Stem blight progress in roselle was studied in four periods and two planting methods. Experimental design was in randomized blocks with four replicates and eight treatments in 4 (periods) x 2 (planting methods) factorial arrangement: direct sowing on 05/02/03, 06/03/03, 05/04/03, 15/05/03 and seedling transplanting on 24/12/02, 25/01/03, 24/02/03, 25/03/03. Immediately after the emergence of symptoms, the disease was evaluated at every 10 days until the end of the cycle, at 205 days, and the percentage of infected stems was calculated. The area under the progress curve of the incidence (AUPCI) was calculated. The disease progress curves for the treatments were adjusted to linear, exponential, monomolecular, logistic and gompertz models. There was significant interaction of methods and planting periods for the disease incidence. Lower incidence of stem blight was noted for seedling transplanting compared to direct sowing in all planting periods. There was increased incidence due to the delay in planting for the two methods. The exponential model best described the disease behavior in all treatments. The statistical differences among treatments considering the progress rate did not reflect the disease intensity in the field. In the same planting period, the initial and maximum quantities of the disease, at the direct sowing, were superior to the treatments related to the seedling transplanting, consistent with the AUPCI values. A direct relationship was recorded between temperature fall and increased incidence of stem blight.
Incidence; AUDPC and progress curve
