3.1. Plant Biomass
In the present study, the fresh and dry weights of roots were evaluated under experimental treatments. The results showed that lower irrigation levels led to a significant decrease in both types of root weight (p < 0.05). It is noteworthy that the effects observed from the three irrigation levels (W60, W80, and W100) were not significantly different from each other (p < 0.05). The lowest fresh and dry weights were observed in response to severe drought treatment (W20) which brought about a decrease of 67% compared to the control. While different irrigation levels had a significant (p < 0.05) impact on the fresh and dry biomass of licorice rhizomes (Figs. 1A and 1B), Si and fungi did not have significant effects on the root weight.
3.2. Root characteristics (length, volume, area, and diameter)
An analysis of licorice root length showed that the longest roots occurred as a result of the well-irrigated treatment (W100), while other treatments showed a significant (p<0.05) reduction in root length, compared to the control (p<0.05). It is noteworthy that no significant difference was observed among other irrigation levels (i.e. W20, W40, W60, and W80) (Fig. 2A).
In the case of root volume, a decrease in irrigation levels caused the root volume to decrease significantly (p<0.05). In this respect, however, the three irrigation levels (W60, W80, and W100) were not significantly different from each other. The smallest root volume was observed in response to the severe drought treatment (W20) which showed a decrease of 64% compared to the W100. While different irrigation levels had a significant (p<0.05) effect on licorice root volume (Fig. 2B), Si and fungi did not have significant effects on this parameter.
According to an analysis of root area (Table S2), maximum root area in well-irrigated treatments was observed as a matter of interaction between Si and F (150.64 cm2). In plants of severe drought stress (W40), however, Si application per se was the cause of maximum root area (113.36 cm2) (Table S2).
In general, root diameter decreased by a lowered level of irrigation (i.e. severe drought stress). F and Si each, per se, maintained the value of root diameter almost akin to the control group under severe drought stress (W20 and W40). However, the interaction between Si and F in W20 caused a significant decrease in root diameter, compared to the control (p<0.05). The highest values were observed by the effect of W80 when using either mycorrhiza (19.45 mm) or Si (18.77 mm) (Table S2).
3.3. Mycorrhizal colonization
Microscopic assessments revealed that non-inoculated plants acquired no or low levels of AMF colonization. C.etunicatum successfully colonized the roots of licorice under various treatments. Plants that were inoculated with C. etunicatum, and were integrated with Si, showed a significant (p<0.05) increase in root colonization. Specifically, these rates of increase were 3321% and 1685% in response to drought stress and well-watered conditions, respectively, compared to the control. Evaluations showed that W40 led to a minimum of root colonization. Meanwhile, Si application increased the amount of this variable in all irrigated levels, except in the case of W20. Evaluations among inoculated plants showed that Si0 treatments exhibited fewer percentages of root colonization, compared to the Si1 treatment. Observations also revealed that the Si improved root colonization by 11.51% under severe drought stress, i.e. at W40. In general, Si had a positive effect on root colonization in all irrigation levels, except at W20 which meant an unrelenting effect of severe drought stress (Fig. 3).
3.4. Glycyrrhizic acid
An analysis of GA in licorice roots showed that the highest amount of this metabolite (26.62 mg/g DW) was obtained at W40 by applying integrated Si and F (Fig. 4). In well-irrigated treatments, the Si application alone was able to enhance GA content (18.39 mg/g DW). Regardless of Si usage, the severe drought level (W40) caused mycorrhizal-inoculated plants to show a notable increase in the amount of GA (16.43 mg/g DW), compared to the same irrigation level in non-mycorrhizal plants (Fig. 4).
3.5. Total flavonoid, antioxidant activity, and IC-50
Total flavonoids in licorice roots were increased significantly by Si and F (p<0.05). They caused 7.85 and 9.28% enhancement in the total flavonoid content, respectively (Figs. 5A and 5B). G. glabra root extracts showed an ability to scavenge DPPH free radicals which have been often seen as a measure of total antioxidant activity (Fig. 6). The total antioxidant activity in licorice roots at different levels of irrigation and mycorrhizal inoculation showed that the well-irrigated treatment (W100) led to the highest antioxidant activity (60.86%). In severe drought stress (W20), the presence of mycorrhiza ultimately maintained the level of antioxidant activity at 59.65%, although this had no significant difference with the control. Regardless of Si and F, root extracts of plants in the 80% FC irrigation demonstrated a higher ability to scavenge free radicals (IC-50 = 0.73 mg/ml) as compared to other irrigation levels. As a matter of Si and mycorrhizal interaction, severe drought stress (W40) induced a higher capacity in plants to scavenge free radicals (IC-50 =1.63 mg/ml). Si alone was able to keep the IC-50 at a low quantity (0.95 mg/ml) in response to W80, hence its great capacity to scavenge free radicals (Table S3).
