Ratings of the percentage of infected plants allow a detailed description of disease progress and enable an evaluation of differences between virus species as well as between different inoculation time points within a virus species. However, these ratings are only suitable for correlation with yield data to a limited extent. For example, regression analysis between the percentage of infected plants at harvest (%) and WSY (t/ha) showed a negative correlation for all virus species in both trials, but with smaller coefficients of determination compared to the correlation between WSY and AUDPC. One reason for this is that the percentage of infected plants at harvest was mostly 100% with a big variance in WSY. Disease severity at harvest does not take into account the course of the epidemiology (e.g., date of onset of disease symptoms and increase of symptoms over time), which is encompassed by the AUDPC. However, it has been used in other crops (Madden et al., 2017) as well as for fungal disease in sugar beet (Kaiser et al., 2010; Wolf & Verreet, 2009). In our study, the calculated AUDPC for BChV, BMYV and BYV in the inoculation time-point trial enabled a good differentiation of the effect of different inoculation time points on disease severity. It clearly shows that, in the case of BChV and BYV, late inoculation time points (BBCH 35–39) had significantly lower disease duration than early time points and that BYV generally led to a stronger disease expression than BChV and BMYV.
The influence of the rating period is particularly evident in the mixed infection trial. Except for the single infection of BMYV, the VY inoculated treatments showed little differentiation at harvest, with 98%–100% infected plants. However, this does not consider that 50% of the plants of the two BYV treatments (BYV and BYV + BtMV) already displayed symptoms in July, compared to less than 5% in the other treatments. Thus, major and significant differences between the VY species were found only for the AUDPC. In addition, the observations in the initial ratings support the above-mentioned effect that a semipersistent virus (BYV) leads to faster transmission and so faster spread and symptom expression compared to poleroviruses.
The yield results of the inoculation time-point trial clearly show that the combination of low inoculation densities and timing of inoculation adapted to the developmental stages (BBCH stages) of sugar beet are suitable for generating data that can be used to derive economic control thresholds. WSY losses (mean of both sites) decreased with increasing stage of development at inoculation (BBCH 10/12–39) from 26.8% to 3.6% for BChV, from 22.0% to 1.7% for BMYV, and from 37.0% to 3.7% for BYV. Stevens et al. (2004) reported a much narrower range of WSY loss depending on inoculation time point (May–July) of 19.2%–14.7% for BChV and 24.5%–21.3% for BMYV. Furthermore, in our study, WSY loss from BChV for the first inoculation time point was much higher than reported by Stevens et al. (2004). The results of Smith and Hallsworth (1990) for BMYV with 27%–0% WSY loss depending on inoculation time point (June–August) were closer to the results from our study. However, Smith and Hallsworth (1990) reported a yield reduction of 27%–11% for inoculation time points in June–July, which is less variation than in our study. For BYV, Smith and Hallsworth (1990) reported a range of WSY loss of 22%–0%, which was clearly lower than in our study with hardly any differentiation between the early time points (22%–18%). The wide range of WSY losses between the different inoculation time points in our study can be explained by the different levels of secondary infection in the plots. Our rating results show that the primary infection, which is restricted to individual plants and mimics natural infection, resulted in limited spread of virus by wingless aphids (secondary infection) in the plots for the later inoculation time points. In contrast, complete spread throughout the plot was observed for the early inoculation time points. Hossain et al. (2021) recommend that low inoculation densities for BChV, BMYV and BYV, mimicking natural infection, are better suited to detect differences between varieties or treatments; our study shows that this is also true for testing the effect of different inoculation time points on WSY. In addition, our results support the reduced inoculation densities (3%–10%) recommended in the updated EPPO protocol for trials with aphids as virus vectors in sugar beet (EPPO, 2023). For BMYV and BYY, it can be assumed, based on our results of the inoculation time-point trial, that an infestation after BBCH 18/19 does not lead to significant yield losses. In the case of BChV, this could only be observed from BBCH 35. This confirms that early control of M. persicae as a vector of VY is particularly important and that later infestation does not result in significant yield losses, even if symptoms develop in the field.
