Table 4.

Effects of disease severity class, tissue type, and cultivar on the isolation frequency of Fusarium pseudograminearum, F. pseudograminearum biomass, and DON and D3G contents.

Factors	Isolation frequency			F. pesudograminearum DNA content			DON contentx			D3G content
	Deviance	F	P	Sum Sq	F	P	Sum Sq	F	P	Sum Sq	F	P
Disease severity classes	15.11	3.78	0.004	18.85	14.04	<0.001	14.39	38.57	<0.001	22.33	62.81	<0.001
Segment samples	22.86	3.27	0.002	47.81	20.34	<0.001	12.37	18.94	<0.001	8.90	14.30	<0.001
Cultivar	0.23	0.23	0.630	8.95	26.64	<0.001	0.03	0.34	0.561	3.07	34.59	<0.001
Residuals	43.85			22.49			6.25			5.95
Total variation	82.05			98.10			33.04			40.25

^xDON, deoxynivalenol; D3G, DON-3-glucoside. Multiple factor analysis of Fusarium pseudograminearum biomass, DON, and D3G content on different disease severity classes, segment samples, and cultivars. The F. pseudograminearum DNA, DON, and D3G content is depicted as the log(x+1) of the their content. The degree of freedom of cultivar, disease severity classes, segment samples, and residuals are 4, 7, 1, and 67, respectively. Based on the distribution characteristics of data, multiple linear regression analysis was then used to establish the association of DON, D3G, and F. pseudograminearum DNA with segment samples, disease severity classes, and cultivar. The contribution of each factor to the variability of the isolation frequency of F. pseudograminearum, the F. pseudograminearum DNA, DON, and D3G content variation was determined by the ratio of the factor’s sum square to the total sum square in the ANOVA model. Generalized linear modeling (GLM) was then used to establish the association of the isolation frequency of F. pseudograminearum with disease severity classes, segment samples, and cultivar used. Residual errors were assued to follow a binomial distribution. The effect of disease severity classes, segment samples, and cultivars were determined based on the F test of deviences.