3.6. Total phenol and polyphenol profile
The results of total phenol determination showed that Si and F, separately, did not have a significant effect on the amount of total phenol at different drought stress levels. But interactions between mycorrhiza and Si significantly (p<0.05) increased the amount of total phenol in the W40 group (1.74 mg/g DW) compared to other irrigation levels. In plants without Si and mycorrhizal inoculation, no significant difference was observed between the effects of the well-irrigated treatment (W100) and those of the severe drought stress (W20 and W40) on total phenol content (Table S3).
The HPLC examination of polyphenolic compounds in licorice roots and their profile showed that the highest amount of sinapic acid (10.11 mg/g DW) occurred in well-irrigated treatments when plants were inoculated with mycorrhiza. In severe drought stress levels (W40), however, the combined application of Si and F could significantly (p<0.05) increase the amount of sinapic acid (16.46 mg/g DW). As the irrigation level decreased, the quercetin content reduced significantly (44.1%) compared to the well-irrigated treatment (W100) (p<0.05). Irrespective of Si and F, more severe drought stress levels caused a significant increase (p<0.05) in the amount of trans-ferulic acid (0.74-0.91 mg/g DW). Nonetheless, Si application on plants of the W100 treatment maximized the amount of trans-ferulic acid (1.11 mg/g DW) (p<0.05) (Table S4).
3.7. Concentration of plant nutrients
Concentrations of select nutrients were evaluated in licorice roots (Table S5). Regardless of Si and mycorrhiza treatments, the concentration of K reached a maximum amount in the well-irrigated treatment (W100). In plants of the Si or F treatments, there was no significant variation in K amount at various drought stress levels. The amount of N in inoculated plants, at all irrigation levels, showed a significant increase (p < 0.05), compared to non-inoculated plants. Si and F interactions were able to maximize the N concentration in plants of the severe drought stress treatment (W20). However, Si alone could not significantly increase the N concentration in response to the severe drought treatment (W20). Regardless of Si, the concentrations of P in inoculated plants were significantly (p<0.05) higher than in non-inoculated plants in response to well-irrigated and moderated drought levels (W100, W80, and W60). In the case of Si and F interaction, W80 and W20 treatments led to the maximum amount of P. Drought stress had no significant effect on P concentration at 40, 60, 80, and 100% FC, irrespective of AMF and Si treatments (Table S5).
3.8. Principal component analysis of the measurable traits of G. glabra in a combination of effects caused by mycorrhiza, silicon, and drought
In the biplot of PC analysis, the first two PCs represented 46% of the variation in the measured traits among the treatment groups (Fig. 7). The first PC explained 24% of the variation and comprised root diameter, root fresh and dry weights, root volume, root area, root length, quercetin, and Si. In contrast, the second PC accounted for 21.9% of trait-related variations, comprising root colonization, total phenol and flavonoid contents, and nitrogen content. In the bi-plot and PC analyses, the cosine of the angles between vectors showed the extent of correlation between traits. In the present study, the projection of treatment groups on the two PCs in the bi-plot showed that the studied treatments were divided into two distinct groups. The first group comprised 12 treatments, i.e. T1, T2, … and T12 in association with the first PC-linked measurements of parameters. Meanwhile, the treatments T13, T14, … and T20 were categorized into the second group for having a higher association with the second PC-linked traits (Fig. 7).
3.9. Correlation and path analysis of measured parameters in G. glabra under a combination of mycorrhiza, silicon, and drought
Using correlation and path coefficient analysis, there were successful evaluations of the relations among licorice root traits, including root colonization, elemental analysis, and secondary metabolite production (Figs. 8 A and 8 B). Stronger positive correlations were found between root traits. These were, namely, root fresh weight, dry weight, volume, and diameter. Among the secondary metabolites, total phenol and total flavonoids showed a good positive correlation between each other (Fig. 8 A). In comparison with the simple correlation analysis, path analysis was aimed at the dependent and independent traits of licorice. The results demonstrated that sinapic acid and GA maintained the highest direct influence as expressed in β=0.719 (p<0.001). Also, total phenol and total flavonoids showed a significant direct effect (p<0.01) as expressed in β=0.707 (Fig. 8 B). In sum, these traits can be improved in their performance through path analysis.