E ﬀ ects of Arbuscular Mycorrhizal Fungi on Growth and Physiological Performance of Catalpa bungei C.A.Mey. under Drought Stress

: Catalpa bungei C.A.Mey. is a common ornamental timber species. Its survival and growth are greatly a ﬀ ected by water scarcity in arid and semi-arid areas of Northwest China. Evidence suggests arbuscular mycorrhizal fungus (AMF) may improve plant drought resistance. However, there is limited information on the systematic e ﬀ ects of AMF on drought resistance in C. bungei seedlings. Here, a pot experiment was used to explore the e ﬀ ects of inoculation with the AMF Rhizophagus intraradices on the growth and physiological performance of C. bungei under di ﬀ erent water treatment conditions. Three water levels and two mycorrhizal inoculation treatments were used with factorial design. The results showed that drought stress noticeably a ﬀ ected the growth and physiological performance of C. bungei seedlings. However, inoculation with R . intraradices signiﬁcantly ameliorated the growth, and alleviated the e ﬀ ects of drought stress. The growth parameters of AMF-inoculated seedlings signiﬁcantly increased regardless of water status. AMF changed the biomass allocation in seedlings by reducing the root mass ratio (RMR) and root / shoot ratio. AMF-inoculated seedlings displayed higher gas exchange parameters, photosynthetic pigment concentrations, speciﬁc leaf area (SLA), but lower speciﬁc leaf weight (SLW), regardless of water status. AMF alleviated drought-induced oxidative stress by attenuating the excess generation of reactive oxygen species (ROS), especially H 2 O 2 and O 2 − , in leaves. Inoculation with AMF under drought stress also dramatically augmented indole-3-acetic acid (IAA) and gibberellins (GA 3 ) levels and the IAA / abscisic acid (ABA) and GA 3 / ABA ratios, but reduced ABA and zeatin (ZT) levels in leaves. AMF symbiosis improved root morphology and promoted the absorption of nitrogen (N) and phosphorus (P) in seedlings. We conclude that inoculation with R. intraradices is potentially useful for a ﬀ orestation and cultivation of C. bungei in Northwest China. Furthermore, AMF improved soil structure by increasing the glomalin-related soil protein (GRSP) contents and the proportion of macro-aggregates (0.25–0.5 mm) in the rhizosphere soil.


Introduction
Drought is one of the most important factors affecting plant growth in terrestrial ecosystems [1]. Climate change increases the odds of severe drought conditions around the world [2,3]. Drought stress can change plant water status, and has obvious effects on seed germination, plant morphological structure, biomass distribution, soil nutrient availability, photosynthesis, and metabolism, thereby seriously restricting the normal growth, development, survival, and productivity of plants [3][4][5][6]. Plants adapt to drought by making morphological and physiological adjustments through the synergistic

Experimental Design
The pot experiment was conducted from April to September 2019 in the greenhouse of the Northwest A & F University, using factorial design, including water and inoculation treatments. Six treatments were performed, with 10 replicates each. Water treatments were divided into three categories: well-watered (WW), moderate drought (MD), and severe drought (SD), which corresponded to a soil water content of 70%, 50%, and 30% of the maximum field capacity (Maximum field capacity was determined according to the method reported by Yan et al. [32]), respectively. The inoculation treatments were performed with R. intraradices and non-inoculated seedlings were used as controls.
C. bungei seedlings with uniform growth trends were selected and transplanted into plastic flowerpots containing about 4.2 kg of cultivation substrate on 15 May 2019. At the time of transplanting, 3 g of R. intraradices inoculum was spread evenly on the roots until full contact (inoculated treatment group). The control group was treated with the same amount of inactivated inoculum (121 • C, 2 h). After transplanting, normal water supply was maintained until the seedlings grew stably (approximately 45 d). Root samples of three seedlings were randomly selected for the detection of AMF colonization. After successful colonization, water treatments were conducted, and the soil water content was controlled using the weighing method for 2.5 months (from 10 July to 25 September 2019).

Measurement of AMF Colonization
Seedling roots were first washed with running tap water and then several times with distilled water. Samples were randomly cut into small root segments (approximately 1 cm) and immersed in 10% KOH solution, after which they were placed in a water bath at 90 • C for 30 min. After the root segments became transparent, any residual KOH was washed off with distilled water, and the root segments were dyed with 0.05% Trypan blue staining solution at 90 • C for 30 min. Samples were then decolorized with a mixture of lactic acid and glycerin (v/v = 1:1) three times [33]. Next, samples were observed under a 400× optical microscope (Olympus Bx43, Tokyo, Japan), and the AMF colonization rate was calculated according to the magnifying cross method [34].

Determination of Plant Growth Parameters and Biomass Allocation
Seedling height and basal diameter were measured before harvest using a tape measure and vernier caliper, respectively. Root, stem, and leaf fresh weights were measured immediately after the harvest, whereas dry weights were determined after drying the biomass (at 65 • C, until constant weight). Next, the biomass allocation of each part was calculated. Leaf mass ratio (LMR), stem mass ratio (SMR), and root mass ratio (RMR) were determined as the ratios of leaf, stem and root biomass to the total biomass, respectively. Root/shoot ratio was calculated as the ratio of underground biomass to the aboveground biomass [36].

Analysis of Nutrient Absorption and Distribution
Powdered root, stem, and leaf samples were sieved through a 0.45 mm sieve, and N, P, and K concentrations were determined using the Kjeldahl, molybdenum-stibium colorimetric, and flame photometric method [37], respectively.

Determination of Specific Leaf Area and Specific Leaf Weight
Leaf area was calculated from a single leaf at the same position, using the transparent grid method [34]. Leaves were dried at 65 • C (to constant weight), and their dry weights were determined. Specific leaf area (SLA) was calculated as the ratio of a single leaf area to its dry weight, whereas the specific leaf weight (SLW) was calculated as the ratio of a single leaf dry weight to its area, which is essentially the reciprocal of the SLA.

Estimation of the Root System Architecture
Cleaned roots of the fresh seedlings were scanned and analyzed using the root scanner system (WinRHIZO 2013e). Morphological parameters, such as root length, root surface area, projected area, average root diameter, and root branching number were calculated [38].

Determination of ROS
Fresh leaf and root samples were ground in pre-cooled mortars with liquid nitrogen, and hydrogen peroxide (H 2 O 2 ) and superoxide anion radical (O 2 − ) contents were determined according to the method reported by Gao [35].

Determination of GRSP Content
The rhizosphere soil was shaken off, collected, air-dried and filtered through a 1 mm sieve. Half gram of soil was put into a plastic centrifuge tube with 4 mL of extraction solution. The mixture was well suspended, and placed in an autoclave (121 • C, 30-60 min), to extract the easily extractable glomalin-related soil protein (EE-GRSP; 20 mmol·L −1 sodium citrate solution, pH 7.0) and the total glomalin-related soil protein (T-GRSP; 50 mmol·L −1 sodium citrate solution, pH 8.0). Tube contents were centrifuged at 10,000 rpm for 15 min, and the supernatants were collected. Absorbance of the supernatants was determined at 595 nm, using the Coomassie Brilliant Blue method. EE-GRSP and T-GRSP contents were calculated according to a standard bovine serum albumin curve [41].

