Effects of stem canker (Leptosphaeria maculans) and light leaf spot (Pyrenopeziza brassicae) on yield of winter oilseed rape (Brassica napus) in southern England

The relationships between yield loss and incidence or severity of stem canker and light leaf spot in winter oilseed rape were analysed by correlation and regression analyses, using data from experiments at Rothamsted, England in 1992/93, 1994/95 and 1995/96. Growth stages (GS) 6,3/6,4 and 4,0/4,5 were identified as the critical points for relating percentage yield loss to stem canker and light leaf spot (on stems), respectively. Critical point (CP) and area under disease progress curve (AUDPC) models relating percentage yield loss to combined incidence or severity of stem canker and light leaf spot (stems) in each experiment were constructed by linear regression. There were no differences in the CP models for incidence between 1992/93, 1994/95 and 1995/96 experiments, or in the AUDPC models for incidence between 1992/93 and 1994/95 experiments. Therefore, a general CP model relating percentage yield loss (Delta Y) to combined incidence of stem canker (S-i) at GS 6,3/6,4 and light leaf spot (stems) (L-i) at GS 4,0/4,5 was constructed using data from the three experiments: Delta Y = 0.85 + 0.079S(i) + 0.065L(i) (R-2 = 43.7%, P < 0.001, 92 df). A general AUDPC model relating Delta Y to the AUDPC of combined incidence of stem canker (S-ia) from GS 5.7 to GS 6.5 and light leaf spot (stems) (L-ia) from GS 4.0 to GS 6.3 was constructed using data from the 1992/93 and 1994/95 experiments: Delta Y = 0.07 + 0.00096S(ia) + 0.0026L(ia) (R-2 = 43.6%, P < 0.001, 68 df). These two general yield-loss models were tested with data from Rothamsted in 1993/94 and Boxworth in 1992/93. The predictive accuracy of the CP model based on combined incidence of stem canker and light leaf spot (stems) was better than that of the AUDPC model. Yield losses predicted by summing the estimates from individual models for incidence of stem canker alone (GS 6,3/6,4) and light leaf spot alone (on leaves at GS 3,3) were greater than observed yield losses in experiments at Rothamsted in 1992/93, 1993/94, 1994/95 and 1995/96 and at Boxworth in 1992/93.