Table 1. Results of the multinomial logistic regression models fitted for each variable separately and the multivariate multinomial logistic regression model fitted with a stepwise method.

	LRT testb
Predictora	Chi-square	P> Chi-square	Max-rescaled R2 b	Correct classification (%)b
SPAD	423.05	<0.0001	0.8964	59.57
Ethylene	122.30	<0.0001	0.4854	56.38
Tl-Ta	26.65	<0.0001	0.1358	49.47
Fs	16.78	0.0021	0.0869	47.74
PRI	13.17	0.0105	0.0888	46.28
SPAD, Ethylene, Tl-Ta, Fs	819.35	<0.0001	0.9875	73.40

^aA multinomial logistic regression model was fitted to each of the stress parameters as an independent variable (predictor) and disease severity class as the dependent variable, using healthy plants as the reference category. To assess the combined effect of all stress-related variables, a multiple logistic regression model was fitted using the stepwise procedure. SPAD: Chlorophyll content; Ethylene: Ethylene production; Tl-Ta: Leaf temperature minus air temperature; Fs: Steady-state chlorophyll fluorescence.

^bThe likelihood ratio test (LRT), maximum rescaled R² determination coefficient, and correct classification rate were obtained when using the models for prediction.