Statistical Analysis
Data were analyzed using the SPSS 21.0 software, and the differences among treatments were compared using one-way analysis of variance (ANOVA) and independent-sample T test. Differences among the treatments were examined using the Duncan test at p < 0.05. A two-way ANOVA was used to determine the significance of the water and inoculation treatments, and their interaction with the experimental parameters. Pearson's correlation coefficients among parameters were analyzed using SPSS [8]. All data are reported as mean ± standard error (SE). Mapping was carried out using SigmaPlot 12.0.

AMF Root Colonization
No AMF colonization was found in the roots of non-inoculated C. bungei seedlings (Figure 2A), whereas clear colonization by R. intraradices was observed in the mycorrhizal inoculated seedlings. Mycorrhizal hyphae ultimately formed distinct morphological structures, such as intraradical hyphae, arbuscules, and vesicles, under WW ( Figure 2B), MD ( Figure 2C), and SD ( Figure 2D) conditions. Root colonization rate in the R. intraradices-inoculated seedlings was above 60% under the three water treatment conditions ( Figure 2E), indicating that R. intraradices could form a good symbiotic relationship with the seedling roots. Colonization rate of the mycorrhizal fungi was 86.89% under the MD treatment, which was only slightly lower than that for WW (87.55%), whereas it decreased sharply under SD stress (65.74%, p < 0.05) ( Figure 2E). symbiotic relationship with the seedling roots. Colonization rate of the mycorrhizal fungi was 86.89% under the MD treatment, which was only slightly lower than that for WW (87.55%), whereas it decreased sharply under SD stress (65.74%, p < 0.05) ( Figure 2E).

Growth Parameters
Drought stress noticeably inhibited the growth of C. bungei seedlings ( Figure 3). Both under inoculated and non-inoculated conditions, the plant height and leaf area of the C. bungei seedlings showed significant decreases with an increase in drought stress ( Figure 3A,C); the basal diameter and total biomass also decreased, but only under SD conditions (in the inoculation treatment) was this difference significant (p < 0.05) ( Figure 3B,D). Plant height, basal diameter, leaf area, and the total biomass of mycorrhizal inoculated C. bungei seedlings increased significantly under both WW and MD conditions, compared with those for the control treatment ( Figure 3). Plant height and total biomass increased more than 1.9-fold, basal diameter approximately 1.35-fold, and leaf area 3.7-(WW condition) and 5.4-(MD stress) fold, compared with those for the control treatment. Under SD stress, the growth parameters of mycorrhizal inoculated seedlings (basal diameter, leaf area, and total biomass) increased, but these differences were not significant (p < 0.05). Thus, AMF inoculation

Growth Parameters
Drought stress noticeably inhibited the growth of C. bungei seedlings ( Figure 3). Both under inoculated and non-inoculated conditions, the plant height and leaf area of the C. bungei seedlings showed significant decreases with an increase in drought stress ( Figure 3A,C); the basal diameter and total biomass also decreased, but only under SD conditions (in the inoculation treatment) was this difference significant (p < 0.05) ( Figure 3B,D). Plant height, basal diameter, leaf area, and the total biomass of mycorrhizal inoculated C. bungei seedlings increased significantly under both WW and MD conditions, compared with those for the control treatment ( Figure 3). Plant height and total biomass increased more than 1.9-fold, basal diameter approximately 1.35-fold, and leaf area 3.7-(WW condition) and 5.4-(MD stress) fold, compared with those for the control treatment. Under SD stress, the growth parameters of mycorrhizal inoculated seedlings (basal diameter, leaf area, and total biomass) increased, but these differences were not significant (p < 0.05). Thus, AMF inoculation could significantly promote the growth of C. bungei seedlings and alleviate the inhibition of growth by drought stress, but this effect was not obvious under SD stress.
Forests 2020, 11, x FOR PEER REVIEW 7 of 31 could significantly promote the growth of C. bungei seedlings and alleviate the inhibition of growth by drought stress, but this effect was not obvious under SD stress. non-AMF-inoculated; AM, AMF-inoculated; WW, well-watered; MD, moderate drought; SD, severe drought; DS, drought stress. Different lowercase letters above the bars indicate significant differences (p < 0.05) among treatments. Values are means ± SE. Plant heights and basal diameters are based on 9 biological replicates, while leaf area and total biomass represents the average of three biological replicates. Two-way ANOVA output: ns, not significant; * p < 0.05; ** p < 0.01.

Biomass Allocation
Drought stress and AMF inoculation significantly changed the biomass allocation pattern of C. bungei (Table 1). Increasing drought stress caused a decrease in LMR and leaf and stem biomass of the C. bungei seedlings (both inoculated and non-inoculated). There was a slight increase in SMR under the same condition. The root biomass of both inoculated and non-inoculated seedlings, together with RMR and the root/shoot ratio of non-inoculated seedlings, increased under MD stress and decreased under SD stress, but there were no significant differences between the water treatments (p > 0.05). RMR and the root/shoot ratio of inoculated seedlings, on the other hand, showed a notable increase (2.43 and 3.41-fold higher than that under WW, respectively; Table 1). Under SD stress, the leaf biomass of the non-mycorrhizal treated seedlings, as well as the leaf and stem biomass and LMR of the mycorrhizal inoculated seedlings, decreased significantly (52.13%, 20.91%, 34.34%, and 65.63% compared to that of the WW conditions, respectively). Under WW conditions, most of the non-inoculated seedling biomass was distributed in the leaves, followed by the roots, and with the least proportion allocated to stems (leaf > root > stem), whereas under drought stress it showed a root > leaf > stem distribution. The biomass of the mycorrhizal inoculated seedlings, however, tended to be distributed in leaves, followed by the stems, while the NAM, non-AMF-inoculated; AM, AMF-inoculated; WW, well-watered; MD, moderate drought; SD, severe drought; DS, drought stress. Different lowercase letters above the bars indicate significant differences (p < 0.05) among treatments. Values are means ± SE. Plant heights and basal diameters are based on 9 biological replicates, while leaf area and total biomass represents the average of three biological replicates. Two-way ANOVA output: ns, not significant; * p < 0.05; ** p < 0.01.

Biomass Allocation
Drought stress and AMF inoculation significantly changed the biomass allocation pattern of C. bungei (Table 1). Increasing drought stress caused a decrease in LMR and leaf and stem biomass of the C. bungei seedlings (both inoculated and non-inoculated). There was a slight increase in SMR under the same condition. The root biomass of both inoculated and non-inoculated seedlings, together with RMR and the root/shoot ratio of non-inoculated seedlings, increased under MD stress and decreased under SD stress, but there were no significant differences between the water treatments (p > 0.05). RMR and the root/shoot ratio of inoculated seedlings, on the other hand, showed a notable increase (2.43 and 3.41-fold higher than that under WW, respectively; Table 1). Under SD stress, the leaf biomass of the non-mycorrhizal treated seedlings, as well as the leaf and stem biomass and LMR of the mycorrhizal inoculated seedlings, decreased significantly (52.13%, 20.91%, 34.34%, and 65.63% compared to that of the WW conditions, respectively). Under WW conditions, most of the non-inoculated seedling biomass was distributed in the leaves, followed by the roots, and with the least proportion allocated to stems (leaf > root > stem), whereas under drought stress it showed a root > leaf > stem distribution. The biomass of the mycorrhizal inoculated seedlings, however, tended to be distributed in leaves, followed by the stems, while the least allocated proportion was to the roots (leaf > stem > root). Under drought stress, the biomass allocation trend was as follows: leaf > root > stem.
AMF inoculation promoted the biomass accumulation in leaves and stems of C. bungei seedlings, increased LMR and SMR, but inhibited the accumulation of root biomass, and decreased RMR and root/shoot ratio under both WW and drought stress, compared with those in the non-inoculated conditions ( Table 1). Leaf and stem biomass of the mycorrhizal inoculated seedlings were significantly higher, while RMR was significantly lower both under WW (3-fold, 3.5-fold and 40%, respectively) and MD (2.8-fold, 2.7-fold and 44.23%, respectively), than in the non-mycorrhizal treated seedlings. However, there were no significant differences in all the indices between the inoculated and control treatments under SD stress (p > 0.05).

