**2. Material and methods**

Seeds of soybean cultivar conquista (MG/BR 46) and suprema were sown in pots containing 5 kg of soil mixture, sand and organic matter (manure) in the proportion 2:1:0.5 in a green house. Thinning was performed 15 days after planting, leaving two plants per pot, forming the experimental unit. The plants were kept in a green house until the V3 stage, according to the soybean phenological scale proposed by Ritchie et al. (1982). During the same period, inoculation of *M. diffusa* was done stirring soybean plants on healthy plants which were then randomly placed next to diseased plants (Demski & Phillips, 1974). According to Grau (1975), because of the ease with which conidia are disseminated, it becomes hard to test inoculation of *M. diffusa* with different isolates without contamination.

The plants were transferred to growth chambers and arranged in randomized blocks, factorial 4 x 5 with three replicates, considering four air temperatures (15, 20, 25 and 30 degrees C) and five leaf wetness periods (0 , 6, 12, 18 and 24 hours). For the different periods of leaf wetness, recently sprayed plants were kept in a moist chamber with transparent plastic bags, during the period used for each treatment. In the treatment of 0 h of wetness, the plants were taken without a moist chamber for the growth chambers. Irrigation was performed by spraying water directly on the stem of the plants.

There were four incidence and severity assessments every five days after the beginning of the experiment. The severity was assessed on all central leaflet of each plant with trifoliate leaves at 9, 11, 13 and 15 days after inoculation, using the grading scale published Sartorato & Yorinori (2001), adopting grade 1 = 1% of affected leaf area; grade 2 = 5%, grade 3 = 10%, grade 4 = 25%, grade 5 = 50%, grade 6 = 100%.

The intensity data were integrated using the area under incidence progress curve over time, according to Campbell & Madden (1990).

$$ALIDPCS = \sum\_{i}^{n-1} (\frac{ys\_i + ys\_{i+1}}{2})(t\_{i+1} - t\_i) \tag{1}$$

$$ALDDPCI = \sum\_{i}^{n-1} (\frac{y i\_i + y i\_{i+1}}{2})(t\_{i+1} - t\_i) \tag{2}$$

Where:

184 Soybean Physiology and Biochemistry

Yorinori , 2001). In general, the lower leaves of young plants are more susceptible than the

In relation to physiological changes in the host, Mignucci & Boyer (1979) studied the inhibition of photosynthesis and transpiration of soybean infected with powdery mildew and found lower photosynthesis and transpiration with increased infection. With 82% of leaf area infected, more than half of the leaf photosynthetic activity had been lost and transpiration dropped to 36% compared to control, considering the direct result of the change in metabolic activity induced by the pathogen. Because infection occurs primarily in the lower leaves and poorly lit, it is unlikely that the reduction in rates of photosynthesis and leaf transpiration resulted in great reduction in soybean yield, however, favorable climatic conditions may enabled the infection of upper leaves leading to high losses

Susceptibility of cultivars and influence of the climate has caused outbreaks of powdery mildew in successive years in Brazil. The lack of resistance in most cultivars have required chemical control mainly in the south and the high plateaus of the savannahs. In the U.S.A., powdery mildew caused economic damage reached in the 70's and early 80's. Since then, the

Reactions of different soybean varieties to powdery mildew and the effect of environmental variables in the progress of the disease have been reported (Arny et al. 1975; Buzzell et al. 1975; Degree & Laurence, 1975, Johnson & Phillips, 1961; Mignucci 1977; Mignucci & Boyer, 1979; Mignucci & Lim, 1980; Lohnes & Bernard, 1992; Lohnes & Nickell, 1994). According to Bedendo (1995), in Brazil, powdery mildew may occur in the humid and cold climates, but are favored by hot dry conditions (20-25 °C). According to the author, conidia do not germinate when is present a film of water on the leaf surface, however, relative humidity

Mignucci et al. (1977) reported temperatures of 18 ºC as favorable to the development of powdery mildew on susceptible cultivars and at temperatures of 30 ºC disease progress was inhibited. Degree & Laurence (1975) also observed lower disease severity at 30 °C. According to Sartorato & Yorinori (2001) the information about the effects of relative humidity, leaf wetness, rainfall, solar radiation or other environmental factors in the

Therefore, the intensity of powdery mildew of soybean under different temperatures and

Seeds of soybean cultivar conquista (MG/BR 46) and suprema were sown in pots containing 5 kg of soil mixture, sand and organic matter (manure) in the proportion 2:1:0.5 in a green house. Thinning was performed 15 days after planting, leaving two plants per pot, forming the experimental unit. The plants were kept in a green house until the V3 stage, according to the soybean phenological scale proposed by Ritchie et al. (1982). During the same period, inoculation of *M. diffusa* was done stirring soybean plants on healthy plants which were then randomly placed next to diseased plants (Demski & Phillips, 1974). According to Grau (1975), because of the ease with which conidia are disseminated, it becomes hard to test

The plants were transferred to growth chambers and arranged in randomized blocks, factorial 4 x 5 with three replicates, considering four air temperatures (15, 20, 25 and 30

periods of leaf wetness on the cultivars conquista and suprema was evaluated.

inoculation of *M. diffusa* with different isolates without contamination.

use of resistant cultivars has dispensed chemical control (Sartorato & Yorinori, 2001).

upper leaves (Mignucci & Lim, 1980).

near 95% is required for germination.

**2. Material and methods** 

(Mignucci & Boyer, 1979; Sartorato & Yorinori, 2001).

progress of powdery mildew in soybeans was not precise.

AUDPCI was the area under the progress curve of powdery mildew incidence; AUDPCS was the area under the progress curve of powdery mildew severity; ys and yi were the disease severity and incidence over time i and i+1, respectively; t was the time in days and n was the number of evaluations along the time.

Plants with chlorosis, green islands, rusty stains, and combination of these symptoms were considered infected by powdery mildew (Figure 1).

Fig. 1. Evaluated signals of powdery mildew in soybean plants.

The significant variables in the F test by the variance analysis of AUDPCS were subjected to regression analysis, linear and nonlinear adjustment models (Leite & Amorim, 2002, Reis et al., 2004). In the case of significant interaction, the combined effect of temperature and leaf wetness duration in disease intensity was modeled (Reis et al., 2004).
