. 2016 Nov 2;7:1569. doi: 10.3389/fpls.2016.01569

Table 2.

Effect of seed soaking with 24-epibrassinolide (EBR) on growth parameters in Brassica juncea seedlings grown under imidacloprid (IMI) toxicity.

Treatments		Seedling length (cm/seedling)			Dry weight (mg/seedling)
IMI (mg/L)	24-EBR (nM)
0	0	14.53^ab ± 0.62			18.6^ab ± 2.17
0	100	15.09^a ± 1.59			19.9^a ± 1.66
150	0	10.87^c ± 1.14			16.1^abc ± 1.18
150	100	13.71^ab ± 0.73			19.6^a ± 0.97
200	0	6.76^d ± 0.80			9.2^d ± 0.91
200	100	12.10^bc ± 0.60			15.4^bc ± 1.57
250	0	2.79^e ± 0.22			5.8^d ± 0.92
250	100	7.79^d ± 0.54			13.5^c ± 1.35
TWO-WAY ANOVA
F-ratios & HSD		F_IMI = 439^***			F_IMI = 208^***
		F_EBR = 310^***			F_EBR = 220^***
		F_{IMI × EBR} = 32.2^***			F_{IMI × EBR} = 20.3^***
		HSD = 2.46^*			HSD = 3.98^*
Multiple linear regression equation		β regression coefficients			MLR	ANN
		β_IMI	β_EBR	β_{IMI × EBR}	r	r (validation)
Seedling length (cm) = 15.38−0.03 X₁ + 0.0042 X₂ + 0.0001 X₁X₂		−0.9825	0.0522	0.4873	0.8984^**	0.9549^***
Dry weight (mg/seedling) = 19.77−0.033 X₁ + 0.0088 X₂ + 0.0002 X₁X₂		−0.9279	0.0929	0.5245	0.8801^**	0.9666^***

Treatments with same superscripts indicates no significant difference at p < 0.05,

^{*, **, and ***}

indicate significant at p < 0.05, p < 0.01, and p < 0.001 respectively. X₁, IMI; X₂, EBR; r, correlation coefficient. Data are mean±standard deviation of 10 seedlings, two-way ANOVA, Tukey's HSD, multiple linear regression analysis and and artificial neural networks (ANN).