Photosynthetic Gas Exchange Capacity
Drought stress significantly reduced the gas exchange parameters of C. bungei seedlings ( Table 2). The Pn, Gs, Ci, Tr, and WUE of the C. bungei seedlings decreased gradually with the intensification of stress, with or without AMF inoculation. Among those, Pn, Gs, and Tr were significantly lower under MD and SD stress than under WW, and WUE was significantly lower under SD stress than under WW. Under SD stress, Pn, Gs, Ci, Tr and WUE of both the control (or inoculated) treatments were only 30.12% (37.88%), 27.27% (33.33%), 65.14% (83.82%), 38.12% (48.48%) and 77.65% (76.22%) of those under WW conditions. Under the same water conditions, Pn, Gs, Tr, and WUE of seedlings inoculated with AMF were all higher than those of non-inoculated seedlings, while Ci was lower than that of the non-inoculated seedlings ( Table 2). Under MD stress, Gs and WUE of the mycorrhizal inoculated seedlings were significantly higher, whereas Ci was significantly lower than those of the non-inoculated seedlings (1.5-fold, 1.15-fold and 81.03%, respectively). However, there was no significant difference between the inoculated and control treatments under SD stress (p > 0.05). These results showed that AMF could improve the photosynthetic capacity and WUE of C. bungei seedlings to some extent, but this positive effect was inhibited under SD stress.

Photosynthetic Pigment Concentrations
Regardless of the AMF status, the concentrations of photosynthetic pigments (chlorophyll a, b, and carotenoids) decreased in the C. bungei leaves with increasing drought stress, whereas the chlorophyll a/b values increased gradually (Table 3). Under SD stress, the concentrations of pigments decreased significantly both in the non-inoculated (or inoculated) seedlings, being only 31.32% (45.65%), 27.45% (41.38%), and 35.29% (55.88%) of those under WW treatments, respectively. AMF inoculation increased leaf photosynthetic pigment concentrations to some extent, especially under drought stress (Table 3). Although there was no significant difference in the concentration of photosynthetic pigments between the inoculated and control treatments (p > 0.05), the pigment concentrations of the mycorrhizal inoculated seedlings was higher than those of the non-mycorrhizal treated seedlings-except for the chlorophyll a/b value, which was lower-under both WW and drought stress.

SLA and SLW
With increasing drought stress, the SLA of both inoculated and non-inoculated C. bungei seedlings decreased gradually, while SLW increased significantly ( Figure 4). SLA and SLW were significantly different between WW and drought stress treatments (p < 0.05). Under SD, the SLA and SLW of AMF-inoculated (or non-inoculated) seedlings were 66.53% (61.19%) and 1.5-fold (1.6-fold) of those under WW treatments, respectively. The SLA of mycorrhizal inoculated seedlings was significantly higher under WW and MD stress, while the SLW was significantly lower than that of non-mycorrhizal treated seedlings under the three water treatments ( Figure 4).

SLA and SLW
With increasing drought stress, the SLA of both inoculated and non-inoculated C. bungei seedlings decreased gradually, while SLW increased significantly ( Figure 4). SLA and SLW were significantly different between WW and drought stress treatments (p < 0.05). Under SD, the SLA and SLW of AMF-inoculated (or non-inoculated) seedlings were 66.53% (61.19%) and 1.5-fold (1.6-fold) of those under WW treatments, respectively. The SLA of mycorrhizal inoculated seedlings was significantly higher under WW and MD stress, while the SLW was significantly lower than that of non-mycorrhizal treated seedlings under the three water treatments ( Figure 4).

ROS (H2O2 and O2 − ) Contents
Regardless of the inoculation status, the O2 -and H2O2 content of C. bungei leaves and roots increased gradually with the intensification of drought (with the exception of H2O2 in roots under MD stress) ( Figure 5). Moreover, the O2 − and H2O2 content of the inoculated (or non-inoculated) showed no significant difference between WW and drought stress (p > 0.05) ( Figure 5C,D). The above parameters were significantly higher (2.64-and 2.55-fold, respectively) in the leaves of non-inoculated seedlings under SD stress, than those under WW treatments ( Figure 5A,B non-inoculated seedlings under SD stress, than those under WW treatments ( Figure 5A,B). However, there was no significant difference in the ROS content of inoculated seedlings among different water treatments (p > 0.05). Under both WW and drought stress conditions, neither H2O2 nor O2 -were significantly different between the roots of mycorrhizal and non-mycorrhizal treated seedlings, while their contents were significantly reduced (to 45.70% and 59.25% of those in the control treatments, respectively) in the leaves under SD stress.

Root System Architecture
Drought stress significantly inhibited root growth, but AMF inoculation significantly improved root development ( Figure 6). Regardless of the inoculation status, all root morphological parameters-except the average root diameter-decreased with an increase in drought stress ( Table  4). None of the total root length, surface area, projected area, root volume, or branching number were significantly different under MD in non-inoculated seedlings from those under WW, but they all significantly decreased under SD conditions. In addition, the total root tips of non-inoculated seedlings were reduced significantly, while the length and surface area of fine roots (0 < d ≤ 0.5 mm) were not significantly different among the three watering conditions (p > 0.05). Most of the root morphological parameters, such as total root length, surface area, volume, root tips, length, and surface area of fine roots (0 < d ≤ 0.5 mm) decreased significantly under drought stress, while total root projected area and branching number were reduced notably only under SD stress. The average root diameter of either inoculated or non-inoculated seedlings did not change significantly among NAM, non-AMF-inoculated; AM, AMF-inoculated; WW, well-watered; MD, moderate drought; SD, severe drought; DS, drought stress. Different lowercase letters above the bars indicate significant differences (p < 0.05) among treatments. Values are means ± SE (n = 3). Two-way ANOVA output: ns, not significant; * p < 0.05; ** p < 0.01.