In the BtMV mixed infection trial, BtMV symptoms spread rapidly in all BtMV-inoculated treatments up to 100% from September onwards. This shows that the combination of 30% mechanical inoculation with a 10 days delay before aphid inoculation (at 100% density) is suitable to achieve infection of the entire plot. By July (31 dpi), more than 85% of the plants in all BtMV treatments showed BtMV symptoms and VY symptoms increased from 73% in July (21 dpi) to 100% in August (63 dpi); therefore, we assume that the field trial was suitable for generation of yield effects. For the single BtMV infection, we found a nonsignificant reduction (6%) of WSY and no reduction of sugar content compared to the noninoculated control. One possible reason is that, unlike the VY species, BtMV infection does not lead to chlorosis or necrosis on the leaves and the mosaic-like patches have little or no effect on the plant's photosynthesis. There was no significant difference in WSY reduction by single infections with BChV or BYV compared to mixed infections of these viruses with BtMV; that is, no significant synergistic effect of a co-infection with BtMV was detected. In the case of BYV and BtMV, these effects were also reported by Shepherd et al. (1964). In addition, there were no significant differences in symptom expression (symptom development and AUDPC) between single and mixed infections with BtMV for either virus species in our study. In contrast, Wintermantel (2005) reported significant stronger and faster symptom development, stunted plant growth and a reduced plant biomass in the case of co-infection of BYV + BtMV. However, Wintermantel (2005) carried out greenhouse experiments, which have limited comparability with field trials, and inoculated earlier than in our study, which may also have enhanced symptom expression and yield effects.
In the case of BMYV, we observed a higher proportion of infected plants during the rating period and a significantly higher AUDPC for the BMYV + BtMV mixed infection than for the single BMYV infection. In addition, the WSY loss was significantly higher for the mixed infection (37%) than for the single infection (24%), indicating a significant synergistic effect of co-infection. These effects have not previously been reported for BMYV and BtMV. However, Wintermantel (2005) described significant effects for the co-infection of BWYV, a close relative of BMYV and BtMV on symptom expression, plant growth and biomass.
Infection by natural infestations of M. persicae or contamination from neighbouring plots could not be entirely prevented in either trial. However, the proportion of symptomatic plants was low, ranging from 3% to 13%. Symptoms did not become more widespread until August and there was no visible increase up to harvest, as infestations were confined to individual plants, mainly at the front sides of the plots. In the inoculation time-point trial, the proportion of infected plants was slightly higher for the BYV control than for the BChV and BMYV controls. Reasons for this could be the higher inoculation density in the BYV subtrial, which resulted in a higher number of M. persicae being present in the plots that spread to neighbouring plots, and that BVY is a semipersistent virus. The application of insecticides in the control plots was deliberately avoided to reduce negative effects on the inoculated neighbouring plots. It cannot be ruled out that even very low infestation in the noninoculated controls had a negative effect on yield and quality parameters, leading to an underestimation of the yield effects; however, this does not affect the absolute differences between the inoculated treatments. On the other hand, natural infestation cannot be ruled out in the inoculated treatments either, but the June disease ratings clearly show that there was no natural infestation or contamination in the late-inoculated treatments before inoculation. The challenges described are difficult to eliminate in VY field trials and were also described by Stevens et al. (2004).
Taken together, the results of our study showed for the first time that low inoculation densities of BChV, BMYV and BYV are suitable to achieve good differentiation between different inoculation time points in symptom expression, yield and quality parameters in sugar beet. The epidemiological data are well suited for calculating AUDPC values for BChV, BMYV, BYV and BtMV, according to Madden et al. (2017). Thus, we have shown for the first time that AUDPC values can be calculated for VY in sugar beet and used as an estimate of infestation–loss ratios for VY. Based on these findings, economic control thresholds can be developed as part of the advice on insecticide applications and decision support systems in sugar beet cultivation.
In addition, our study is the first to describe the effects of mixed infections of BtMV with BChV, BMYV or BYV under field conditions, proving a synergistic effect of BtMV and BMYV on yield loss and proportion of symptomatic plants compared to single BMYV infection. BtMV had negligible effects in single infections or even in mixed infections with BChV or BYV. The synergistic effect of BtMV with BMYV needs to be investigated further. Therefore, the breeding for tolerance or resistance to BChV, BMYV and BYV should be prioritized, although there is also a need for further studies with a wider range of varieties and environments.
ACKNOWLEDGEMENTS
We thank all participating technical staff at IfZ, especially Saskia Flentje and Anna Schwarz, for their support during sowing, inoculation, ratings and harvest. We would also like to thank Christine Kenter for her helpful comments on the statistical analysis and proof reading and Sebastian Liebe for his helpful comments on the statistical analysis. The project was partly supported by funds of the Federal Ministry of Food and Agriculture (BMEL) based on a decision of the Parliament of the Federal Republic of Germany via the Federal Office for Agriculture and Food (BLE) through the project ‘EntoProg’. Open Access funding enabled and organized by Projekt DEAL.
Open Research
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.
