No abstract

ABSTRACT:The objective of this study was to verify the efficiency of the control of white mold on soybean with the use of fungicides applied alone and in rotation, at different growth stages and in a mixture of two active ingredients at three locations: Arapoti, PR, Mauá-da-Serra, PR and Pinhão, PR, Brazil. The fungicides used were carbendazim (Ca), thiophanate methyl (Tm), procymidone (Pr) and fluazinam (Fl). The experiments consisted on 17 treatments and 4 replications in a randomized block design. The analyzed variables were severity, incidence, number of sclerotia and yield. Mauá-da-Serra and Pinhão presented the highest incidences (31% and 29.8% in the control, respectively). At these two locations most of treatments with fungicides decreased the incidence and production of sclerotia, when compared to control; however, no differences in terms of yield were observed. Arapoti presented the lowest incidence (15.8% in the control) where most of treatments with fungicides did not present differences for the variables incidence, production of sclerotia and yield, when compared to the control. No differences were also observed for severity in any of three locations. In conclusion, fungicides applied in soybean areas with historically white mold incidence up to 31% can reduce the disease incidence and sclerotia production levels.

White mould (Sclerotinia sclerotiorum) is a damaging disease of soybean crops in Brazil. Within-field spatial analyses of white mould disease and inoculum, as well as yield components at the field level may provide insights into their relationship, but this information is lacking in the country. During three years, spatially explicit within-field data on white mould and soybean stand and yield were measured at three fields, with sizes ranging from 4 to 12 ha, located in the Campos Gerais region of Paraná State, Brazil and all naturally infested with sclerotia. Crop stand, white mould incidence and severity, soybean yield and sclerotia collected at harvest were assessed in quadrats of 7.2 m 2 distant from each other by 8 to 50 m. Soilborne sclerotia were counted in four quadrats (0.25 m 2 and 0.05 m depth) at each sampling point. The semivariograms fitted the pure nugget effect, linear, spherical, and exponential models. The pure nugget model best fitted sclerotia data, suggesting a random distribution. Significant and positive associations were found between disease incidence and sclerotia amount at harvest, which were both negative associated with yield.

No abstract