Root System Architecture
Drought stress significantly inhibited root growth, but AMF inoculation significantly improved root development ( Figure 6). Regardless of the inoculation status, all root morphological parameters-except the average root diameter-decreased with an increase in drought stress ( Table 4). None of the total root length, surface area, projected area, root volume, or branching number were significantly different under MD in non-inoculated seedlings from those under WW, but they all significantly decreased under SD conditions. In addition, the total root tips of non-inoculated seedlings were reduced significantly, while the length and surface area of fine roots (0 < d ≤ 0.5 mm) were not significantly different among the three watering conditions (p > 0.05). Most of the root morphological parameters, such as total root length, surface area, volume, root tips, length, and surface area of fine roots (0 < d ≤ 0.5 mm) decreased significantly under drought stress, while total root projected area and branching number were reduced notably only under SD stress. The average root diameter of either inoculated or non-inoculated seedlings did not change significantly among the three water treatments. Compared with those of the non-inoculated seedlings under WW conditions, all root morphological parameters of mycorrhizal inoculated seedlings were significantly higher than those of non-mycorrhizal treated seedlings, except the average root diameter, which decreased significantly (by 29.52%) ( Table 4). Under MD stress, the total root volume, the number of root tips, branching number, and the surface area of fine roots (0 < d ≤ 0.5 mm) were significantly higher in the inoculated vs. the non-inoculated seedlings (1.41-, 1.56-, 1.45-, and 1.89-fold, respectively) ( Table 4). However, under SD stress, only the root branching number of the mycorrhizal inoculated seedlings was significantly higher (1.84-fold), while the other parameters showed no significant difference between inoculated and control treatments (p > 0.05) ( Table 4).
Forests 2020, 11, x FOR PEER REVIEW 13 of 31 the three water treatments. Compared with those of the non-inoculated seedlings under WW conditions, all root morphological parameters of mycorrhizal inoculated seedlings were significantly higher than those of non-mycorrhizal treated seedlings, except the average root diameter, which decreased significantly (by 29.52%) ( Table 4). Under MD stress, the total root volume, the number of root tips, branching number, and the surface area of fine roots (0 < d ≤ 0.5 mm) were significantly higher in the inoculated vs. the non-inoculated seedlings (1.41-, 1.56-, 1.45-, and 1.89-fold, respectively) (Table 4). However, under SD stress, only the root branching number of the mycorrhizal inoculated seedlings was significantly higher (1.84-fold), while the other parameters showed no significant difference between inoculated and control treatments (p > 0.05) ( Table 4).

Nutrient Absorption and Distribution
N concentration was highest in the roots, stems, and leaves of C. bungei, while K concentration was the lowest. The concentrations of N and P in the aboveground parts, especially leaves, were significantly higher than those in the underground part under the three water treatments (Figure 7). Compared with the WW conditions, drought stress significantly reduced the N, P, and K concentrations in leaves, N in the stems of non-mycorrhizal treated seedlings, the N, P, and K concentrations in the roots, N in the leaves and stems of the mycorrhizal inoculated seedlings, and increased the concentration of K in the stems of both mycorrhizal and non-mycorrhizal treated seedlings, but had no significant effect on the P concentration in the stem (p > 0.05). Moderate drought stress significantly increased the N and P concentrations in the roots of non-mycorrhizal treated seedlings but decreased them under SD stress. Compared with those in the control treatments, root, stem, and leaf N and P concentrations and leaf K were all higher in inoculated seedlings under both WW and drought stress conditions (except for N concentration in stems, and P and K in leaves, under WW conditions) (Figure 7). N concentration reached a significant difference under the three water conditions, while P concentration in leaves and roots was significantly different under SD stress (p < 0.05). In addition, AMF inoculation significantly reduced K concentration in the stems and roots of C. bungei seedlings under drought stress (p < 0.05). treatments, root, stem, and leaf N and P concentrations and leaf K were all higher in inoculated seedlings under both WW and drought stress conditions (except for N concentration in stems, and P and K in leaves, under WW conditions) (Figure 7). N concentration reached a significant difference under the three water conditions, while P concentration in leaves and roots was significantly different under SD stress (p < 0.05). In addition, AMF inoculation significantly reduced K concentration in the stems and roots of C. bungei seedlings under drought stress (p < 0.05).

Phytohormones Levels
Phytohormone contents (ABA, IAA, ZT, GA3) and proportions (IAA/ABA, ZT/ABA, GA3/ABA) increased initially in the C. bungei leaves, after which they decreased with the aggravation of drought stress, reaching the maximum under MD stress (except ZT/ABA without inoculation), regardless of the AMF inoculation status (Figure 8). Under MD stress, the ABA, IAA, GA3, and ZT contents in non-mycorrhizal treated seedlings, and the IAA and GA3 contents in mycorrhizal inoculated seedlings were significantly (p < 0.05) higher than those under WW or SD conditions. In

Phytohormones Levels
Phytohormone contents (ABA, IAA, ZT, GA 3 ) and proportions (IAA/ABA, ZT/ABA, GA 3 /ABA) increased initially in the C. bungei leaves, after which they decreased with the aggravation of drought stress, reaching the maximum under MD stress (except ZT/ABA without inoculation), regardless of the AMF inoculation status (Figure 8). Under MD stress, the ABA, IAA, GA 3 , and ZT contents in non-mycorrhizal treated seedlings, and the IAA and GA 3 contents in mycorrhizal inoculated seedlings were significantly (p < 0.05) higher than those under WW or SD conditions. In addition, under drought stress, the ZT/ABA and GA 3 /ABA ratios were significantly higher in the leaves of non-mycorrhizal treated seedlings than under WW conditions. These results showed that drought stress could stimulate the accumulation of hormones to a certain extent, could change their proportion, and had an overall greater impact on these growth regulators in the leaves of non-mycorrhizal rather than mycorrhizal inoculated seedlings. Under both WW and drought stress, the ABA and ZT contents, as well as the ZT/ABA ratio, were lower in the leaves of inoculated compared to the non-inoculated seedlings ( Figure 8A,C,F). IAA and GA 3 contents, and the IAA/ABA and GA 3 /ABA ratios, on the other hand, were higher than those under control treatments (except the IAA content under MD stress) ( Figure 8B,D,E,G). Thus, hormone contents and proportions showed significant differences between inoculated and control treatments (p < 0.05) under drought stress.

GRSP Content
Under drought stress, there was a slight decrease in the EE-GRSP content in the non-inoculated rhizosphere soil, while T-GRSP (both in inoculated and control treatments) and EE-GRSP (in the inoculated treatment) were higher than those under WW conditions ( Figure 9). However, there was no significant difference in the EE-GRSP and T-GRSP contents among the three water treatments, regardless of the inoculation status (p > 0.5), which indicated that water stress had little effect on the GRSP content in the rhizosphere soil of C. bungei. Compared with the non-inoculated control, the EE-GRSP and T-GRSP contents were higher in the AMF-inoculated rhizosphere soils vs. the non-inoculated soils, under both WW and drought stress conditions (Figure 9). EE-GRSP content increased significantly-by 1.79-fold (in MD) and 1.46-fold (in SD), compared to the non-inoculated control-under drought stress, while the T-GRSP content increased significantly under WW and SD (1.15-and 1.12-fold that of control), respectively. addition, under drought stress, the ZT/ABA and GA3/ABA ratios were significantly higher in the leaves of non-mycorrhizal treated seedlings than under WW conditions. These results showed that drought stress could stimulate the accumulation of hormones to a certain extent, could change their proportion, and had an overall greater impact on these growth regulators in the leaves of non-mycorrhizal rather than mycorrhizal inoculated seedlings. Under both WW and drought stress, the ABA and ZT contents, as well as the ZT/ABA ratio, were lower in the leaves of inoculated compared to the non-inoculated seedlings ( Figure 8A,C,F). IAA and GA3 contents, and the IAA/ABA and GA3/ABA ratios, on the other hand, were higher than those under control treatments (except the IAA content under MD stress) ( Figure 8B,D,E,G). Thus, hormone contents and proportions showed significant differences between inoculated and control treatments (p < 0.05) under drought stress. GRSP content in the rhizosphere soil of C. bungei. Compared with the non-inoculated control, the EE-GRSP and T-GRSP contents were higher in the AMF-inoculated rhizosphere soils vs. the non-inoculated soils, under both WW and drought stress conditions (Figure 9). EE-GRSP content increased significantly-by 1.79-fold (in MD) and 1.46-fold (in SD), compared to the non-inoculated control-under drought stress, while the T-GRSP content increased significantly under WW and SD (1.15-and 1.12-fold that of control), respectively.

