Bacterial Blight of Soybean: Survival, Yield Reduction, and Disease Progress Model

106 p.Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 1984.Overwintering of Pseudomonas syringae pv. glycinea (PSG) associated with soybean leaves in the field was evaluated using a streptomycin-rifampicin resistant isolate (PRS-81) of the bacterium. In the 1981-82 trial, PRS-81 was recovered from buried samples (10, 20, and 30 cm below the surface) until February 1982 and from soil surface samples until March 1982. In the 1982-83 trial, Viable cells of the bacterium were recovered from samples at all depths until March 1983. PRS-81 survived better under dry cold conditions than under wet warm conditions. Effects of soil moisture on survival of PRS-81 were greater at 4 C than at 12 C or -12 C. The relationship between bacterial blight severity and yield and 300-seed weight of soybean cultivars Wells II and Williams 79 was studied in 1981 and 1982. Inoculations at five different growth stages of soybean plants resulted in bacterial blight development on different levels of plant canopy. Decrease in disease severity was noticed in the midseason in both years. Williams 79 was not as severely blighted as Wells II in both years. Yield and 300-seed weight of Wells II were significantly different between treatments in 1981 but only 300-seed weight was different in 1982. Bacterial blight severity at the R6 stages of Wells II was negatively correlated with yield in 1981 and with 300-seed weight in both years. Yield and 300-seed weight of Williams 79 were not significantly different between treatments in both years. Growth functions that have been commonly used to depict plant disease progress are restricted by the asymptote value which is fixed at 1. Disease progress curves with asymptote values of less than 1 cannot be accurately described by growth models with the asymptote value of 1. Van der Plant's apparent infection rate, which is the rate parameter of the logistic function with the asymptote value of 1, can not only underestimate the rate of increase in disease severity but also can change the rank of the rates of epidemics to be compared. The weighted mean absolute and the weighted mean relative growth rates of the Richards function can be used as the absolute (Ra) and the relative (Rr) infection rates, respectively, for describing and comparing epidemics with different shapes of development and different asymptote values. The Ra and Rr values are biologically meaningful and universally comparable parameters for disease progress models.U of I OnlyRestricted to the U of I community idenfinitely during batch ingest of legacy ETD