Soil Aggregates
Regardless of the AMF status, the soil particle size in the C. bungei rhizosphere was mainly distributed in the 0.01-0.25 mm range (micro-aggregates), accounting for 68%-72%, followed by 0.25-0.5 mm (macro-aggregates), 13%-18%, and 0.5-1.0 mm (macro-aggregates), which was the smallest proportion ( Figure 10). With an increase in drought stress, the percentage of soil aggregates with different particle sizes was not significant in the rhizosphere of both inoculated and control treatments. Compared with the control treatment, AMF inoculation reduced the proportion of soil micro-aggregates (0-0.25 mm), but increased the proportion of macro-aggregates (0.25-0.50 mm) under both WW and MD stress (Figure 10), indicating that AMF can promote the formation of soil micro-aggregates into macro-aggregates to a certain extent.

Soil Aggregates
Regardless of the AMF status, the soil particle size in the C. bungei rhizosphere was mainly distributed in the 0.01-0.25 mm range (micro-aggregates), accounting for 68%-72%, followed by 0.25-0.5 mm (macro-aggregates), 13%-18%, and 0.5-1.0 mm (macro-aggregates), which was the smallest proportion ( Figure 10). With an increase in drought stress, the percentage of soil aggregates with different particle sizes was not significant in the rhizosphere of both inoculated and control treatments. Compared with the control treatment, AMF inoculation reduced the proportion of soil micro-aggregates (0-0.25 mm), but increased the proportion of macro-aggregates (0.25-0.50 mm) under both WW and MD stress (Figure 10), indicating that AMF can promote the formation of soil micro-aggregates into macro-aggregates to a certain extent.

Correlation Between AMF Colonization and Other Parameters
We carried out Pearson correlation analysis between AMF colonization and the other parameters to clarify the potential mechanisms surrounding mycorrhizal colonization and drought stress response. Results are presented in Table 5. Significant positive correlations were found between the AMF colonization and seedling growth parameters (except root biomass and SLW), root morphological parameters (except average root diameter), gas exchange parameters (Pn, Gs, WUE),

Correlation Between AMF Colonization and Other Parameters
We carried out Pearson correlation analysis between AMF colonization and the other parameters to clarify the potential mechanisms surrounding mycorrhizal colonization and drought stress response. Results are presented in Table 5. Significant positive correlations were found between the AMF colonization and seedling growth parameters (except root biomass and SLW), root morphological parameters (except average root diameter), gas exchange parameters (Pn, Gs, WUE), photosynthetic pigment concentration, nutrient concentration (N in leaves and roots), and hormone levels (GA 3 ) (r ranged from 0.695 to 0.962, p < 0.01, or 0.05). Positive correlations were also found with EE-GRSP, root biomass, N and P in stems, P and K in roots, IAA, ZT, and ABA in leaves, albeit not to a significant level. In contrast, significant negative correlations were found between AMF colonization and SLW, K concentration in stems, P and O 2 − content in leaves, and H 2 O 2 content in roots (r in the range of −0.971 to −0.702, p < 0.01, or 0.05). Moreover, correlations with root/shoot ratio, average root diameter, leaf K and H 2 O 2 content, and root O 2 − concentration were also negative, but these differences were not significant. T-GRSP, total glomalin-related soil protein; LF, length of fine roots (0 < d ≤ 0.5 mm); SAF, surface area of fine roots (0 < d ≤ 0.5 mm); * p < 0.05; ** p < 0.01.

Discussion
AMF promotes the absorption of water and nutrients by plant roots and facilitates water and nutrient transport from the roots to the leaves, leading to enhanced photosynthetic rate and other gas exchange-related traits. Moreover, it improves plant nutrition by increasing the availability and affecting the distribution of nutrients, and promotes the growth of woody tree seedlings (height and basal diameter), biomass accumulation, and stress resistance [8,13,16,43].
The colonization rate of mycorrhizal fungi reflects the degree of infection and affinity between AMF and the host plant [6,44]. Under all water conditions, the mycorrhizal colonization rate was over 60%, indicating a relatively high affinity between the selected AMF and C. bungei. The degree of arbuscule staining was deeper under drought stress than the WW treatments, which could be due to the stimulation of branching or thickening of the arbuscular by the drought stress, resulting in morphological changes. Under SD stress, however, mycorrhizal colonization rate decreased significantly, which likely had an impact on the microbial activities, and affected the function of mycorrhizal symbionts to a certain extent. These findings are in line with those reported in Cyclobalanopsis glauca seedlings [45], black locust (Robinia pseudoacacia L.) [46], and trifoliate orange (Poncirus trifoliata) [47]. This decrease may be due to the low carbon availability in the host plants under drought stress, or because drought stress could have inhibited spore germination and hyphal growth in the rhizosphere soil [1,45]. However, the response of different AMF to drought stress may be significantly different because of their diverse stress resistance. Chen et al. showed that the mycorrhizal colonization rate of Amorpha fruticosa by Funneliformis mosseae and F. constrictum was significantly higher under drought stress than in WW conditions [48]. In addition, previous studies on the hybrid poplar [49], Astragalus adsurgens Pall [50], and Malus hupehensis [3] reported that the AMF colonization rate was not affected by water deficit conditions. These phenomena indicate that the mycorrhizal colonization rate depends not only on the external environment factors, but also on the fungi and the colonized plant species [51].
Plant adaptability to the external environment and the influence of the environment on plant growth are to some extent reflected by biomass. Drought stress inhibits plant root extension along with stem and leaf growth, which ultimately results in a reduced biomass. Thus, biomass is the most intuitive index to assess plant performance under drought [52]. In this study, drought stress reduced the growth parameters (plant height, basal diameter, leaf area and biomass) of the C. bungei seedlings. However, colonization by AMF had positive effects on the growth parameters of C. bungei seedlings under both WW and drought stress. Similar effects were previously reported in other species; Sophora davidii [53], black locust [8], Cyclobalanopsis glauca [45] and Arizona cypress (Cupressus arizonica G.) [16]. We recorded significant positive correlations (p < 0.01) between AMF colonization and morphological parameters, which corroborate the above effects. This growth benefit is likely due to the large extracellular hyphae AMF network, which transports water and nutrients to roots, increasing their absorption range and helping plants cope with the drought stress [8,52]. However, the relationship between AMF and plants under drought stress is very complex. Other studies found no detectable effect of AMF colonization [54,55], or even a negative effect on plant growth under drought stress [56], possibly for a variety of reasons, such as high root density, nutrient depletion, or insufficient light [57,58]. These indicate a myriad of interspecific differences between different AMF species and plants, and that their symbiosis may also be affected and regulated by environmental conditions [59].
Pronounced changes in the biomass allocation pattern could be interpreted as an adaptive strategy to overcome water and/or nutrient depletion more quickly [60,61]. In this study, water status significantly affected biomass allocation, regardless of the inoculation status. Under MD stress, C. bungei seedlings allocated more biomass to their roots, increasing their root/shoot ratio. This adaptive response indicates a preferential distribution of carbohydrates to the roots, probably to increase water absorption and to accelerate root growth under water scarcity. This survival strategy is complemented by a decrease in leaf area and the induction of leaf senescence and abscission, forming a drought resistance strategy that is consistent with results of previous studies [59,60]. However, inoculation with R. intraradices promoted biomass accumulation in the stems and leaves of C. bungei seedlings but inhibited it in the root. Under the same drought stress level, both the biomass ratio of inoculated seedlings to roots and the root/shoot ratio were significantly lower than those in the non-inoculated seedlings. A negative correlation between AMF colonization and root/shoot ratio (r = −0.658, p > 0.05) in mycorrhizal inoculated seedlings was also observed. This indicated that AMF changed the biomass allocation in the seedlings. Under drought stress, roots of the symbiotic seedlings could absorb relatively large amounts of water or could allocate a relatively high proportion of carbohydrates for the aboveground growth, thus alleviating the restriction imposed by water scarcity on the growth of C. bungei seedlings. Similar results were also obtained for mycorrhizal inoculated Broussonetia papyrifera [60], Bauhinia faberi var. microphylla [4], Retama monosperma, and Acacia gummifera [58] under water deficit conditions. In this study, the biomass allocation pattern of non-mycorrhizal treated seedlings under WW conditions was consistent with that of inoculated seedlings under drought stress (leaf > root > stem), which was also the most direct evidence that AMF could promote the absorption of water in C. bungei seedlings and alleviate drought stress.
Stomata are channels for gas exchange and transpiration control between plants and the external environment. They effectively control water loss and CO 2 absorption [62]. Stomatal behavior is closely related to the soil and climate conditions of plants, especially water and nutrient status. Stomata function is the key to modulate transpiration and photosynthesis in leaves [63]. Both inoculated and non-inoculated C. bungei seedlings subjected to drought stress had strongly reduced Gs, and consequently reduced photosynthesis. Reducing Gs to decrease the transpiration-induced water loss is a key strategy for maintaining adequate levels of tissue hydration and cell expansion to sustain growth and biomass production, even under water deficit conditions [64,65]. Studies have shown the Gs, Tr, and Pn of mycorrhizal inoculated plants to be higher than those of non-mycorrhizal treated plants under water deficit [8,47]. In the current study, mycorrhizal inoculated seedlings displayed higher Pn, Gs, and Tr, but lower Ci concentrations under both WW and drought stress conditions than their non-mycorrhizal treated counterparts, which was similar in woody species, such as black locust [8], P. trifoliata [66], carob (Ceratonia siliqua L.) [65], and M. hupehensis [3]. Studies have also shown that by increasing leaf area and P content, to regulate the physiological status, AMF optimizes the photosynthetic rate in the host plants [12]. Our results of a higher leaf area and P content in the inoculated seedlings also confirmed this. On the other hand, the higher Pn of the inoculated plants under drought stress was also caused by the increase in Gs, indicating that the stomatal opening of inoculated plants was prolonged, to the benefit of both gas exchange and photosynthesis [15]. This was also indicated by the correlations found between Pn and Gs (r = 0.967, p < 0.01) in the present study (data not shown in table).
Leaf water status is an important feature for assessing the effects of AMF on drought stress, as it can provide information on how to evaluate the physiological and biochemical changes in plants [67]. A higher WUE is conducive to moving water to the evaporating surface and keeping leaf stomata open [8]. Therefore, a higher WUE in the mycorrhizal inoculated plants under drought stress is an indicator of their higher photosynthetic performance [65]. This was also confirmed by our results. Indeed, AMF improved WUE of the C. bungei seedlings, irrespective of water status. A positive and highly significant correlation (r = 0.809, p < 0.01) was also evident between AMF colonization and WUE. Similarly, inoculation with AMF enhanced WUE in S. davidii [53], black locust [8] and hybrid poplar [49] under drought stress conditions. The positive effects of AMF on WUE can be attributed to the absorption of water and nutrients from the soil by the external hyphae, the regulation of Gs through hormonal signals, optimization of osmotic regulation, or the enhancement of root hydraulic conductivity [8,68,69]. However, contrary to the results of most studies, Doubková et al. [70] documented that inoculation with AMF in Knautia arvensis had no positive effect on WUE. In fact, the WUE of mycorrhizal inoculated plants was lower than that of the non-mycorrhizal treated plants under WW conditions. This possible discrepancy was explained by larger root systems connected to the extraradical hyphal network, that made the use of soil water more efficient.
Chlorophyll is one of the most important pigments in higher plants. It is crucial for photosynthesis, through which plants obtain energy from light [3,71]. Therefore, chlorophyll content is an important index for the evaluation of photosynthetic intensity, efficiency, and environmental stress [8]. Several studies have shown that inoculation with AMF can notably increase the chlorophyll content of plants under drought stress [22,46,47,72]. Similarly, our study also shows that inoculation with AMF increased the chlorophyll a, chlorophyll b, and carotenoid concentrations, and enhanced the utilization of light energy in the leaves of C. bungei. We found significant positive correlations between AMF colonization and chlorophyll a (r = 0.800, p < 0.01), chlorophyll b (r = 0.757, p < 0.05), and carotenoids (r = 0.757, p < 0.05), which further confirmed the beneficial role of mycorrhizal fungi on the photosynthetic pigments. The increase in chlorophyll content may be related to the increase in P and Mg uptake by AMF [71]. Indeed, AMF promoted the absorption of P in leaves also in our study. Moreover, increased chlorophyll and gas exchange parameters indicate that AMF increased photosynthetic carbon (C) fixation, which would not only stimulate AMF growth, but would also increase the photosynthetic rate because of the reciprocal C-P relationship between AMF and their host plants [71].
SLA is the main leaf functional trait connected with the nutrient retention capacity, which reflects the plant's ability to exploit and utilize the environmental resources [64,73]. SLA directly affects the assimilation of CO 2 and the light, water, and nutrient utilization efficiency [64]. Our results showed that seedlings inoculated with R. intraradices had markedly increased SLA and decreased SLW under both WW and drought stress conditions. Moreover, the correlation analysis revealed a significant positive relationship between AMF colonization and SLA (r = 0.954, p < 0.01) and Pn (r = 0.695, p < 0.05). Increased SLA enhances the photosynthetic area and thus the leaf photosynthetic rate, which is conducive to the accumulation of photosynthetic products [61]. Similar results were also found in Cynophalla flexuosa [64] and wheat (Triticum spp.) [74], confirming that mycorrhizal inoculated plants are beneficial for the adaptation to arid environments. In addition, SLA is also related to photosynthetic N allocation, meaning that a higher SLA would enable a higher N allocation ratio to photosynthesis [75]. A significantly positive correlation (r = 0.882, p < 0.01) between SLA and leaf N concentration was also confirmed by our results (data not shown in table). Therefore, a high SLA may help plants increase their ability to acquire available resources for better growth and reproduction [75]. In contrast, Hernández-Ortega et al. [76] reported that Melilotus albus inoculated with the Glomus Zac-19 AMF-inoculum had lower SLA on diesel-contaminated substrate. Busquets et al. [77] found that the SLW of Anthyllis cytisoides inoculated with G. mosseae was significantly higher than that of the non-inoculated plants following drought. This may be due to an improved cell wall resistance and leaf thickening (typical xerophytic characteristics), as a result of the symbiotic relationships [73,77].
In plants, ROS production and removal (e.g., O 2 − and H 2 O 2 ) are in a dynamic equilibrium state to maintain normal physiological and metabolic functions [78]. When plants are subjected to drought stress, this balance is disrupted, leading to an overproduction of ROS that will result in oxidative stress and consequently membrane lipid peroxidation, protein denaturation, nucleic acid chain fracture, cell membrane damage, and even cell death in severe cases [6,15,79]. Avoidance of oxidative stress by preventing ROS accumulation is the most effective way for mycorrhizal inoculated plants to cope with drought stress [80]. p < 0.05) in mycorrhizal inoculated seedlings. These results are similar to those previously reported in black locust [46], C. arizonica [16] and M. hupehensis [3], indicating that mycorrhizal fungi could mitigate drought-induced oxidative stress in plants by reducing the production of ROS, and thus can maintain the integrity of cell membranes and stabilize proteins under drought stress [15,80]. Low ROS accumulation in the mycorrhizal inoculated seedlings may be related to an intensive hyphal and arbuscular growth within the roots [46]. Plant roots not only provide physical support for plants but are also important in providing the material basis for photosynthesis (water and nutrient absorption). Root development directly affects plant growth, performance, and survival [4]. It has been confirmed that AMF inoculation can significantly affect the host root architecture and promote plant growth under drought stress [7,19]. Our results showed significantly higher root length, total surface area, projected area, volume, tips, branching number, length, and surface area of fine roots in the AMF-inoculated C. bungei seedlings under WW conditions. These morphological parameters were also improved under drought stress when compared with the non-AMF seedlings. This is consistent with previous reports from Prunus cerasifera L. [81], P. trifoliata [82], Amygdalus pedunculata [83], and Acacia seyal Del. seedlings [84]. AMF colonization was positively correlated (p < 0.01) with all root morphological parameters (except average root diameter), indicating that likely due to a better root morphology, mycorrhizal inoculated plants have a stronger capacity to adapt to drier soils than their non-mycorrhizal treated counterparts.
Although AMF can affect root plasticity in a number of ways, the most common one is to increase root branching and to generate a larger proportion of roots with small diameters [81,85]. An increased root branching caused by mycorrhizal symbiosis resulted in the enhancement of root hairy zones, which are the main sites for P acquisition [85]. This was also supported by our findings since mycorrhizal symbiosis significantly increased P uptake and root branching in the C. bungei seedlings. The average root diameter is a parameter that reflects the absorption efficiency of the root system, and to a certain extent reflects the soil volume that can be occupied by the root system per unit biomass [60]. In this study, under the same water conditions, the average root diameter of the inoculated seedlings was lower than that of the non-inoculated ones. This is similar to the results of Zhang et al. [86] in Cyclobalanopsis glauca seedlings inoculated with AMF, indicating that AMF inoculation can induce inoculated plant roots to become smaller, thus making it easier for the roots to absorb water and nutrients. The above results indicate that AMF can effectively regulate the root morphological parameters of the host plants to improve the absorption of water and nutrients, thereby promoting growth, root development and stress resistance. However, some studies have suggested that inoculation had no detectable effect on the root morphology of Bidens frondosa L., such as root length, volume, or surface area, under three levels of water availability [55]. This may be possible, as the regulation of AMF on plant root morphology is dependent on nutrient and water availability, the AMF species, and the host species itself [81,86,87].
As basic macronutrients, N, P and K play important roles in plant growth and metabolism [88]. Nutrient distribution and accumulation are results of the interaction between plant and environmental factors and reflect not only the adaptability of plants to environmental factors, but also the impact of the environment on plants [89]. Previous studies suggested that the promotion of nutrient uptake by AMF is an important physiological mechanism of the mycorrhiza-induced drought resistance of the host plant [2,7]. Yang et al. [8] discovered that black locust seedlings inoculated with F. mosseae and R. intraradices had enhanced leaf C, N, and P concentrations under both WW and drought stress conditions. Zhang et al. [90] reported that inoculation with A. laevis, G. mosseae, and G. caledonium increased N, P, K, calcium (Ca), Mg, zinc (Zn), and copper (Cu) uptake in loquat (Eriobotrya japonica Lindl.) seedlings. Zarik et al. [91] conducted similar research on Cupressus atlantica, and the results suggested that inoculation with AMF significantly improved the uptake of minerals independently of water regimes. A recent study by Al-Arjani et al. [22] also found that under drought stress, mycorrhizal inoculated Ephedra foliata seedlings had higher uptake of essential nutrients, such as K, Mg, and Ca, than the non-inoculated seedlings. In the present study, AMF inoculation increased the N and P concentrations in roots, stems, and leaves, and K concentration in the leaves of C. bungei. These data suggest that the higher drought resistance of C. bungei seedlings following AMF inoculation could partly be attributed to the improved nutrient uptake by AMF, as nutrient deficiency would greatly contribute to stress. Moreover, the enhanced drought resistance is likely connected to the improved photosynthetic performance of C. bungei by the AMF inoculation.
Phytohormones, including ABA, GA 3 , CTKs, and IAA, are known to act as signaling molecules and to be involved in various developmental processes of the mycorrhizal symbiosis [10,[92][93][94]. AMF could improve drought resistance by regulating endogenous hormones after symbiosis with host plants [6,7]. Typically, ABA is the most involved hormone in the drought responses, mainly because stomatal behavior is closely related to changes in the ABA levels [66,94,95]. According to Ouledali et al. [94], ABA content increased in the leaves of AMF-inoculated olive tree (Olea europaea L.) seedlings under MD stress, but decreased significantly under SD stress, compared with that in the non-inoculated seedlings. Some studies suggested ABA content to be significantly higher in drought-stressed mycorrhizal inoculated plants than in non-mycorrhizal treated plants [19,66,96]. Their view was that plants respond to drought by synthesizing ABA to promote stomatal closure, thereby reducing water loss. However, contrary to their findings, our study showed that AMF-inoculated C. bungei seedlings exhibited considerably lower ABA levels in leaves under drought stress, which supports other findings [18,97]. This apparent discrepancy might be because of the delayed stomatal closure in mycorrhizal inoculated seedlings by the reduction of ABA levels, which improved the gas exchange capacity and enhanced the photosynthetic effect. Our previous results also supported the hypothesis that seedlings inoculated with AMF had higher Gs and Pn than non-inoculated seedlings, regardless of the water status.
In addition, CTKs may also affect leaf senescence and nutrient mobilization [98]. These hormones, alone or in combination, regulate plant growth and development [22,66]. Our results indicated that inoculation with AMF considerably augmented IAA (with the exception of MD stress) and GA 3 levels, but reduced zeatin (ZT, a cytokinin) in the leaves of C. bungei seedlings, exposed either to WW or drought stress conditions. This is in accordance with previous findings [22,66,99]. It implies that AMF could postpone leaf senescence and cause a quick adaptation to drought stress by influencing the auxin signal transduction. Regardless of the water balance, AMF inoculation significantly increased the IAA/IBA and GA 3 /ABA values in the leaves of C. bungei, indicating that the positive effects of C. bungei on growth were strengthened under drought stress. As proposed by Al-Arjani et al. [22], the decrease in GA 3 levels and the increase in ABA levels are supposed to be the key determinants of stress tolerance. However, in our study, the opposite results were obtained. This, however, should not be alarming, as plants inoculated with AMF had a disposition to maintain both growth promotion and stress adaptation, and the final result of the hormone interaction seemed to have a minimal impact on cell metabolism and the increased stress resistance [100]. Such changes in phytohormones by AMF may provide important clues to improve the drought resistance of the host plants.
AMF hyphae and spores release GRSP, a type of iron-containing, heat-stable glycoprotein [7,11] while growing with the root system. GRSP is an important binding agent for the formation of soil aggregates [50,101]. It is abundant in soils, usually in the range of several to 10 mg·g −1 soil, which helps combine tiny particles into small aggregates of different sizes [11]. In addition, to ensure the transportation of water into the mycelium, it can seal the AMF extraradical hyphae in the rhizosphere soil [7], improve soil environment (enter the soil along with the degradation of fungal biomass), which can enhance soil stability, porous structure and moisture penetration, and provide necessary space and improved gas exchange channels for plant root growth [41,50]. Most studies suggest that T-GRSP is a stable glycoprotein accumulated by AMF over a long period of time, while EE-GRSP is a newly produced glycoprotein [102]. In this study, we found no detectable changes in the EE-GRSP and T-GRSP levels in the rhizosphere soil without AMF inoculation upon the aggravation of drought stress. This may be due to the fact that the sterilized substrate was mixed evenly, and AMF colonization did not occur, therefore, no new GRSP could be produced. After inoculation, EE-GRSP and T-GRSP increased under MD stress, indicating that a certain degree of drought stress could promote the production of GRSP, and this increase in the EE-GRSP further led to an increase in T-GRSP.
Wu [103] investigated GRSP in the rhizosphere of pot-cultured P. trifoliata seedlings (without AMF inoculation) and concluded that there was no GRSP without AMF colonization. However, our results showed the presence of GRSP (both EE-GRSP and T-GRSP) in sterilized, non-inoculated soil. This may be because the GRSP having been present in the rhizosphere soil before sterilization, which was also confirmed by Bedini et al. [41]. Indeed, GRSP is very stable in nature and highly persistent in the soil [7,11]. In the current study, under the same water regime, the EE-GRSP and T-GRSP contents increased significantly in the rhizosphere soils inoculated with AMF, compared to those without inoculation, which was in accordance with the results of Ji et al. [50], who reported a significant increase in the GRSP content in the rhizosphere of A. adsurgens Pall., inoculated with AMF, under both WW and drought stress conditions. AMF also elevated GRSP content in the rhizosphere soil of black locust [104]. GRSP concentration in potted soil was significantly higher after inoculation than the initial value without inoculation, indicating a causal relationship between mycorrhizal symbiosis and GRSP content. Our results provide strong evidence for the differences in GRSP production between mycorrhizal and non-mycorrhizal treated plants.
Soil structure refers to the grouping of soil particles into porous compounds, also known as aggregates. Particle size distribution and the stability of soil aggregates impacts not only soil pore distribution but is also related to the movement and storage of water [104,105]. In addition to improving soil structure by producing GRSP, the extraradical mycelia of AMF can directly create the skeletal structure of soil aggregates by physically entangling soil particles, and then form micro-aggregates. Finally, hyphae and the roots can combine these smaller micro-aggregates (<0.25 mm) to form larger and more stable macro-aggregates (>0.25 mm) [101,[104][105][106]. Due to their better air and water permeability and stability towards wind and water erosion, well-aggregated soils are conducive to plant and microbial growth [11]. Zhang et al. [104] reported that soil aggregate stability parameters, such as water stable aggregates, geometric mean diameter, and mean weight diameter were elevated by AMF inoculation in the rhizosphere of black locust. Wu et al. [105] showed that mycorrhizas-enhanced > 2, 1-2, and > 0.25 mm water-stable aggregate fractions in the rhizosphere soil of P. trifoliata seedlings. In the current study, AMF inoculation reduced the proportion of soil micro-aggregates (<0.002, 0.002-0.01, and 0.01-0.25 mm, respectively) and increased the proportion of macro-aggregates (0.25-0.5 mm) under the same water conditions, compared with the control treatments. These indicate that AMF inoculation improved soil structure and maintained soil permeability and water-holding capacity by consolidating small micro-aggregates into large aggregates. We therefore conclude that beyond a direct mineral nutritional effect, AMF colonization-through the effects of GRSP production and hyphal entanglement on soil aggregates-promoted the growth of C. bungei seedlings also indirectly, under drought stress. Although the results obtained in pot experiments cannot represent field conditions, microscopic approaches are helpful to better understand the causal relationship between AMF symbiosis and soil quality. Our findings underscore the link between GRSP and AMF and add new proof to their ability to enhance soil aggregate stability.

Conclusions
This study demonstrated that inoculation of R. intraradices significantly ameliorated the growth of C. bungei seedlings under drought conditions. AMF could enhance photosynthesis and WUE (by increasing leaf area), SLA and photosynthetic pigment concentration, change the biomass allocation in seedlings (by reducing RMR and root/shoot ratio) and promote the growth of aboveground parts (leaves and stems). Moreover, mycorrhizal treatment alleviated drought-induced oxidative stress in plants by reducing the excess generation of ROS (especially H 2 O 2 and O 2 − in leaves). Inoculation with AMF under drought stress dramatically augmented IAA and GA 3 levels, but reduced ABA, and ZT levels in the leaves. Under drought stress, R. intraradices could form a good symbiotic relationship with C. bungei seedlings, and AMF symbiosis indeed improved root morphology and structure, and promoted the absorption of N and P in seedlings, which was conducive to plant adaptation to a dry environment. Furthermore, by increasing the GRSP content and proportion of macro-aggregates (0.25-0.5 mm) in the rhizosphere soil, AMF improved soil structure. Thus, it is feasible to use mycorrhizal biotechnology for vegetation restoration and reconstruction in arid and semi-arid areas.
This article mainly involved physiological and biochemical processes; however, we cannot ignore that the genetics of the AMF samples chosen to treat plants with could have a significant impact on the outcome of the interaction, which should be regarded as another area of uncertainty and future focus. The molecular mechanism of AMF improving drought resistance of C. bungei also needs to be further explored.